Ada Operating System

Ada Operating System

[Dann Corbit]

Has anyone tried to build a complete operating system in Ada?  Does anyone
know of a paper or URL for such an effort?
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
 "The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

Re: Ada Operating System

[Adrian Knoth]

Dann Corbit <dcorbit@connx.com> wrote:

> Has anyone tried to build a complete operating system in Ada?  

www.adaos.org


-- 
mail: adi@thur.de  	http://adi.thur.de	PGP: v2-key via keyserver

Weisst Du - ich bin sooo bloed. (von langer@gmx.net)

Re: Ada Operating System

[chris.danx]

"Dann Corbit" <dcorbit@connx.com> wrote in message
news:a4u9d00vqe@enews2.newsguy.com...
> Has anyone tried to build a complete operating system in Ada?  Does anyone
> know of a paper or URL for such an effort?

www.adaos.org
http://site.voila.fr/ProjetADX/index.htm (in french, if you can't read it
try running it through http://www.freetranslation.com/)

Re: Ada Operating System

[Hugues Jerome]

In article <5Tzc8.111999$H37.15018616@news2-win.server.ntlworld.com>, 
chris.danx wrote:
> 
> "Dann Corbit" <dcorbit@connx.com> wrote in message
> news:a4u9d00vqe@enews2.newsguy.com...
>> Has anyone tried to build a complete operating system in Ada?  Does anyone
>> know of a paper or URL for such an effort?
> 
> www.adaos.org
> http://site.voila.fr/ProjetADX/index.htm (in french, if you can't read it
> try running it through http://www.freetranslation.com/)

Simple translation for ProjetADX :

it is a micro-kernel, built using the facility of the Ada95 language.
early 2001 there were some specs, describing the overall architecture, since
then nothing. unless otherwise stated, this project seems dead ...

rtems is (was ?) also an embedded OS written in Ada (another version was in C)
by the US Army, more info here http://www.rtems.army.mil, perhaps you can 
have other info somewhere else ... 


-- 
Jerome

Re: Ada Operating System

[Dann Corbit]

"Hugues Jerome" <hugues@neraka.rezel.enst.fr> wrote in message
news:slrna75lja.g7h.hugues@neraka.rezel.enst.fr...
> In article <5Tzc8.111999$H37.15018616@news2-win.server.ntlworld.com>,
> chris.danx wrote:
> >
> > "Dann Corbit" <dcorbit@connx.com> wrote in message
> > news:a4u9d00vqe@enews2.newsguy.com...
> >> Has anyone tried to build a complete operating system in Ada?  Does
anyone
> >> know of a paper or URL for such an effort?
> >
> > www.adaos.org

This one has specs only at this time.

> > http://site.voila.fr/ProjetADX/index.htm (in french, if you can't read
it
> > try running it through http://www.freetranslation.com/)

I have a couple years of French, but the translation below was handy anyway.

> Simple translation for ProjetADX :
>
> it is a micro-kernel, built using the facility of the Ada95 language.
> early 2001 there were some specs, describing the overall architecture,
since
> then nothing. unless otherwise stated, this project seems dead ...
>
> rtems is (was ?) also an embedded OS written in Ada (another version was
in C)
> by the US Army, more info here http://www.rtems.army.mil, perhaps you can
> have other info somewhere else ...

Looks like the new versions are in C.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
 "The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

Re: Ada Operating System

[eunux]

www.eunux.org ?

"Dann Corbit" <dcorbit@connx.com> wrote in message
news:a4u9d00vqe@enews2.newsguy.com...
> Has anyone tried to build a complete operating system in Ada?  Does anyone
> know of a paper or URL for such an effort?
> --
> C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
>  "The C-FAQ Book" ISBN 0-201-84519-9
> C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm
>
>

Re: Ada Operating System

[Stephen Leake]

hugues@neraka.rezel.enst.fr (Hugues Jerome) writes:

> rtems is (was ?) also an embedded OS written in Ada (another version
> was in C) by the US Army, more info here http://www.rtems.army.mil,
> perhaps you can have other info somewhere else ...

Currently maintained by OAR Corp, in C. See www.oarcorp.com. 

-- 
-- Stephe

Re: Ada Operating System

[Joel Sherrill]

"Dann Corbit" <dcorbit@connx.com> wrote in message news:<a4uqu402315@enews2.newsguy.com>...
> "Hugues Jerome" <hugues@neraka.rezel.enst.fr> wrote in message
> news:slrna75lja.g7h.hugues@neraka.rezel.enst.fr...
> > In article <5Tzc8.111999$H37.15018616@news2-win.server.ntlworld.com>,
> > chris.danx wrote:
> > >
> > > "Dann Corbit" <dcorbit@connx.com> wrote in message
> > > news:a4u9d00vqe@enews2.newsguy.com...
> > >> Has anyone tried to build a complete operating system in Ada?  Does
>  anyone
> > >> know of a paper or URL for such an effort?
> > >
> > > www.adaos.org
> 
> This one has specs only at this time.
> 
> > > http://site.voila.fr/ProjetADX/index.htm (in french, if you can't read
>  it
> > > try running it through http://www.freetranslation.com/)
> 
> I have a couple years of French, but the translation below was handy anyway.
> 
> > Simple translation for ProjetADX :
> >
> > it is a micro-kernel, built using the facility of the Ada95 language.
> > early 2001 there were some specs, describing the overall architecture,
>  since
> > then nothing. unless otherwise stated, this project seems dead ...
> >
> > rtems is (was ?) also an embedded OS written in Ada (another version was
>  in C)
> > by the US Army, more info here http://www.rtems.army.mil, perhaps you can
> > have other info somewhere else ...

I don't think there is a public Army RTEMS site any longer.  The 
current RTEMS source is at ftp.oarcorp.com.  The last Ada
release is also there since more than a few people had asked for it.
 
> Looks like the new versions are in C.

This is correct.  V3.2.1 was the last version in which the C
and Ada implementations were feature-wise in sync.  About this
time, the GNAT project was started and we decided that it made
more sense to provide Ada support for RTEMS by focusing on 
providing a free full-featured embedded run-time for GNAT.  

This is accomplished via as complete an implementation 
of POSIX 1003.1b as is possible without actually having 
UNIX processes.  We made the decision because we felt this 
resulted in a solution more appealing to more people.  I have
always enjoyed the irony that C/C++ supporters enhancing RTEMS 
are improving it as an Ada run-time. :)

--joel
Joel Sherrill, Ph.D.             Director of Research & Development
joel@OARcorp.com                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
Support Available                (256) 722-9985

Re: Ada Operating System

[Hyman Rosen]

Joel Sherrill wrote:
> I have always enjoyed the irony that C/C++ supporters enhancing RTEMS 
> are improving it as an Ada run-time. :)

It's not really ironic. Remember that, while the Ada world thinks of
C/C++ as the enemy and the competition, the C/C++ world barely realizes
that Ada exists, and certainly wouldn't mind giving it a helping hand.

Re: Ada Operating System

[tony gair]

"Hyman Rosen" <hyrosen@mail.com> wrote in message
news:3C73F480.5010806@mail.com...
> Joel Sherrill wrote:
> > I have always enjoyed the irony that C/C++ supporters enhancing RTEMS
> > are improving it as an Ada run-time. :)
>
> It's not really ironic. Remember that, while the Ada world thinks of
> C/C++ as the enemy and the competition, the C/C++ world barely realizes
> that Ada exists, and certainly wouldn't mind giving it a helping hand.
>

I think the C++ /Ada rivally is born of some peoples need to gang fight.
I would look like at C++ and Ada not as friends but as occasional lovers
who don't use any contraception.
Ada and C++ offspring  include
1) GNAT translator
2) Gtkada
3) ......

GNAT and GTKada would not exist in their present useful form without
C++

and many C++ programmers would not exist if their planes or trains or
automobiles
had not used Ada with their subsystems.

PS you could also say there would be a lot more innocent civilians around
without
Ada or C++ supported weapon systems.

Re: Ada Operating System

[Chad R. Meiners]

So how exactly is GNAT's existance dependent on C++?

-CRM

"tony gair" <tonygair@nounsolitations.btinternet.com> wrote in message
news:a533vq$sts$1@knossos.btinternet.com...
>
> "Hyman Rosen" <hyrosen@mail.com> wrote in message
> news:3C73F480.5010806@mail.com...
> > Joel Sherrill wrote:
> > > I have always enjoyed the irony that C/C++ supporters enhancing RTEMS
> > > are improving it as an Ada run-time. :)
> >
> > It's not really ironic. Remember that, while the Ada world thinks of
> > C/C++ as the enemy and the competition, the C/C++ world barely realizes
> > that Ada exists, and certainly wouldn't mind giving it a helping hand.
> >
>
> I think the C++ /Ada rivally is born of some peoples need to gang fight.
> I would look like at C++ and Ada not as friends but as occasional lovers
> who don't use any contraception.
> Ada and C++ offspring  include
> 1) GNAT translator
> 2) Gtkada
> 3) ......
>
> GNAT and GTKada would not exist in their present useful form without
> C++
>
> and many C++ programmers would not exist if their planes or trains or
> automobiles
> had not used Ada with their subsystems.
>
> PS you could also say there would be a lot more innocent civilians around
> without
> Ada or C++ supported weapon systems.
>
>
>

Re: Ada Operating System

[Larry Kilgallen]

In article <a533vq$sts$1@knossos.btinternet.com>, "tony gair" <tonygair@nounsolitations.btinternet.com> writes:

> PS you could also say there would be a lot more innocent civilians around
> without
> Ada or C++ supported weapon systems.

I am not convinced that ill-controlled explosives are better than
well-controlled explosives.  Plenty of explosive were used in the
US Civil War (1861-1865) -- as much as possible, despite the lack
of fine-grained control mechanisms.

Re: Ada Operating System

[David Starner]

On Thu, 21 Feb 2002 15:33:15 +0000 (UTC), tony gair <tonygair@nounsolitations.btinternet.com> wrote:
> GNAT and GTKada would not exist in their present useful form without
> C++

Why? Both GCC and GTK are pure C code.

-- 
David Starner - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Ada Operating System

[tony gair]

"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:$D13NMeU$AWl@eisner.encompasserve.org...
> In article <a533vq$sts$1@knossos.btinternet.com>, "tony gair"
<tonygair@nounsolitations.btinternet.com> writes:
>
> > PS you could also say there would be a lot more innocent civilians
around
> > without
> > Ada or C++ supported weapon systems.
>
> I am not convinced that ill-controlled explosives are better than
> well-controlled explosives.  Plenty of explosive were used in the
> US Civil War (1861-1865) -- as much as possible, despite the lack
> of fine-grained control mechanisms.

They may be well controlled by the software, but they are certainly ill
controlled
by vicous vile, cowards. Come the world wide revolution they will each be
tied in a chair and forced to watch a looped tape of an episode of Dallas,
for the rest of their lives...........

Re: Ada Operating System

[Larry Kilgallen]

In article <a543mf$s4e$1@paris.btinternet.com>, "tony gair" <tonygair@nounsolitations.btinternet.com> writes:
> 
> "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
> news:$D13NMeU$AWl@eisner.encompasserve.org...
>> In article <a533vq$sts$1@knossos.btinternet.com>, "tony gair"
> <tonygair@nounsolitations.btinternet.com> writes:
>>
>> > PS you could also say there would be a lot more innocent civilians
> around
>> > without
>> > Ada or C++ supported weapon systems.
>>
>> I am not convinced that ill-controlled explosives are better than
>> well-controlled explosives.  Plenty of explosive were used in the
>> US Civil War (1861-1865) -- as much as possible, despite the lack
>> of fine-grained control mechanisms.
> 
> They may be well controlled by the software, but they are certainly ill
> controlled
> by vicous vile, cowards. Come the world wide revolution they will each be
> tied in a chair and forced to watch a looped tape of an episode of Dallas,
> for the rest of their lives...........

I gather you live in a backward country that does not yet have
broadcasts of Jenny Jones, Sally Jessy Raphael and Jerry Springer :-)

Re: Ada Operating System

[Adrian Hoe]

"tony gair" <tonygair@nounsolitations.btinternet.com> wrote in message news:<a533vq$sts$1@knossos.btinternet.com>...
> "Hyman Rosen" <hyrosen@mail.com> wrote in message
> news:3C73F480.5010806@mail.com...
> > Joel Sherrill wrote:
> > > I have always enjoyed the irony that C/C++ supporters enhancing RTEMS
> > > are improving it as an Ada run-time. :)
> >
> > It's not really ironic. Remember that, while the Ada world thinks of
> > C/C++ as the enemy and the competition, the C/C++ world barely realizes
> > that Ada exists, and certainly wouldn't mind giving it a helping hand.
> >
> 
> I think the C++ /Ada rivally is born of some peoples need to gang fight.
> I would look like at C++ and Ada not as friends but as occasional lovers
> who don't use any contraception.
> Ada and C++ offspring  include
> 1) GNAT translator
> 2) Gtkada
> 3) ......
> 
> GNAT and GTKada would not exist in their present useful form without
> C++

But, if I am correct, GNAT is in Ada. And GtkAda is of course in Ada,
sitting on top of GTK which is in C/C++.



> and many C++ programmers would not exist if their planes or trains or
> automobiles
> had not used Ada with their subsystems.
> 
> PS you could also say there would be a lot more innocent civilians around
> without
> Ada or C++ supported weapon systems.

Perhaps the world will have more peace if without Ada (or any other
programming languages) supported advanced weapon systems. :)

As in Chinese's Ying-Yang, if Ying exists, so does Yang. If good
purpose exists, so does the bad! :(

                              -- Adrian Hoe
                              -- http://greenlime.com/users/adrian.hoe

Re: Ada Operating System

[David Starner]

On 22 Feb 2002 01:08:28 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
> But, if I am correct, GNAT is in Ada. 

GNAT is in Ada and C. The bulk of the frontend is in Ada, and the
backend is in C.

> And GtkAda is of course in Ada,
> sitting on top of GTK which is in C/C++.

GTK is pure C.

-- 
David Starner - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Re: Ada Operating System

[Robert Dewar]

"tony gair" <tonygair@nounsolitations.btinternet.com> wrote in message news:<a533vq$sts$1@knossos.btinternet.com>...
> Ada and C++ offspring  include
> 1) GNAT translator
> 2) Gtkada
> 3) ......
> 
> GNAT and GTKada would not exist in their present useful 
> form without C++


I am afraid this is MUCN (More uninformed CLA nonsense :-)

Neither GNAT not GTKAda have anything whatsoever to do with C++.

Re: Ada Operating System

[Dann Corbit]

"Robert Dewar" <dewar@gnat.com> wrote in message
news:5ee5b646.0202221911.3b96252a@posting.google.com...
> "tony gair" <tonygair@nounsolitations.btinternet.com> wrote in message
news:<a533vq$sts$1@knossos.btinternet.com>...
> > Ada and C++ offspring  include
> > 1) GNAT translator
> > 2) Gtkada
> > 3) ......
> >
> > GNAT and GTKada would not exist in their present useful
> > form without C++
>
>
> I am afraid this is MUCN (More uninformed CLA nonsense :-)
>
> Neither GNAT not GTKAda have anything whatsoever to do with C++.

If (on the other hand) they are built from [for instance] C or some other
language of that nature, then the problem of insecurity is not removed, only
abstracted to be hidden.

If some library routine somewhere deep down in the bowels calls gets() you
can-be/are still hosed.
In C (and Fortran and many other languages) the sources for undefined
behavior are plentiful.
scanf("%s"....
i = i++;
The list goes on and on.

If you use a library routine or tool composed using a dangerous language,
the result is a dangerous tool.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
 "The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

Re: Ada Operating System

[Larry Kilgallen]

In article <a572f3023tj@enews2.newsguy.com>, "Dann Corbit" <dcorbit@connx.com> writes:

> If you use a library routine or tool composed using a dangerous language,
> the result is a dangerous tool.

The dangerous tool might blow up in the environment where the tool
is used -- the machine where the programmer is logged in.  To have
something blow up in the environment where the completed program
is used one does not need a dangerous tool, merely one that is not
always correct.  You can do the same thing with a programmer who
is not always correct.  Some choose to combat this with testing,
formal inspection, design review, etc.  The possible failures in
a program vastly outnumber those that might be due to using a
"dangerous" tool.

Re: Ada Operating System

[Adrian Hoe]

David Starner <dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote in message news:<a56a64$8201@news.cis.okstate.edu>...
> On 22 Feb 2002 01:08:28 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
> > But, if I am correct, GNAT is in Ada. 
> 
> GNAT is in Ada and C. The bulk of the frontend is in Ada, and the
> backend is in C.

OK. My memory was wrong.

Sometime ago, I was thrown a question while giving talks in
Universities. "What language do they use to build the Ada compiler?"
and an (assumed) answer from the audience was "C". Then raised another
question: "If Ada has many safety features that C does not have and
the Ada compiler was written in C, how exactly safe is Ada?"

I know there is ACVC test suite out there to certify Ada compilers
before they reach out the market. But then, why still use C to write
an Ada compiler?

                             -- Adrian Hoe
                             -- http://greenlime.com/users/adrian.hoe

Re: Ada Operating System

[Mark Biggar]

Adrian Hoe wrote:
 
> Sometime ago, I was thrown a question while giving talks in
> Universities. "What language do they use to build the Ada compiler?"
> and an (assumed) answer from the audience was "C". Then raised another
> question: "If Ada has many safety features that C does not have and
> the Ada compiler was written in C, how exactly safe is Ada?"
> 
> I know there is ACVC test suite out there to certify Ada compilers
> before they reach out the market. But then, why still use C to write
> an Ada compiler?

Gnat is based on the GNU C/C++ compiler suite and uses the already
existing backend optimizer/code generator virtually unchanged.  The
Gnat front end is written in Ada.

So, it's not a matter of writing part of the Ada compiler in C is more
a matter of not bothering to rewrite the already existing parts in Ada.

--
Mark Biggar
mark.a.biggar@attbi.com

Re: Ada Operating System

[Wannabe h4x0r]

On Fri, 22 Feb 2002 15:37:24 -0500, David Starner wrote:

> On 22 Feb 2002 01:08:28 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
>> But, if I am correct, GNAT is in Ada.
> 
> GNAT is in Ada and C. The bulk of the frontend is in Ada, and the
> backend is in C.

Okay, the statement that GNAT is in Ada and C is being repeated so often
that people are now actually starting to beleive it.

The correct statement would be that GNAT is written in Ada95 and the GCC
RTL. i.e. GCCs internal representation tokens. As far as I know, there is
not a single line of C code anywhere in the GNAT source code "proper".
I could be mistaken, but I dont think I am.

Chris

Re: Ada Operating System

[Jeffrey Creem]

"Wannabe h4x0r" <chris@dont.spam.me> wrote in message
news:9SMd8.122423$Pz4.542721@rwcrnsc53...
>> RTL. i.e. GCCs internal representation tokens. As far as I know, there is
> not a single line of C code anywhere in the GNAT source code "proper".
> I could be mistaken, but I dont think I am.
>
> Chris

Where is Robert when you need him..Download the gnat source code and take a
look...There are certainly some
.c files that are part of what I would consider the GNAT portion of
GCC..Mostly in the area of the interface to the
various OS runtime services... Not a lot them...But there are some there..

Watch sometime especially during the build phase where you do a make with a
target of gnatlib and you
will see a few.

Re: Ada Operating System

[Samuel Tardieu]

| As far as I know, there is not a single line of C code anywhere in the
| GNAT source code "proper".  I could be mistaken, but I dont think I am.

I'm sorry, but you are :)

A few low-level glue is written in C, mostly because it involves C
constants (which are not exported, as functions are) whose value is not
standard. Look at a-*.c files in the ada directory.

  Sam

Re: Ada Operating System

[Florian Weimer]

Wannabe h4x0r <chris@dont.spam.me> writes:

> The correct statement would be that GNAT is written in Ada95 and the GCC
> RTL. i.e. GCCs internal representation tokens.

I wouldn't phrase it this way.  GNAT trees are converted to GCC trees,
and this portion of the compiler is written in K&R C.  AFAIK, most of
the time, it doesn't deal with RTL generation directly, either.  RTL
generation from trees is the job of the GCC backend.

You can't implement GCC functionality in RTL (only by transforming
RTL, but that's a different matter ;-).

> As far as I know, there is not a single line of C code anywhere in
> the GNAT source code "proper".  I could be mistaken, but I dont
> think I am.

GNAT performs file I/O using the standard C library, for example.
Some low-level parts of the runtime library are written in C, too.

Re: Ada Operating System

[Nick Roberts]

On Tue, 19 Feb 2002 11:39:11 -0800, "Dann Corbit" <dcorbit@connx.com>
strongly typed:

>Has anyone tried to build a complete operating system in Ada?  Does anyone
>know of a paper or URL for such an effort?

The simplest, perhaps slightly glib, answer is that a _complete_ operating
system could not reasonably be built in Ada, as such:

there would always be a few low-level things that had to be done in machine
code (assembly), although of course these bits could be embedded in an Ada
program as 'machine code insertions';

there would certainly be lots of bits (a majority) which would be better
programmed (or otherwise originated) using other languages and tools,
although most or all of these bits _could_ be programmed in Ada (by the
terminally bloody minded).

The AdaOS project seeks to write in Ada the bits that make sense being
written in Ada.

This comes to quite a lot really, potentially:

most of the 'middleware' found in operating systems today (from the upper
parts and front-ends of device drivers to the more speed-critical parts of
a graphical user interface);

anything that combines complexity with a need for speed (e.g. many games,
all sorts of data translators).

I think we will eventually get around to writing most of this stuff, but in
what order, and how quickly, is mostly in the realm of clairvoyance at this
time. Plainly, some low-level stuff will have to come first.

There will be more activity at

   http://www.adaos.org

soon!
-- 
Nick Roberts

Re: Ada Operating System

[Nick Roberts]

On Wed, 20 Feb 2002 14:09:52 -0500, Hyman Rosen <hyrosen@mail.com> strongly
typed:

>Joel Sherrill wrote:
>> I have always enjoyed the irony that C/C++ supporters enhancing RTEMS 
>> are improving it as an Ada run-time. :)
>
>It's not really ironic. Remember that, while the Ada world thinks of
>C/C++ as the enemy and the competition, the C/C++ world barely realizes
>that Ada exists, and certainly wouldn't mind giving it a helping hand.

Well, I don't see C or C++ as the enemy; they are just languages.

What I see as unfortunate is the number of people who make an inappropriate
choice to use those languages for reasons of innocence, ignorance,
indolence, or intransigence.

I see as 'the enemy' those who use fear, uncertainty, and doubt as the
weapons to promote these inappropriate choices for their own cynical ends.

In this light, I would suggest that the choice made by the US Army and/or
OAR Corp to continue RTEMS in C was probably very carefully considered, and
probably not inappropriate.
-- 
Nick Roberts

Re: Ada Operating System

[Nick Roberts]

On 22 Feb 2002 01:08:28 -0800, byhoe@greenlime.com (Adrian Hoe) strongly
typed:

>Perhaps the world will have more peace if without Ada (or any other
>programming languages) supported advanced weapon systems. :)
>
>As in Chinese's Ying-Yang, if Ying exists, so does Yang. If good
>purpose exists, so does the bad! :(

I think the world might have a little more peace if the manufacturers of
advanced weapon systems didn't (or weren't permitted to) sell these weapons
to people who are likely to abuse them. Of all the world's problems, this
would solve but a few, but it _would_ solve a few.
-- 
Nick Roberts

Re: Ada Operating System

[Richard Riehle]

Adrian Hoe wrote:

> Perhaps the world will have more peace if without Ada (or any other
> programming languages) supported advanced weapon systems. :)

A study of history suggests that war is the natural state of human interaction,
not peace.   Every country that has unilaterally disarmed, or welcomed a
stronger country as a guest, has been finally conquered by that stronger
country.

An example is Tibet.  Tibet was once a militarily strong nation, with warriors
feared by every neighbor.  Tibet disarmed and sought a peaceful way of life.
China conquered and annexed Tibet.    Mexico once welcomed citizens of the
United States to live in its country as guests.   The United States , in 1846,
decided guest status was not good enough, and conquered much of Mexico,
annexing Texas, New Mexico, Arizona, California, etc.   The Nazi's ran
amuck through Europe and were only stopped by military might.   It would
be wonderful if the disagreements between countries could be solved with
peaceful means, but it is rarely possible.   Al Capone used to say about
getting results with a kind word, "In my neighborhood we get things done
with a kind word and a gun."    Reality is awful.  But it is real.

If Ada makes a contribution to the dependability of our weapon systems,
and makes us slightly more effective than our potential enemies, so be it.
If, as the science fiction authors suggest, we are someday to be overcome
by another race of intelligent creatures from beyond our own solar system,
they will be nothing more than conquerors, even if benevolent ones.  Our
technology will seem puny to them.

Richard Riehle

Re: Ada Operating System

[Richard Riehle]

Adrian Hoe wrote:

> Sometime ago, I was thrown a question while giving talks in
> Universities. "What language do they use to build the Ada compiler?"
> and an (assumed) answer from the audience was "C". Then raised another
> question: "If Ada has many safety features that C does not have and
> the Ada compiler was written in C, how exactly safe is Ada?"

This is actually quite a good question.

In discussing another programming language that uses a C-Path compiler
this question came up with regard to certain features of C that are inherently
unsafe.    It seems that, for the compiler under discussion, things such as
integer overflow, etc. were not directly addressed.   (Yes, I know GNAT
is not a C-Path compiler).

Several Ada compilers, other than GNAT, are C-Path technology.   For those
compilers to pass the conformity tests, they must account for the the issues
inherent in C.   I recall a conversation with Tucker where he described the
amount of effort that went into making certain his company's C-Path compiler
covered all the reliability issues before sending it off for validation.

Richard Riehle

Re: Ada Operating System

[David Starner]

On 22 Feb 2002 20:44:21 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
> I know there is ACVC test suite out there to certify Ada compilers
> before they reach out the market. But then, why still use C to write
> an Ada compiler?

Because GCC handles many languages, and there's a lot of work going into
GCC to improve the C, C++ and Fortran compilers that improve the Ada
compiler as a side effect. Because GCC handles pretty much every
platform known to man, as long as it's got 32 bits and runs some form of
Unix (or DOS or Windows or VMS.) While that doesn't nessecarily mean
they will run GNAT, it simplifies the problem to fixing up the threading
and exceptions (the latter I understand will go away once GNAT uses GCC
exceptions.)

-- 
David Starner - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Re: Ada Operating System

[David Starner]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 678 bytes --]

On Fri, 22 Feb 2002 19:36:09 -0800, Dann Corbit <dcorbit@connx.com> wrote:
> If some library routine somewhere deep down in the bowels calls gets() you
> can-be/are still hosed.

If some library routine calls the reboot syscall with the right magic
numbers, you are hosed. And? I can say with confidence that none of the
major libraries on this syste call gets, as the creators are totally
brain-damaged, and the linker will spit out a warning if asked to link
to gets. 

-- 
David Starner / Давид Старнэр - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Re: Ada Operating System

[Darren New]

Adrian Hoe wrote:
> Sometime ago, I was thrown a question while giving talks in
> Universities. "What language do they use to build the Ada compiler?"
> and an (assumed) answer from the audience was "C". Then raised another
> question: "If Ada has many safety features that C does not have and
> the Ada compiler was written in C, how exactly safe is Ada?"

It's possible for the Ada compiler to be unsafe yet the code it
generates is safe, yes? That is, if the compiler crashes, is that
considered a safety-critical failure? I wouldn't think so, but I don't
write safety-critical code.

In any case, regardless of what you do, the code you generate eventually
turns into machine language, and is therefore only as "safe" as the
underlying machine. Language safety comes from preventing a programmer
from stepping outside the defined semantics of the language. It doesn't
really matter what language you use to check whether a program is safe;
it only matters what language you are checking.

If you wrote a C compiler in Ada, would C suddenly become "safer"? No,
of course not. So why would an Ada compiler in C be less safe?

Richard wrote:
>  It seems that, for the compiler under discussion, things such as
> integer overflow, etc. were not directly addressed. 

This is yet another question, which is how one generates the machine
code using other tools, and how safe those tools are. I.e., this is
confusing "the code the compiler's written in" with "the code the
compiler generates", whereas the question seemed to be about "the code
the compiler's written in" vs "the code the compiler understands how to
compile". Three different systems there - Input, Compiler, and Output.
Depending on how the compiler's built, the output may be more or less
easy to get "right", i.e., accurately reflecting the semantics of the
input.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
  To the user, everything works just as expected,
    assuming the user's expectations are correct.

Re: Ada Operating System

[Wannabe h4x0r]

On Sat, 23 Feb 2002 09:31:18 -0500, Florian Weimer wrote:

> Wannabe h4x0r <chris@dont.spam.me> writes:
> 
>> The correct statement would be that GNAT is written in Ada95 and the
>> GCC RTL. i.e. GCCs internal representation tokens.
> 
> I wouldn't phrase it this way.  GNAT trees are converted to GCC trees,
> and this portion of the compiler is written in K&R C.  AFAIK, most of
> the time, it doesn't deal with RTL generation directly, either.  RTL
> generation from trees is the job of the GCC backend.
> 
> You can't implement GCC functionality in RTL (only by transforming RTL,
> but that's a different matter ;-).
> 
>> As far as I know, there is not a single line of C code anywhere in the
>> GNAT source code "proper".  I could be mistaken, but I dont think I am.
> 
> GNAT performs file I/O using the standard C library, for example. Some
> low-level parts of the runtime library are written in C, too.

Alright. I stand corrected. I did take a look at the source and I did
find some "glue" code written in C.

Nonetheless, it wouldn't be too difficult to build a stand alone compiler
 distinct from GCC using the GNAT sources and some creative use of Annex
C in the LRM I suspect. No?

Chris

Re: Ada Operating System

[Larry Kilgallen]

In article <a58lop$a2g1@news.cis.okstate.edu>, David Starner <dvdeug@x8b4e53cd.dhcp.okstate.edu> writes:
> On 22 Feb 2002 20:44:21 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
>> I know there is ACVC test suite out there to certify Ada compilers
>> before they reach out the market. But then, why still use C to write
>> an Ada compiler?
> 
> Because GCC handles many languages, and there's a lot of work going into
> GCC to improve the C, C++ and Fortran compilers that improve the Ada
> compiler as a side effect. Because GCC handles pretty much every
> platform known to man, as long as it's got 32 bits and runs some form of
> Unix (or DOS or Windows or VMS.)

"Some form of Unix (or DOS or Windows or VMS)" seems to me quite a
bit shy of "every platform known to man".  It omits HP/MPE, MVS
and OS/400.  It also omits MacOS (pre-X) for which only the PowerPC
platform can be targeted by GNAT and neither platform can host GNAT.
(While using the Unix emulator to host a compiler is a neat trick,
it is not an example of GCC/GNAT being adapted to run on MacOS.)

Re: Ada Operating System

[David Starner]

On 23 Feb 2002 14:30:57 -0600, Larry Kilgallen <Kilgallen@SpamCop.net> wrote:
>> Because GCC handles pretty much every
>> platform known to man, as long as it's got 32 bits and runs some form of
>> Unix (or DOS or Windows or VMS.)
> 
> "Some form of Unix (or DOS or Windows or VMS)" seems to me quite a
> bit shy of "every platform known to man".  It omits HP/MPE, MVS
> and OS/400.  

That's why "runs some form of ..." modifies "every platform known to
man". (And GCC does target MVS.) True, GCC doesn't handle archaic
systems and misses some Big Iron. In any case, since we're talking
compilers here, do you know of another compiler that targets Unix, DOS,
Windows and VMS non-trivially?

-- 
David Starner - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Re: Ada Operating System

[Larry Kilgallen]

In article <a59jtr$7pu1@news.cis.okstate.edu>, David Starner <dvdeug@x8b4e53cd.dhcp.okstate.edu> writes:
> On 23 Feb 2002 14:30:57 -0600, Larry Kilgallen <Kilgallen@SpamCop.net> wrote:
>>> Because GCC handles pretty much every
>>> platform known to man, as long as it's got 32 bits and runs some form of
>>> Unix (or DOS or Windows or VMS.)
>> 
>> "Some form of Unix (or DOS or Windows or VMS)" seems to me quite a
>> bit shy of "every platform known to man".  It omits HP/MPE, MVS
>> and OS/400.  
> 
> That's why "runs some form of ..." modifies "every platform known to
> man". (And GCC does target MVS.) True, GCC doesn't handle archaic
> systems and misses some Big Iron. In any case, since we're talking
> compilers here, do you know of another compiler that targets Unix, DOS,
> Windows and VMS non-trivially?

I agree that GCC also makes great toast and coffee for breakfast,
so long as you like both cold and extremely weak.

Please note the modifier clause :-)

Re: Ada Operating System

[David Starner]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1503 bytes --]

On 23 Feb 2002 22:27:42 -0600, Larry Kilgallen <Kilgallen@SpamCop.net> wrote:
>>> "Some form of Unix (or DOS or Windows or VMS)" seems to me quite a
>>> bit shy of "every platform known to man".  It omits HP/MPE, MVS
>>> and OS/400.  
>> 
>> That's why "runs some form of ..." modifies "every platform known to
>> man". (And GCC does target MVS.) True, GCC doesn't handle archaic
>> systems and misses some Big Iron. In any case, since we're talking
>> compilers here, do you know of another compiler that targets Unix, DOS,
>> Windows and VMS non-trivially?
> 
> I agree that GCC also makes great toast and coffee for breakfast,
> so long as you like both cold and extremely weak.

In other words, because GCC doesn't support two platforms, the fact that
it supports a hundred others garners no awards.

As for reality, for all the support, nobody who has a MVS system
bothered to build it and follow the build instructions to send an email
to gcc@gcc.gnu.org. In fact the list of systems where people have done
so is limited to, hmm, Linux, Cygwin, the major proprietary Unixes, and
. . . a few of the minor Unixes. There's a major Mac group hacking on
the compiler, and many embedded/real-time hackers, but the number of
people who mail the list over something else can be counted on one paw.

-- 
David Starner / Давид Старнэр - starner@okstate.edu
What we've got is a blue-light special on truth. It's the hottest thing 
with the youth. -- Information Society, "Peace and Love, Inc."

Re: Ada Operating System

[Adrian Hoe]

Wannabe h4x0r <chris@dont.spam.me> wrote in message news:<9SMd8.122423$Pz4.542721@rwcrnsc53>...
> On Fri, 22 Feb 2002 15:37:24 -0500, David Starner wrote:
> 
> > On 22 Feb 2002 01:08:28 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
> >> But, if I am correct, GNAT is in Ada.
> > 
> > GNAT is in Ada and C. The bulk of the frontend is in Ada, and the
> > backend is in C.
> 
> Okay, the statement that GNAT is in Ada and C is being repeated so often
> that people are now actually starting to beleive it.
> 
> The correct statement would be that GNAT is written in Ada95 and the GCC
> RTL. i.e. GCCs internal representation tokens. As far as I know, there is
> not a single line of C code anywhere in the GNAT source code "proper".
> I could be mistaken, but I dont think I am.

This sounds familiar, Chris. I remembered someone talked about this sometime ago.

                              -- Adrian Hoe
                              -- http://greenlime.com/users/adrian.hoe

> Chris

Re: Ada Operating System

[Ian S. Nelson]

Wannabe h4x0r wrote:
> On Fri, 22 Feb 2002 15:37:24 -0500, David Starner wrote:
> 
> 
>>On 22 Feb 2002 01:08:28 -0800, Adrian Hoe <byhoe@greenlime.com> wrote:
>>
>>>But, if I am correct, GNAT is in Ada.
>>>
>>GNAT is in Ada and C. The bulk of the frontend is in Ada, and the
>>backend is in C.
>>
> 
> Okay, the statement that GNAT is in Ada and C is being repeated so often
> that people are now actually starting to beleive it.
> 
> The correct statement would be that GNAT is written in Ada95 and the GCC
> RTL. i.e. GCCs internal representation tokens. As far as I know, there is
> not a single line of C code anywhere in the GNAT source code "proper".
> I could be mistaken, but I dont think I am.
> 
> Chris

That is a technically correct statement, if that makes some people feel 
better.  Other than some "glue" GNAT looks like it's pure Ada. 
Spiritually, those who say it's in Ada and C are accurate also.  If I 
give you only source code and you don't have a C compiler, you're not 
going to end up with an Ada compiler with the GNAT source tree.

Ian

Re: Ada Operating System

[Marin David Condic]

1) Which Ada compiler? There is no "The Ada Compiler". Each one is going to
be implemented differently. Given that, some compilers may be written in C,
others in Ada, some possibly in obscure languages like BLISS - did DEC do
that with their compiler? The first validated compiler was written in SETL.

2) Note that Gnat is developed in Ada. Sure, it has the gcc back end and all
that, but that's just leveraging some existing body of code. Ada can be, and
has been, used to implement Ada compilers.

3) Assuming for a moment that a given Ada compiler is developed in C - and
that the implementation (being in C) is particularly buggy, how does that
have anything to do with the quality of the output? Even a buggy program
might make highly reliable output - if and when it works.

4) The ACVC doesn't really have much to do with the "quality" of an Ada
compiler. What it does is test it for conformance to the rules - which may
act as a useful test suite & help you produce a better quality product.
(Conformance to the rules being one aspect of "quality", I suppose we could
consider it an aid to getting there.) A compiler can pass the ACVC and still
produce inefficient, buggy code, so its no guarantee.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Adrian Hoe" <byhoe@greenlime.com> wrote in message
news:9ff447f2.0202222044.78c44b25@posting.google.com...
>
> Sometime ago, I was thrown a question while giving talks in
> Universities. "What language do they use to build the Ada compiler?"
> and an (assumed) answer from the audience was "C". Then raised another
> question: "If Ada has many safety features that C does not have and
> the Ada compiler was written in C, how exactly safe is Ada?"
>
> I know there is ACVC test suite out there to certify Ada compilers
> before they reach out the market. But then, why still use C to write
> an Ada compiler?
>

Re: Ada Operating System

[Marin David Condic]

Its probably important to once again note that just because something is
written in C doesn't automatically mean that it *must* be buggy and
unreliable. The correct contention for Ada to make is that it is just harder
and more expensive to get there with C. Much like it is possible to produce
a correctly spelled document with a word processor that lacks a
spell-checker. You can get there quicker and with less cost by using one
that does - but without it, its still possible to correct all the spelling
errors.

Certainly time is a major ally in this. Something like gcc has been around
for a fairly long time (in computer years) and has had a large number of
eyes pouring over it. That tends to give one increased confidence in it even
if it *is* written in C.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:8nH9bwDSZjm1@eisner.encompasserve.org...
>
> The dangerous tool might blow up in the environment where the tool
> is used -- the machine where the programmer is logged in.  To have
> something blow up in the environment where the completed program
> is used one does not need a dangerous tool, merely one that is not
> always correct.  You can do the same thing with a programmer who
> is not always correct.  Some choose to combat this with testing,
> formal inspection, design review, etc.  The possible failures in
> a program vastly outnumber those that might be due to using a
> "dangerous" tool.

Re: Ada Operating System

[Marin David Condic]

"Nick Roberts" <nickroberts@ukf.net> wrote in message
news:3c77b476.322111671@news.cis.dfn.de...
>
> I think we will eventually get around to writing most of this stuff, but
in
> what order, and how quickly, is mostly in the realm of clairvoyance at
this
> time. Plainly, some low-level stuff will have to come first.
>
I'd suggest that if you concentrated on getting a bootstrap going with some
minimal level of device drivers & scheduling such that small, primitive apps
might be buildable, you'd really have something there that would start
spawning interest. Do it with the Gnat compiler - at least initially - and
it will make it easy for people to participate. Aim for a small, achievable
goal & the rest will start falling in place.

> There will be more activity at
>
>    http://www.adaos.org
>
> soon!

That would be good to see. I think an OS in Ada is a noble goal that would
likely spawn more interest in the language. You might consider aiming at
having some realtime capabilities because this market could really use a
good, reliable OS and it wouldn't require having a bazillion features and
applications to make it useful. (A PC OS would be nice as a growth goal, but
if you got as far as a scheduler that is capable of realtime & some I/O
support, that would be enough for a lot of realtime & embedded projects.)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Pascal Obry]

byhoe@greenlime.com (Adrian Hoe) writes:

> OK. My memory was wrong.
> 
> Sometime ago, I was thrown a question while giving talks in
> Universities. "What language do they use to build the Ada compiler?"
> and an (assumed) answer from the audience was "C". Then raised another
> question: "If Ada has many safety features that C does not have and
> the Ada compiler was written in C, how exactly safe is Ada?"

And everything is translated to machine code !!! Serriouly, the above comment
is just about the Ada compiler itself... but for Ada users there is BIG
difference, the Ada compiler to check for Ada SYNTAX and SEMANTICS!

Pascal.

-- 

--|------------------------------------------------------
--| Pascal Obry                           Team-Ada Member
--| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE
--|------------------------------------------------------
--|         http://perso.wanadoo.fr/pascal.obry
--|
--| "The best way to travel is by means of imagination"

Re: Ada Operating System

[Pascal Obry]

"Jeffrey Creem" <jeff@thecreems.com> writes:

> Where is Robert when you need him..Download the gnat source code and take a
> look...There are certainly some
> .c files that are part of what I would consider the GNAT portion of
> GCC..Mostly in the area of the interface to the
> various OS runtime services... Not a lot them...But there are some there..

Indeed most of them just to get the value of C macros and handle easily the
OS specific parts.

Pascal.

-- 

--|------------------------------------------------------
--| Pascal Obry                           Team-Ada Member
--| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE
--|------------------------------------------------------
--|         http://perso.wanadoo.fr/pascal.obry
--|
--| "The best way to travel is by means of imagination"

Re: Ada Operating System

[Randy Brukardt]

>Sometime ago, I was thrown a question while giving talks in
>Universities. "What language do they use to build the Ada compiler?"
>and an (assumed) answer from the audience was "C". Then raised another
>question: "If Ada has many safety features that C does not have and
>the Ada compiler was written in C, how exactly safe is Ada?"
>
>I know there is ACVC test suite out there to certify Ada compilers
>before they reach out the market. But then, why still use C to write
>an Ada compiler?

For what its worth, Janus/Ada is 99.5% written in Ada, and always has
been (since June 1981, long before it was first released). There is a
small amount of assembler in the compiler (for performance reasons; we
have Ada equivalent code that we use when porting the compiler to
another platform). The runtime is also all in Ada except for about 8000
lines of assembler that do things not possible in Ada 83, including task
context switching, implementing the heap, exception handling, and so on.
Much of that could be written in Ada 95; we haven't done so mainly
because it isn't broken (important rule of computing: "If it ain't
broke, don't fix it!"). We've also built all of the compiler's tools
(Make, Debugger, etc.) in Ada. The only exception is the programming
environment, which we acquired from another company; we're planning to
replace it with a Claw application as we can't fix it (no source code),
which will eliminate that as non-Ada code.

I believe at least one other Ada compiler is virtually all in Ada.

                    Randy.

Re: Ada Operating System

[Wes Groleau]

> GNAT performs file I/O using the standard C library, for example.
> Some low-level parts of the runtime library are written in C, too.

Including much if not all of Ada.Numerics.*

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Ted Dennison]

"Randy Brukardt" <randy@rrsoftware.com> wrote in message news:<u7l5vg5kn26uad@corp.supernews.com>...
> For what its worth, Janus/Ada is 99.5% written in Ada, and always has
...
> broke, don't fix it!"). We've also built all of the compiler's tools
> (Make, Debugger, etc.) in Ada. The only exception is the programming
> environment, which we acquired from another company; we're planning to
> replace it with a Claw application as we can't fix it (no source code),
> which will eliminate that as non-Ada code.

Since you don't have the sources, technically you have no evidence
(other than perhaps someone's word) that it isn't Ada code. ;-)

-- 
T.E.D. 
Home     -  mailto:dennison@telepath.com (Yahoo: Ted_Dennison)
Homepage -  http://www.telepath.com/dennison/Ted/TED.html

Re: Ada Operating System

[Larry Kilgallen]

In article <a5dfhe$k82$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
> 1) Which Ada compiler? There is no "The Ada Compiler". Each one is going to
> be implemented differently. Given that, some compilers may be written in C,
> others in Ada, some possibly in obscure languages like BLISS - did DEC do
> that with their compiler?

Yes, although Bliss was far from obscure with their workforce.

Re: Ada Operating System

[Larry Kilgallen]

In article <a5dgc4$khi$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
> Its probably important to once again note that just because something is
> written in C doesn't automatically mean that it *must* be buggy and
> unreliable. The correct contention for Ada to make is that it is just harder
> and more expensive to get there with C. Much like it is possible to produce
> a correctly spelled document with a word processor that lacks a
> spell-checker. You can get there quicker and with less cost by using one
> that does - but without it, its still possible to correct all the spelling
> errors.

You must have a better spell-checker than the one I got from Bill.

Re: Ada Operating System

[James Ross]

>> There will be more activity at
>>
>>    http://www.adaos.org
>>
>> soon!
>
>That would be good to see. I think an OS in Ada is a noble goal that would
>likely spawn more interest in the language. You might consider aiming at
>having some realtime capabilities because this market could really use a
>good, reliable OS and it wouldn't require having a bazillion features and
>applications to make it useful. (A PC OS would be nice as a growth goal, but
>if you got as far as a scheduler that is capable of realtime & some I/O
>support, that would be enough for a lot of realtime & embedded projects.)
>
>MDC

I would like to ditto this statement.

Because I have had a renewed interest in Ada I thought I would follow
the AdaOS project and possibly even help out if I could.   I thought
that an OS written primarily in Ada would be a great platform to
learn, experiment, or "play" with Ada.   If this project ever gets off
the ground and proves to be useful I think it would do wonders for Ada
advocacy.  

I get the impression from comments that there is progress being made
behind the scenes.   However, I assume nothing (i.e. source code) is
going to be released until something is in a complete "working" state.
If this is the case,  I kind feel they have jumped the gun with
promoting the project as open source GPL and then keeping the progress
"closed".

My personal experience is I often get bogged down on the design of a
new from scratch project. I am not very good at sitting down and
planning it all out to last detail, especially a large project. When
this happens to me there are two solutions.  Take a bottom-up approach
and just start working on pieces that you know are going to be needed
in some form, or start plagiarizing someone else's code.  (At least
the latter is legal with GPLed code!)  Perhaps this is what is needed
in the case of AdaOS?

James Ross

Re: Ada Operating System

[Jorge Real]

[-- Attachment #1: Type: text/plain, Size: 413 bytes --]

You can take a look at 
  http://marte.unican.es 
for a real-time kernel on bare x86.

Jorge.

Dann Corbit ha escrito:
> 
> Has anyone tried to build a complete operating system in Ada?  Does anyone
> know of a paper or URL for such an effort?
> --
> C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
>  "The C-FAQ Book" ISBN 0-201-84519-9
> C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Tarjeta para Jorge Real --]
[-- Type: text/x-vcard; charset=us-ascii; name="jorge.vcf", Size: 320 bytes --]

begin:vcard 
n:Real S�ez;Jorge
tel;fax:+34 96 387 7579
tel;work:+34 96 387 9702
x-mozilla-html:FALSE
org:Universidad Polit�cnica de Valencia;Departamento de Inform�tica de Sistemas y Computadores
adr:;;Camino de Vera 14;Valencia;;E-46020;Spain
version:2.1
email;internet:jorge@disca.upv.es
fn:Jorge Real
end:vcard

Re: Ada Operating System

[Ken Pinard]

About 12 years ago there was a computer OS called Biin. It was a joint
effort by Intel and a German company. It worked great, distributing Ada
Tasks across multiple processors in the box. Unfortunitly it was ill timed
and did not prosper. The OS it self ran well.

Ken Pinard

"Dann Corbit" <dcorbit@connx.com> wrote in message
news:a4u9d00vqe@enews2.newsguy.com...
> Has anyone tried to build a complete operating system in Ada?  Does anyone
> know of a paper or URL for such an effort?
> --
> C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
>  "The C-FAQ Book" ISBN 0-201-84519-9
> C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm
>
>

Re: Ada Operating System

[Dave Poirier]

Ken Pinard wrote:
> About 12 years ago there was a computer OS called Biin. It was a joint
> effort by Intel and a German company. It worked great, distributing Ada
> Tasks across multiple processors in the box. Unfortunitly it was ill timed
> and did not prosper. The OS it self ran well.
> 
> Ken Pinard
> 
> "Dann Corbit" <dcorbit@connx.com> wrote in message
> news:a4u9d00vqe@enews2.newsguy.com...
> 
>>Has anyone tried to build a complete operating system in Ada?  Does anyone
>>know of a paper or URL for such an effort?
>>--
>>C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
>> "The C-FAQ Book" ISBN 0-201-84519-9
>>C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

The Biin/OS is indeed very interesting (google is your friend), I wonder 
if the sources could be had or made open source, this could be very 
interesting for the Ada community.

EKS - Dave Poirier

Re: Ada Operating System

[tmoran]

Is there a description somewhere of Windows' NTFS disk layout?  It would
be wonderful to be able to write an Ada program that would be able to
read/write an NTFS partition without hindrance from Windows.

Re: Ada Operating System

[chris.danx]

<tmoran@acm.org> wrote in message
news:Bjjf8.48764$967.2706237501@newssvr14.news.prodigy.com...
> Is there a description somewhere of Windows' NTFS disk layout?  It would
> be wonderful to be able to write an Ada program that would be able to
> read/write an NTFS partition without hindrance from Windows.

There is some information available at
http://sourceforge.net/projects/linux-ntfs/, under ntfs documentation.  The
project members are gathering information from as many sources as they can
for a linux driver, it doesn't look complete but they're making progress.

Chris

Re: Ada Operating System

[Georg Bauhaus]

tmoran@acm.org wrote:
: Is there a description somewhere of Windows' NTFS disk layout?  It would
: be wonderful to be able to write an Ada program that would be able to
: read/write an NTFS partition without hindrance from Windows.

there is an o'Reilly book `NTFS internals' (or some such).
don't have it here, so details off-group if you wish.

-- georg
---
Microsoft Windows--a fresh perspective on information hiding

Re: Ada Operating System

[Wes Groleau]

> About 12 years ago there was a computer OS called Biin. It was a joint
> effort by Intel and a German company. It worked great, distributing Ada
> Tasks across multiple processors in the box. Unfortunitly it was ill timed
> and did not prosper. The OS it self ran well.

And Rational had Delta, which if not written in Ada,
had an Ada-like user interface.  But when they dropped
their hardware line, this software went with it.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

> Is there a description somewhere of Windows' NTFS disk layout?  It would
> be wonderful to be able to write an Ada program that would be able to
> read/write an NTFS partition without hindrance from Windows.

One could use that to generate a lot of hindrance  TO  Windows!  :-)

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

> http://sourceforge.net/projects/linux-ntfs/, under ntfs documentation.  The
> project members are gathering information from as many sources as they can
> for a linux driver, it doesn't look complete but they're making progress.

Don't say that too loud.  If Bill hears, they'll change
the format!

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[chris.danx]

"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C7E3C70.665A2687@despammed.com...
>
> > Is there a description somewhere of Windows' NTFS disk layout?  It would
> > be wonderful to be able to write an Ada program that would be able to
> > read/write an NTFS partition without hindrance from Windows.
>
> One could use that to generate a lot of hindrance  TO  Windows!  :-)

Joking aside... you could violate NTFS integrity and its security model, so
it's probably not a good idea especially if the program was boot from a
floppy.  You could delete system files, change some attributes, etc.  Not
good.

OTOH it'd be nice to implement for another OS, like the Linux NTFS guys are
doing.  I'm seriously considering making NTFS the native FS of the OS I'm
developing (the microkernel is being designed now, it's called Nu.  The
complete system as yet has no name), if sufficient information exists.
Since win2k was installed I sometimes wonder how I ever did without NTFS
(performance has been about the same as FAT32 on this machine if not
slightly better).

Chris

Re: Ada Operating System

[Ian Wild]

"chris.danx" wrote:

> Joking aside... you could violate NTFS integrity and its security model, so
> it's probably not a good idea especially if the program was boot from a
> floppy.

What sort of "security model" is predicated on the
bad guys not having a bootable floppy?

Re: Ada Operating System

[chris.danx]

"Ian Wild" <ian@cfmu.eurocontrol.be> wrote in message
news:3C7E4DE6.F184C408@cfmu.eurocontrol.be...
> "chris.danx" wrote:
>
> > Joking aside... you could violate NTFS integrity and its security model,
so
> > it's probably not a good idea especially if the program was boot from a
> > floppy.
>
> What sort of "security model" is predicated on the
> bad guys not having a bootable floppy?

I don't understand exactly what your asking, but if you mean what security
model does NTFS have when someone boots from a floppy, then none -- it's
just a bunch of ones and zeros on a disk.  NTFS security model has to be
enforced by the operating system in some manner, so an OS has to be present
to ensure the 'correctness' of the FS and enforce security.  The point was
that NTFS and windows have a security model which would be circumvented by
such a program, hence it's probably not a good idea to write such a program.
The integrity of the NTFS volume could be compromised and the users' data no
longer secure (as secure as the OS allows).


Chris.0
00000

Re: Ada Operating System

[Dave Poirier]

tmoran@acm.org wrote:
> Is there a description somewhere of Windows' NTFS disk layout?  It would
> be wonderful to be able to write an Ada program that would be able to
> read/write an NTFS partition without hindrance from Windows.
> 

The only somewhat useful information I have found about the NTFS disk
layout can be found here:

    http://www.informatik.hu-berlin.de/~loewis/ntfs/

Good Luck
EKS - Dave Poirier

Re: Ada Operating System

[Dave Poirier]

chris.danx wrote:
> "Ian Wild" <ian@cfmu.eurocontrol.be> wrote in message
> news:3C7E4DE6.F184C408@cfmu.eurocontrol.be...
> 
>>"chris.danx" wrote:
>>
>>
>>>Joking aside... you could violate NTFS integrity and its security model,
>>>
> so
> 
>>>it's probably not a good idea especially if the program was boot from a
>>>floppy.
>>>
>>What sort of "security model" is predicated on the
>>bad guys not having a bootable floppy?
>>
> 
> I don't understand exactly what your asking, but if you mean what security
> model does NTFS have when someone boots from a floppy, then none -- it's
> just a bunch of ones and zeros on a disk.  NTFS security model has to be
> enforced by the operating system in some manner, so an OS has to be present
> to ensure the 'correctness' of the FS and enforce security.  The point was
> that NTFS and windows have a security model which would be circumvented by
> such a program, hence it's probably not a good idea to write such a program.
> The integrity of the NTFS volume could be compromised and the users' data no
> longer secure (as secure as the OS allows).
> 
> 
> Chris.0
> 00000

This seems to be the only technique Microsoft knows, Security by 
obstruction.  They seriously believe that by hiding the "how" you do it, 
you will make your file system secure.  If they would even give it some 
serious thought they would know the only way to be secure on the long 
run is to use encryption.  Using a proper algorithm they could give away 
the algorithm as well as the exact details of how the files are stored 
on disk, after all it would just make it easier to create recovery tools.

Re: Ada Operating System

[Dave Poirier]

> This seems to be the only technique Microsoft knows, Security by 
> obstruction.  They seriously believe that by hiding the "how" you do it, 
(snip)

sorry, should have said "Security by obscurity", my french coming back 
at me ;)

Re: Ada Operating System

[Tom Moran]

> could violate NTFS integrity and its security
> model, so it's probably not a good idea
  If the night watchman has a habit of getting drunk on the job, it's a
security hole, but a good idea, to have a way to get past him.

Re: Ada Operating System

[Chad R. Meiners]

NTFS supports encrypting sensitive files.  So of course MS does not believe
that a propriety disk format is inherently secure!


"Dave Poirier" <instinc@users.sf.net> wrote in message
news:3C7E6E48.2000402@users.sf.net...
> chris.danx wrote:
> > "Ian Wild" <ian@cfmu.eurocontrol.be> wrote in message
> > news:3C7E4DE6.F184C408@cfmu.eurocontrol.be...
> >
> >>"chris.danx" wrote:
> >>
> >>
> >>>Joking aside... you could violate NTFS integrity and its security
model,
> >>>
> > so
> >
> >>>it's probably not a good idea especially if the program was boot from a
> >>>floppy.
> >>>
> >>What sort of "security model" is predicated on the
> >>bad guys not having a bootable floppy?
> >>
> >
> > I don't understand exactly what your asking, but if you mean what
security
> > model does NTFS have when someone boots from a floppy, then none -- it's
> > just a bunch of ones and zeros on a disk.  NTFS security model has to be
> > enforced by the operating system in some manner, so an OS has to be
present
> > to ensure the 'correctness' of the FS and enforce security.  The point
was
> > that NTFS and windows have a security model which would be circumvented
by
> > such a program, hence it's probably not a good idea to write such a
program.
> > The integrity of the NTFS volume could be compromised and the users'
data no
> > longer secure (as secure as the OS allows).
> >
> >
> > Chris.0
> > 00000
>
> This seems to be the only technique Microsoft knows, Security by
> obstruction.  They seriously believe that by hiding the "how" you do it,
> you will make your file system secure.  If they would even give it some
> serious thought they would know the only way to be secure on the long
> run is to use encryption.  Using a proper algorithm they could give away
> the algorithm as well as the exact details of how the files are stored
> on disk, after all it would just make it easier to create recovery tools.
>
>

Re: Ada Operating System

[Marin David Condic]

Fair enough. I knew about BLISS because back when I was in college just
about everything was DEC equipment where I went to school. But it didn't
have much of a following outside of DEC - if any following. Recollection is
that it looked a little like Pascal but with less type checking - more "raw
bits" kind of flavor.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:8nJIRrQtxpn7@eisner.encompasserve.org...
>
> Yes, although Bliss was far from obscure with their workforce.

Re: Ada Operating System

[Marin David Condic]

Its easy for a project to get so overly-ambitious that it fails to get
anywhere constructive. I think the AdaOS project is a really good idea, but
I think it needs to have some sort of phased goals/objectives that at least
bust the problem up into something achievable. You're right about it being
easy to get bogged down in design - sometimes its better to say "We'll burn
that bridge when we get to it!" :-)

If it were my gig, I'd look to getting some kind of bootable kernel going by
utilizing the Gnat compiler (just because its there and readily accessible
to the hobbyist). If it could boot off a floppy and do a little basic
scheduling & I/O so that folks could start playing with it, then it would
likely spark interest.

Of course, my bias would be to provide good realtime support for the
scheduler because I think that's an unaddressed (or insufficiently
addressed) niche by both Windows & Linux. If it could run on a PC and
provide the possibility of running at least one realtime process, it would
have something you don't get elsewhere. That's product distinction! :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"James Ross" <rem.jr@rem.webross.com> wrote in message
news:m4ro7uorulr8cfn9tn3hk30cqi7mf4egdc@4ax.com...
>
> My personal experience is I often get bogged down on the design of a
> new from scratch project. I am not very good at sitting down and
> planning it all out to last detail, especially a large project. When
> this happens to me there are two solutions.  Take a bottom-up approach
> and just start working on pieces that you know are going to be needed
> in some form, or start plagiarizing someone else's code.  (At least
> the latter is legal with GPLed code!)  Perhaps this is what is needed
> in the case of AdaOS?
>

Re: Ada Operating System

[James Ross]

On Tue, 5 Mar 2002 09:28:30 -0500, "Marin David Condic"
<dont.bother.mcondic.auntie.spam@[acm.org> wrote:

>If it were my gig, I'd look to getting some kind of bootable kernel going by
>utilizing the Gnat compiler (just because its there and readily accessible
>to the hobbyist). If it could boot off a floppy and do a little basic
>scheduling & I/O so that folks could start playing with it, then it would
>likely spark interest.

Even this alone is a fairly tall order to build from scratch, let
alone everything else that is envisioned for AdaOS on their site.
Using GNAT is definitely the way to go but it does introduce a
somewhat "chicken and the egg" problem with the build / link / execute
process;  I believe there are issues with the GNAT runtime as well as
getting a executable image as these are platform dependent.  (I might
be wrong about this!) 

>Of course, my bias would be to provide good realtime support for the
>scheduler because I think that's an unaddressed (or insufficiently
>addressed) niche by both Windows & Linux. If it could run on a PC and
>provide the possibility of running at least one realtime process, it would
>have something you don't get elsewhere. That's product distinction! :-)
>
>MDC

Yes, the real lure is to end up with an OS on par with Linux / Windows
/ Mac but somehow better.  And to have hard real-time functionality
built is just one thing on the path to being better.  

I would suggest (or if it were my gig, as you say!) to take the
following pieces of GPL'd code: The Linux kernel, one of the Linux
real-time extensions, the GNU tools and GNAT and then design the AdaOS
API's (kernel, etc..) by creating a, I believe it is called a "thick
binding" to the underlying Linux OS. Then it would be possible to
begin writing tools and such for the new OS.  Then slowly replace the
Linux kernel with a new and improved (?) Ada kernel.

James Ross

Re: Ada Operating System

[Marin David Condic]

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:o2td8ug1chantij2e6j80kk56mrafp6e78@4ax.com...
>
> Even this alone is a fairly tall order to build from scratch, let
> alone everything else that is envisioned for AdaOS on their site.
> Using GNAT is definitely the way to go but it does introduce a
> somewhat "chicken and the egg" problem with the build / link / execute
> process;  I believe there are issues with the GNAT runtime as well as
> getting a executable image as these are platform dependent.  (I might
> be wrong about this!)
>

Sure. That's a lot to get going. But the thing is, (as you say, looking at
the AdaOS website) if you get too ambitious and bite off more than you can
chew, it will just remain A Wonderful Idea & never get too far. I'd think
that what I suggested as a starting point might be an achievable goal by a
small handful of hobyists & all of the rest of the noble Devoutly To Be
Desired Results could be solidified & added at some later point. (The world
changes along the way & entirely new directions might get going, right?) A
bootstrap & scheduler that could execute some toy programs would provide
something hobyists could start to play with & provide a starting point for
something that might actually be a product that could be used commercially.
It gets the ball rolling.

As far as GNAT goes, I seem to recall that there was a "Gnat No Runtime" at
one point & maybe the compiler can be used to produce code that doesn't rely
on a runtime if you stick to some subset of the language. I just don't
know - it would have to be investigated. (Eventually, you implement the
runtime you need - some version of Posix calls IIRC, if you'd want to just
use Gnat as is.) I never tried to get pure (by which I mean "no calls to any
runtime") code out of GNAT, but I'll bet it can be done.

Another possibility would be to utilize something like RTEMS to provide the
RTK & then implement the rest of the OS on top of that - using what you need
and discarding the rest.

Of course, to get a bootstrap & kernel going, you'd need a really capable
linker. You have to be able to control where things go in memory, etc. if
you want to get right down to the bare silicon. I don't know how capable the
standard PC linker is (never had cause to investigate) so you might need to
go to some other tool. I suspect that a solution does exist - it would just
need to be identified.


>
> Yes, the real lure is to end up with an OS on par with Linux / Windows
> / Mac but somehow better.  And to have hard real-time functionality
> built is just one thing on the path to being better.
>
Its commonly observed when this topic comes up that "realtime" and
"workstation" often can't live in the same universe at the same time. Yet
there are folks making money by providing realtime extensions to popular
operating systems. What if the TBD-OS in Ada were to have just one single
process that could be designated "realtime" and it runs at the priority of
the OS (or higher)? You say "Normal usage is pretty much like Windows or
Linux processes - round-robin or whatever scheduling and no guaranteed
latencies. But *if* you create The Magical Process (and have the appropriate
permissions) then The Magical Process gets to take over & just relies on the
OS for certain services that can be built to be deterministic. It gives up
spare time to the rest of the world but as soon as its ready to run, it
runs. The rest is up to you..."

I think if you had that capability, then TBD-OS would offer something to the
market that isn't really there already - except by patches & kludges. It
would offer Product Distinction & would satisfy a real-world need.


> I would suggest (or if it were my gig, as you say!) to take the
> following pieces of GPL'd code: The Linux kernel, one of the Linux
> real-time extensions, the GNU tools and GNAT and then design the AdaOS
> API's (kernel, etc..) by creating a, I believe it is called a "thick
> binding" to the underlying Linux OS. Then it would be possible to
> begin writing tools and such for the new OS.  Then slowly replace the
> Linux kernel with a new and improved (?) Ada kernel.
>
I wouldn't go with the Linux kernel for a handful of reasons. One being that
the objective ought to be to have a truly Ada OS - one that isn't "just like
everyone else". Why be hamstrung with the past? Starting with Linux (even if
it was just a translation of the kernel into Ada) is saying "Me Too!!!" when
it ought to be saying "I'm Better!!!"

Going with RTEMS might be a way to start. Building it from
bottom-dead-center might be possible. It might be worth identifying
someone's OS textbook as a model on which to build.  Whatever would be
underlying it I don't think it would be an impossible goal to shoot for
cobbling together whatever compiler/linker stuff was needed & building a
bootstrap & scheduler. Set some minimal objectives and get those achieved.

You want to team up & go tilting at windmills? :-) I may be able to find
some time for a side project. (Although I've already got a half-dozen or so
going already & none of them making fast enough progress! :-)

MDC

--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Wes Groleau]

A pessimist's viewpoint:

To make an operating system successful, you need
at least one of

 - sufficient quantity of useful applications

 - sufficient geek appeal to attract developers.

 - sufficent robustness/power/features to overcome
   the lack of either of the others.

To get people to write applications that run on it,
it must have enough users to make people want to
write apps for it.  So you have a chicken/egg problem
for the first point.

For the second, you have the prevailing distaste for
Ada against you ("Us" if you will).

On the robustness/power/features point, obviously
Ada has a significant advantage over C/C++/Java.

BUT, when you multiply the language potential
by the number of developers, Linux still comes
out ahead.  In spite of its implementation language,
Linux is quite robust, and it has hundreds of
developers for every AdaOS developer to keep it
that way.  Shift the target to features, and
Microsoft will always have you beat there.
They will always sacrifice security/robustness
to beat any competition in features.  And their
"if you can't beat 'em, steal 'em" technique
also applies.

So on the third point, you have say a hundred
developers and a language that scores ten on
some arbitrary rating scale vs. several thousand
developers and a language that scores one or two.

Do the math and Ada loses.  Unfortunately.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Marin David Condic]

Well, by that reasoning, Linux should have never been able to get going.
From bottom dead center, it had no useful apps or geek appeal or robustness.
It had to get started from *somewhere* to have anything to interest geeks in
developing things for it and it took time before it became something stable
and reliable, right?

Maybe I'm not so pessimistic about the possibilities. One thing is
absolutely certain - if one starts with an attitude of "It'll never work,
its doomed to fail, there's no hope..." then guaranteed it won't happen.

One of the reasons I suggest doing something that would (a) be a small
kernel starting point and (b) provide some realtime capability is because I
think that it alleviates the need for massive amounts of useful apps and
provides geek appeal. (The robustness we get just because, as all
right-thinking people know, its not possible to write bad code in Ada,
right? :-)

Imagine that you had a floppy that would let you boot up a garden variety PC
and be able to load a realtime app you constructed that could access
standard issue devices (serial port, ethernet card, etc.) There wouldn't be
geeks that would want to build toy apps and fool around with them? There
wouldn't be someone saying "Hey! I can still get some use out of this old
'386 hardware because I don't have to run Windoze needing a bajillion bytes
of memory, etc..."? Maybe even some commercial uses for PC compatible
embedded computers or PC based realtime apps? If it had some use in the
realtime/embedded world, it doesn't need an Office-like suite and it pretty
much has some instant geek-appeal as long as it provides some capabilities
you don't get with most other OS's.

I just think there is some potential for success - but it certainly needs
some minimal level of working stuff before it could generate much interest.
Will it happen? I don't know. It depends on getting a few die-hards together
aimed at a common goal and having the thing scoped out into something
achievable. Is the AdaOS project going to be it? I wish them all well and
hope they achieve what they set out to do, but I think its scope needs to be
drawn in a little & some progress made or it remains just a conversation
piece. Maybe it will get somewhere - I wouldn't write it off. Maybe some
spinoff effort derives from it. Its only impossible as long as people
believe it is.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C88E0D1.89161C16@despammed.com...
>
> A pessimist's viewpoint:
>
> To make an operating system successful, you need
> at least one of
>
>  - sufficient quantity of useful applications
>
>  - sufficient geek appeal to attract developers.
>
>  - sufficent robustness/power/features to overcome
>    the lack of either of the others.
>
> To get people to write applications that run on it,
> it must have enough users to make people want to
> write apps for it.  So you have a chicken/egg problem
> for the first point.
>
> For the second, you have the prevailing distaste for
> Ada against you ("Us" if you will).
>
> On the robustness/power/features point, obviously
> Ada has a significant advantage over C/C++/Java.
>
> BUT, when you multiply the language potential
> by the number of developers, Linux still comes
> out ahead.  In spite of its implementation language,
> Linux is quite robust, and it has hundreds of
> developers for every AdaOS developer to keep it
> that way.  Shift the target to features, and
> Microsoft will always have you beat there.
> They will always sacrifice security/robustness
> to beat any competition in features.  And their
> "if you can't beat 'em, steal 'em" technique
> also applies.
>
> So on the third point, you have say a hundred
> developers and a language that scores ten on
> some arbitrary rating scale vs. several thousand
> developers and a language that scores one or two.
>
> Do the math and Ada loses.  Unfortunately.
>
> --
> Wes Groleau
> http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Dann Corbit]

"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C88E0D1.89161C16@despammed.com...
>
> A pessimist's viewpoint:
>
> To make an operating system successful, you need
> at least one of
>
>  - sufficient quantity of useful applications
>
>  - sufficient geek appeal to attract developers.
>
>  - sufficent robustness/power/features to overcome
>    the lack of either of the others.
>
> To get people to write applications that run on it,
> it must have enough users to make people want to
> write apps for it.  So you have a chicken/egg problem
> for the first point.
>
> For the second, you have the prevailing distaste for
> Ada against you ("Us" if you will).
>
> On the robustness/power/features point, obviously
> Ada has a significant advantage over C/C++/Java.
>
> BUT, when you multiply the language potential
> by the number of developers, Linux still comes
> out ahead.  In spite of its implementation language,
> Linux is quite robust, and it has hundreds of
> developers for every AdaOS developer to keep it
> that way.  Shift the target to features, and
> Microsoft will always have you beat there.
> They will always sacrifice security/robustness
> to beat any competition in features.  And their
> "if you can't beat 'em, steal 'em" technique
> also applies.
>
> So on the third point, you have say a hundred
> developers and a language that scores ten on
> some arbitrary rating scale vs. several thousand
> developers and a language that scores one or two.
>
> Do the math and Ada loses.  Unfortunately.


There are billions of dollars lost per month due to virus attacks.  Most of
them are buffer overruns.  A secure OS ought to be interesting to someone.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
 "The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm

Re: Ada Operating System

[David Starner]

On Fri, 8 Mar 2002 11:31:12 -0500, Marin David Condic <dont.bother.mcondic.auntie.spam@[> wrote:
> Well, by that reasoning, Linux should have never been able to get going.
> From bottom dead center, it had no useful apps or geek appeal or robustness.
> It had to get started from *somewhere* to have anything to interest geeks in
> developing things for it and it took time before it became something stable
> and reliable, right?

It was a free Unix system for the x86, which a lot of people were
hungering for, especially after the problems with Minix and BSD. It also
came with a decent selection of GNU utils for application to start with.

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably refering to the Internet)

Re: Ada Operating System

[Marin David Condic]

Fair enough. My point was just that Linux (in whatever early or later forms)
at one time didn't exist. (Same for Windows, Unix, MacOS, VMS, etc. etc.
etc.) It had to come into existance and find some following by satisfying
some need that wasn't being filled adequately by something else. Hence,
there is reason to believe that should some Ada operating system emerge at
some point, it too might find some following and utilization.

Sometimes these sorts of things are invested in by a larger company (or
other entity - like the government or a university) because they need some
particular capability or think there is some market out there. Sometimes
projects get built by hobyists or a university professor or students or
whomever to satisfy some curiosity, or to have fun or to go tilting at
windmills. However it gets there, if it offers something useful to the rest
of the world and is made available in a manner that is accessible (free or
at least affordable) by some large segment of the potential users, it will
likely meet with some success. That's why I think an OS written in Ada that
did some minimal useful stuff could work & find acceptance.

Of course we'll never know if we never try. :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"David Starner" <dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote in message
news:a6clko$8i41@news.cis.okstate.edu...
>
> It was a free Unix system for the x86, which a lot of people were
> hungering for, especially after the problems with Minix and BSD. It also
> came with a decent selection of GNU utils for application to start with.
>

Re: Ada Operating System

[Wes Groleau]

> Well, by that reasoning, Linux should have never been able to get going.
> From bottom dead center, it had no useful apps or geek appeal or robustness.

Good point!  OTOH, it did have _some_ geek appeal--the idea of an
open source, no cost operating system.  And of course, AdaOS would
have the same appeal--were it not for the fact that it's written in
"the dreaded Ada"

> Maybe I'm not so pessimistic about the possibilities. One thing is
> absolutely certain - if one starts with an attitude of "It'll never work,
> its doomed to fail, there's no hope..." then guaranteed it won't happen.

True again.  I have always laughed at the "no user-serviceable
parts inside" stickers.  As a result, I've fixed a few things
that supposedly can't be fixed.  (Made a few worse, too, but
came out ahead overall.)

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

> There are billions of dollars lost per month due to virus attacks.  Most of
> them are buffer overruns.  A secure OS ought to be interesting to someone.

To some, true.  But if security were interesting
to very many, they would take a few seconds to
install some on their computers.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Marin David Condic]

"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C8A2CDB.D8E1B4AA@despammed.com...
>
> Good point!  OTOH, it did have _some_ geek appeal--the idea of an
> open source, no cost operating system.  And of course, AdaOS would
> have the same appeal--were it not for the fact that it's written in
> "the dreaded Ada"
>
I kind of meant at the point in time when there was no Linux code at all.
Once there was code available to play with, it started breeding geeks who
wanted to play with it. Not unlike leaving a mayonase jar out in the sun...
:-)

I think there are enough Ada-Geeks (do we get pocket protectors with a
special logo to identify us? :-) out there who would start playing with an
OS if there was some usable core available. Not *all* geeks (or at least
freebie software developers) program in C.


>
> True again.  I have always laughed at the "no user-serviceable
> parts inside" stickers.  As a result, I've fixed a few things
> that supposedly can't be fixed.  (Made a few worse, too, but
> came out ahead overall.)
>
You know its bad when the last step in assembling a Heath kit is to afix the
"No user-servicable parts inside" sticker to the project you just built.

The "Guy" Test:

You are visited by space aliens who give you a mysterious black box with
buttons and lights and switches on it. They claim it can bring about world
peace, put an end to starvation and cure all disease. You:

a) Take it to the President of the United States.

b) Take it to the United Nations

c) Take it apart.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Ian S. Nelson]

Wes Groleau wrote:
> A pessimist's viewpoint:
> 
> To make an operating system successful, you need
> at least one of
> 
>  - sufficient quantity of useful applications


Needless to say.  That doesn't make it a worthless project though.
Research and demostrations are powerful.

>  - sufficient geek appeal to attract developers.

I think this clearly exists.  Linux and BSD examples and if you do some 
digging, there are probably 50 to 100 other "free OS" projects being 
done be individuals all over the world.  People are interested in system 
level stuff.


>  - sufficent robustness/power/features to overcome
>    the lack of either of the others.
> 
> To get people to write applications that run on it,
> it must have enough users to make people want to
> write apps for it.  So you have a chicken/egg problem
> for the first point.
> 
> For the second, you have the prevailing distaste for
> Ada against you ("Us" if you will).


I'll go on a limb.  I don't think people hate ada.  People dislike 
change to some degree, they dislike hype and they dislike uphill 
battles.  Outside of a few geekier circles I've never heard anything bad 
about ada other than "The DoD uses it" which isn't really a positive or 
negative statement about Ada.


> On the robustness/power/features point, obviously
> Ada has a significant advantage over C/C++/Java.

Not really.  I think that remains to be seen.  Linux doesn't crash a 
lot.  It's not "weapons grade" but we're not talking about that. Code 
written in Ada isn't going to be better than code in C, C++ or Java by 
default, which is suggested so many times.   It should be cheaper to 
write it in Ada at the same level. Right now, the benchmark is high, BSD 
and Linux have solid reputations for being rock solid, and for 99% of 
the tasks done on computers they are.    Implementing a BSD clone in Ada 
would be fun but you couldn't sell it by saying that it is "more stable" 
becuase you could never show that, at least not in an interesting way.


> BUT, when you multiply the language potential
> by the number of developers, Linux still comes
> out ahead.  In spite of its implementation language,
> Linux is quite robust, and it has hundreds of
> developers for every AdaOS developer to keep it
> that way.  Shift the target to features, and
> Microsoft will always have you beat there.
> They will always sacrifice security/robustness
> to beat any competition in features.  And their
> "if you can't beat 'em, steal 'em" technique
> also applies.
> 
> So on the third point, you have say a hundred
> developers and a language that scores ten on
> some arbitrary rating scale vs. several thousand
> developers and a language that scores one or two.
> 
> Do the math and Ada loses.  Unfortunately.


Well what is the goal?  If you're expecting to develop a kernel or full 
on OS and have it sweep the world by storm, that's probably not going to 
happen.  I've been involved in that industry, I've seen how it works and 
you're right, Ada would lose because there are so many other factors 
that are so much bigger than implementation language.  I think that 
there isn't a killer app any more that you could develop and sell ada to 
the world with.

Now if you're trying to build Ada community and show people that good 
things can be done with Ada, out in the public and not in secure 
environments, then I think that doing any projects in Ada is a good 
thing.  Building a kernel in Ada could be a very fun project, I'll tell 
you right now that it's not likely going to displace Linux, BSD or any 
other big kernel but it could be a fun project none the less and get 
people interested.    Kernels in particular seem to be an area where 
lot's of people like to tinker and play right now,  a kernel in Ada 
could be readable enough and clean enough to allow non-system hackers to 
  play around with.

Ian Nelson

Re: Ada Operating System

[tmoran]

>Imagine that you had a floppy that would let you boot up a garden variety PC
>and be able to load a realtime app you constructed that could access
>standard issue devices (serial port, ethernet card, etc.)
  Wouldn't that be pretty straightforward to put on top of MaRTE?
http://marte.unican.es

Re: Ada Operating System

[Marin David Condic]

"Ian S. Nelson" <nelsonis@earthlink.net> wrote in message
news:3C8A3999.2000301@earthlink.net...
>
> I'll go on a limb.  I don't think people hate ada.  People dislike
> change to some degree, they dislike hype and they dislike uphill
> battles.  Outside of a few geekier circles I've never heard anything bad
> about ada other than "The DoD uses it" which isn't really a positive or
> negative statement about Ada.
>
Well.... There are *some* folks out there who - when Ada is brought up -
will dismiss it as basically a bad idea and worthless. Its of no concern to
them because nobody is telling them they have to use it - so "hate" is
pretty hard to muster when it doesn't impact you.

In a way, that's worse. The oposite of "love" isn't "hate" - its
"indifference". Its easier to turn Hate into Love because at least that
person *cares* about it enough to maybe keep learning about it.

Still, I think a number of people are starting to at least admit that Ada
has a lot of good ideas - just that it isn't worth their time and money to
do anything with it.


>
>
> Not really.  I think that remains to be seen.  Linux doesn't crash a
> lot.  It's not "weapons grade" but we're not talking about that. Code
> written in Ada isn't going to be better than code in C, C++ or Java by
> default, which is suggested so many times.   It should be cheaper to
> write it in Ada at the same level. Right now, the benchmark is high, BSD
> and Linux have solid reputations for being rock solid, and for 99% of
> the tasks done on computers they are.    Implementing a BSD clone in Ada
> would be fun but you couldn't sell it by saying that it is "more stable"
> becuase you could never show that, at least not in an interesting way.
>
Put that in the "Good Enough" category. Most people don't need to worry that
much about security and MTBF is only important when it exceeds some pain
threshold to them. But that doesn't mean there are not a lot of people who
*are* concerned with those things.

You could add other interesting and related goals to the design of such an
OS. What about "privacy"? Would it be useful if the OS gave you a guarantee
that some app you have can't be blipping your data to some unknown source or
accessing websites without your knowledge/permission? There are rumors from
time to time that Bill Gates's products are saving up your data to be sent
to Micro$oft when you connect to the net - for marketing purposes? Or worse?
Paranoid? Or just seeing things clearly? :-) Such an OS would probably fare
well with the black chopper crowd. (Or are they on to something? :-)


>
>
>
> Well what is the goal?  If you're expecting to develop a kernel or full
> on OS and have it sweep the world by storm, that's probably not going to
> happen.  I've been involved in that industry, I've seen how it works and
> you're right, Ada would lose because there are so many other factors
> that are so much bigger than implementation language.  I think that
> there isn't a killer app any more that you could develop and sell ada to
> the world with.
>
IIRC, Windows NT was something like ten years in the making from inception
to delivery. That from a big company with lots of resources and talented
people who knew how to build operating systems. A hobbyist effort to do
something similar starts looking kind of bleak. That's why I'd suggest going
for something small & achievable and then waiting for interest to build &
other developers to start signing up. You could get there - just don't
expect it to happen in one swell foop.


> Now if you're trying to build Ada community and show people that good
> things can be done with Ada, out in the public and not in secure
> environments, then I think that doing any projects in Ada is a good
> thing.  Building a kernel in Ada could be a very fun project, I'll tell
> you right now that it's not likely going to displace Linux, BSD or any
> other big kernel but it could be a fun project none the less and get
> people interested.    Kernels in particular seem to be an area where
> lot's of people like to tinker and play right now,  a kernel in Ada
> could be readable enough and clean enough to allow non-system hackers to
>   play around with.
>
It *could* supplant Linux, et alia, but just don't expect it to happen
overnight. Come up with something better than a Unix clone or Windows
knock-off and offer something the others don't do, & maybe it gets there.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Marin David Condic]

<tmoran@acm.org> wrote in message
news:m%ri8.2068$bY3.963358932@newssvr21.news.prodigy.com...
> >Imagine that you had a floppy that would let you boot up a garden variety
PC
> >and be able to load a realtime app you constructed that could access
> >standard issue devices (serial port, ethernet card, etc.)
>   Wouldn't that be pretty straightforward to put on top of MaRTE?
> http://marte.unican.es

Maybe that constitutes "A Good Start". I'd have to look at it more than a
brief view of the web page you posted, but it sure looks like its at least
on the right track.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Hyman Rosen]

Marin David Condic wrote:
> What about "privacy"? Would it be useful if the OS gave you a guarantee
> that some app you have can't be blipping your data to some unknown source or
> accessing websites without your knowledge/permission?

I'm afraid your love for Ada has blinded you to the fact that not
all software errors are because of buffer overflow. Many of the
errors and vulnerabilities that show up these days are "cross-
scripting" errors and the like, where programs execute externally
submitted scripting code because they fail to properly validate
inputs. This is purely an error of program logic - no automated
language checking is going to catch this.

Re: Ada Operating System

[Larry Kilgallen]

In article <3C8B0191.3080705@mail.com>, Hyman Rosen <hyrosen@mail.com> writes:
> Marin David Condic wrote:
>> What about "privacy"? Would it be useful if the OS gave you a guarantee
>> that some app you have can't be blipping your data to some unknown source or
>> accessing websites without your knowledge/permission?
> 
> I'm afraid your love for Ada has blinded you to the fact that not
> all software errors are because of buffer overflow. Many of the
> errors and vulnerabilities that show up these days are "cross-
> scripting" errors and the like, where programs execute externally
> submitted scripting code because they fail to properly validate
> inputs. This is purely an error of program logic - no automated
> language checking is going to catch this.

But automatic language checking does provide the programmer with
more time to worry about those things that only humans can handle.
I rarely see Norm Abrams on PBS using a hand tool in his carpentry.

I do not believe that the automatic checking provided by Ada makes
programmers behave sloppily.  There is some pride in writing code
that the compiler will accept on the first try.  From there on,
the individual zeal for quality can only be further sustained by
addressing those problems that are not susceptible to automation.

But perhaps it would be otherwise if a different class of programmers
were using Ada.  We should be thankful there are other languages to
draw those people who are only interested in what is the latest fad.

Re: Ada Operating System

[Matthew Woodcraft]

Hyman Rosen <hyrosen@mail.com> writes:

> Many of the errors and vulnerabilities that show up these days are
> "cross-scripting" errors and the like, where programs execute
> externally submitted scripting code because they fail to properly
> validate inputs. This is purely an error of program logic - no
> automated language checking is going to catch this.

People can and do include the 'trustedness' of data in their type
systems. Consider taintperl, for an example with dynamic typing. There's
no reason why this couldn't be done with static typing too.

-M-

Re: Ada Operating System

[David Starner]

On Sun, 10 Mar 2002 06:44:05 GMT, Hyman Rosen <hyrosen@mail.com> wrote:
> Many of the
> errors and vulnerabilities that show up these days are "cross-
> scripting" errors and the like, where programs execute externally
> submitted scripting code because they fail to properly validate
> inputs. This is purely an error of program logic - no automated
> language checking is going to catch this.

It can catch at least part of this - think sandboxing or Perl's taint
mode. That's a problem for the scripting language, though.

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably refering to the Internet)

Re: Ada Operating System

[Hyman Rosen]

Larry Kilgallen wrote:
> I do not believe that the automatic checking provided by Ada makes
> programmers behave sloppily.  There is some pride in writing code
> that the compiler will accept on the first try.  From there on,
> the individual zeal for quality can only be further sustained by
> addressing those problems that are not susceptible to automation.

I point out that here in c.l.a. in a recent thread it was reported
that some pice of Adaware (I don't remember if it was GNAT itself or
some associated package) failed to run when installed in a directory
whose pathname contained blanks. This is excatly the kind of issue
that leads to scripting problems - the piece of software fails to
distinguish between blanks which are part of a name and blanks which
are separators. So if you think that Ada programmers will avoid these
kinds of problems because of some kind of inbuilt virtuosity, you can
see that you are wrong.

Re: Ada Operating System

[Hyman Rosen]

David Starner wrote:
> It can catch at least part of this - think sandboxing or Perl's taint
> mode. That's a problem for the scripting language, though.

But this sort of thing isn't part of Ada, so if AdaOS will have it,
it will be because some decides to implement it. It then becomes
difficult to argue that the safety of AdaOS is due to the safety of
Ada.

Re: Ada Operating System

[Simon Wright]

"Dann Corbit" <dcorbit@connx.com> writes:

> Has anyone tried to build a complete operating system in Ada?  Does
> anyone know of a paper or URL for such an effort?

http://www-vs.informatik.uni-ulm.de/DOSinWWW/TextFiles/DOSystems/PULSE.html

(I came across PULSE in a copy of the first reference at the URL that
I caught on its way to the skip in a recent clearout. It is described
as "Ada-based", though)

-- 
Simon Wright                         Email: simon.j.wright@amsjv.com
Alenia Marconi Systems                     Voice: +44(0)23 9270 1778
Integrated Systems Division                  FAX: +44(0)23 9270 1800

Re: Ada Operating System

[Wes Groleau]

> etc.) It had to come into existance and find some following by satisfying
> some need that wasn't being filled adequately by something else. Hence,
> there is reason to believe that should some Ada operating system emerge at
> some point, it too might find some following and utilization.

That's not too different from my point:
some need that isn't filled elsewhere.

Microsoft fills the needs of those who like glitz and hype.

Macintosh fills the needs of those who like visual elegance.

Linux and MANY others fill the needs of those who like to tinker
or who like robustness.

If AdaOS is perceived as just another OS,
the fact that it's in Ada will kill it
in the minds of the ignorant masses.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

> > On the robustness/power/features point, obviously
> > Ada has a significant advantage over C/C++/Java.
> 
> Not really.  I think that remains to be seen.  Linux doesn't crash a
> lot.  It's not "weapons grade" but we're not talking about that. Code

That's because thousands of people have been fixing its bugs
for ... how long?  As good as Ada is in terms of robustness,
a hundred Ada developers cannot compete with ten thousand
C developers.  (A thousand, maybe, but not ten thousand.)

That's why I agree with Marin that it must meet some otherwise
unfilled need.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

Hyman Rosen wrote:
> 
> Marin David Condic wrote:
> > What about "privacy"? Would it be useful if the OS gave you a guarantee
> > that some app you have can't be blipping your data to some unknown source or
> > accessing websites without your knowledge/permission?
> 
> I'm afraid your love for Ada has blinded you to the fact that not
> all software errors are because of buffer overflow. Many of the

Huh?!?!  Marin said nothing about buffer overflows.
He didn't even say anything about errors.  The above
is a reference to spy code put in intentionally by the
softwware vendor.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Wes Groleau]

> I point out that here in c.l.a. in a recent thread it was reported
> that some pice of Adaware (I don't remember if it was GNAT itself or
> some associated package) failed to run when installed in a directory
> whose pathname contained blanks. This is excatly the kind of issue
> that leads to scripting problems - the piece of software fails to

This is exactly the kind of problem that occurs in MANY environments.
Of course Ada is not immune to all oversights.  It's just immune to
more of them than most languages.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Marin David Condic]

O.K. Try re-reading what I wrote and see if I said anything about buffer
overflow or automatic error checking in the part you quote. I'd guess you
just jumped on what you *thought* I said rather than what I *did* say. :-)

What I was suggesting there is that there *might* be some interest on the
part of some users to have some really tight security around all the
possible external gateways within a computer. Some version of firing up a
third-party application that might be making use of the Ethernet to connect
to ports/addresses behind your back and blip data in either direction
without your knowledge. If the OS is doing a good job of monitoring this and
presenting it in an easily interpreted manner such that you can tell that
Bill Gates is attempting to watch you through your use of "Outlook" or
"Explorer", that might have an interest to some people. Being able to block
it might be even more interesting.

I'm not saying you can't do some of that with existing operating systems or
third party security apps. What I'm saying is that an extensive capability
in this area that was inherent in the OS and difficult/impossible to
circumvent might just be a feature that would make such an OS attractive by
offering something new/different/better than what you can get elsewhere.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Hyman Rosen" <hyrosen@mail.com> wrote in message
news:3C8B0191.3080705@mail.com...
> Marin David Condic wrote:
> > What about "privacy"? Would it be useful if the OS gave you a guarantee
> > that some app you have can't be blipping your data to some unknown
source or
> > accessing websites without your knowledge/permission?
>
> I'm afraid your love for Ada has blinded you to the fact that not
> all software errors are because of buffer overflow. Many of the
> errors and vulnerabilities that show up these days are "cross-
> scripting" errors and the like, where programs execute externally
> submitted scripting code because they fail to properly validate
> inputs. This is purely an error of program logic - no automated
> language checking is going to catch this.
>

Re: Ada Operating System

[Marin David Condic]

There's absolutely no getting around the fact that not all errors in
programs are the kinds of things that can be caught by a compiler. Weak
design, incorrect logic, failure to check all conditions, etc. are all
things that no programming language can make up for with compile or runtime
checks. I don't know that anyone here has ever contended that programming in
Ada was going to result in error-free code.

I think the reasoning goes something like this: You can make logic errors in
*any* programming language. You can make a whole slew of simple programming
errors that are catchable by a compiler (what should we name these? "Coding
Errors"? Let's call them that for the time being.) So if Ada allows you to
make Logic Errors, but not Coding Errors and C/C++ lets you make both Logic
*and* Coding errors, then it stands to reason that in general, programs will
have fewer errors if written in Ada. I always use a spell-checker as an
analogy. No spell-checker will stop me from saying stupid things - but it
can help me catch the more mundane errors in what I write & thus reduce the
overall error rate.

BTW: I've had metrics on projects that bear this out. Its not just theory,
but something measurable.

None of that means that an OS written in Ada is going to automagically be a
better thing than Linux or Windows or anything else out on the market. It
*can* be better, but it won't be just by virtue of the fact that it is
written in Ada. I don't know that anyone here ever claimed it would be.

MDC

--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Hyman Rosen" <hyrosen@mail.com> wrote in message
news:3C8C3C4E.9030703@mail.com...
>
> But this sort of thing isn't part of Ada, so if AdaOS will have it,
> it will be because some decides to implement it. It then becomes
> difficult to argue that the safety of AdaOS is due to the safety of
> Ada.
>

Re: Ada Operating System

[Hyman Rosen]

Wes Groleau wrote:
> The above is a reference to spy code put in
 > intentionally by the softwware vendor.

Really? I was assuming that he was talking about
the reliability of Ada carrying over into an OS
developed using it, and that the unauthorized use
was due to code exploiting errors in the code to
bypass normal operation.

If you're just talking about adding some privacy
feature to an OS, I don't see much point in writing
a new OS from scratch just to get that feature. Why
not just add it to an existing OS?

Re: Ada Operating System

[Hyman Rosen]

Wes Groleau wrote:
> Microsoft fills the needs of those who like glitz and hype.

Microsoft also fills the need of those who need to prepare
spreadsheets and documents, of those who like to play games,
and of those who like to purchase from the enormous variety
of software available to fill every need under the sun.

Re: Ada Operating System

[Hyman Rosen]

Wes Groleau wrote:
  This is exactly the kind of problem that occurs in MANY environments.
> Of course Ada is not immune to all oversights.  It's just immune to
> more of them than most languages.

You did not respond to my point here. Larry Kilgallen appeared to
be saying that Ada *programmers* were more immune to this kind of
error, either because Ada's checking gave them more time to find
higher-level bugs, or because Ada programmers were of a higher
class. You therefore have illustrated *my* point exactly, that Ada
programmers have made the same oversight that occurs in many other
environments.

Re: Ada Operating System

[Marin David Condic]

Yeah. You could do that. You could take the Linux code and add just about
any ideas you wanted to add and not have to build an OS from bottom dead
center.

But that was not the point I was trying to make. The point was: "If you want
to write an OS in Ada because you like Ada and want to help promote the use
of Ada then including features/capabilities not found in other OS's would
help make this hypothetical OS attractive to some segment of the market and
help promote its use."

IOW, an Ada application being developed to help promote Ada needs to offer
something more/better than similar applications in other languages. If the
Ada app just duplicates line-for-line something done in C or Java or
something else, then what incentive is there to use the Ada version? Nobody
cares much about what's inside unless it makes some kind of difference to
them outside.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Hyman Rosen" <hyrosen@mail.com> wrote in message
news:3C8CCA42.4080506@mail.com...
>
> If you're just talking about adding some privacy
> feature to an OS, I don't see much point in writing
> a new OS from scratch just to get that feature. Why
> not just add it to an existing OS?
>

Re: Ada Operating System

[Wes Groleau]

Hyman Rosen wrote:
> 
> Wes Groleau wrote:
>   This is exactly the kind of problem that occurs in MANY environments.
> > Of course Ada is not immune to all oversights.  It's just immune to
> > more of them than most languages.
> 
> You did not respond to my point here. Larry Kilgallen appeared to
> be saying that Ada *programmers* were more immune to this kind of
> error, either because Ada's checking gave them more time to find
> higher-level bugs, or because Ada programmers were of a higher
> class. You therefore have illustrated *my* point exactly, that Ada
> programmers have made the same oversight that occurs in many other
> environments.

Wrong.  Ada programmers (some of them) are less vulnerable
to two categories of bugs:

1. Those that the language prevents.
2. Those that they don't commit because category (1)
   leaves them with more available brain power.

I say "some" because there are the lazy ones who let that
freed up brainpower remain idle, but .....

Of course, in the C world, these same lazy ones would not
bother to run 'lint' or to proof-read their code.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Hyman Rosen]

Wes Groleau wrote:
> Wrong.

Why wrong? As reported in another thread, GNAT 3.13p didn't
work if it was installed in a directory whose name contained
a space. Presumably GNAT distributions are prepared by Ada
programmers. And it has been quite some time since the default
place to install Windows programs was 'Program Files'. Can you
blame me, therefore, for believing that Ada programmers will
be as susceptible to making cross-scripting erros as anyone
else?

Re: Ada Operating System

[Wes Groleau]

> blame me, therefore, for believing that Ada programmers will
> be as susceptible to making cross-scripting erros as anyone
> else?

Actually, I misread your post, sorry.  You said that
"Ada programmers have made the same oversight that occurs..."

I interpreted it as "...make the same oversights..."
A subtle difference in English.  The latter would
mean that Ada programs are subject to exactly the
same bugs as any other language.  Which is not true.

But what you actually said would mean that they
are not immune to one particular kind of bug.
Which IS true.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[James Ross]

On Thu, 7 Mar 2002 10:48:56 -0500, "Marin David Condic"
<dont.bother.mcondic.auntie.spam@[acm.org> wrote:

>A bootstrap & scheduler that could execute some toy programs would provide
>something hobyists could start to play with & provide a starting point for
>something that might actually be a product that could be used commercially.
>It gets the ball rolling.

I agree that the method I proposed is not very attractive because
until you have a no kidding Ada kernel then all you have is just YALD
(Yet Another Linux Distro)

>You want to team up & go tilting at windmills? :-) I may be able to find
>some time for a side project. 

Tilting at windmills.. Hmm, now there is an expression I am not
familiar with :) 

I have been waiting to see if the AdaOS project is ever going to
release some code.  But apparently, I might be waiting a long time.   

Sure, I'd be willing to give a go at it, if you mean to just start an
alternate project to the AdaOS one?  Would it be GPL, BSD, or roll our
own?   

I have a  learning curve with Ada 95, OS design, RT design, & x86
protected mode programming... but I am a programmer! :) I did some
modifications on a simulator's sub-system that was in Ada 83 some
years ago. It ran on a DOS / 286 box in real mode. If I remember
correctly the real-time'ness of it was to replace the PC timer
interrupt vector (the 18.2 hertz)  with it it's own and then the main
loop executed certain routines depending on what cycle it was on.   I
also have substantial assembly experience, but not on Intel/PC.

>(Although I've already got a half-dozen or so
>going already & none of them making fast enough progress! :-)

I know what you mean.  I am pretty much in the same category.  Besides
my personal pet projects that seem to never get done, I also have a
day job.

My view as far as open source ( GPL'd ) projects is that many people
contribute and the burden doesn't lie on just a few people.  Also, all
development code is released daily (or weekly least), working or not.
I don't know how well interactive development is possible on the CVS
like at SourceForge, but some sort of web based interactive source
code modification / tracking / approval process would be nice.  

So what's the first step? Presumably, to get a binary executable image
out of GNAT without any OS dependency.

James Ross

Re: Ada Operating System

[Marin David Condic]

One of the things that came up in this thread was:

http://marte.unican.es/

It looks at first blush like it might constitute A Good Start. I'd suggest
looking it over and seeing what ideas come from there.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:1ujq8ucead9v0ofrqlulldsfri3tpln4bu@4ax.com...
>
> So what's the first step? Presumably, to get a binary executable image
> out of GNAT without any OS dependency.
>

Re: Ada Operating System

[Warren W. Gay VE3WWG]

Hyman Rosen wrote:

> Larry Kilgallen wrote:
> 
>> I do not believe that the automatic checking provided by Ada makes
>> programmers behave sloppily.  There is some pride in writing code
>> that the compiler will accept on the first try.  From there on,
>> the individual zeal for quality can only be further sustained by
>> addressing those problems that are not susceptible to automation.
> 
> I point out that here in c.l.a. in a recent thread it was reported
> that some pice of Adaware (I don't remember if it was GNAT itself or
> some associated package) failed to run when installed in a directory
> whose pathname contained blanks. This is excatly the kind of issue
> that leads to scripting problems - the piece of software fails to
> distinguish between blanks which are part of a name and blanks which
> are separators. So if you think that Ada programmers will avoid these
> kinds of problems because of some kind of inbuilt virtuosity, you can
> see that you are wrong.
 
Where I have seen this, is not in Ada code. This was in the shell

script (gate?) that was used for GtkAda. Let's blame the right
software here, eh.
-- 
Warren W. Gay VE3WWG
http://home.cogeco.ca/~ve3wwg

Re: Ada Operating System

[Warren W. Gay VE3WWG]

Marin David Condic wrote:

> Yeah. You could do that. You could take the Linux code and add just about
> any ideas you wanted to add and not have to build an OS from bottom dead
> center.
> 
> But that was not the point I was trying to make. The point was: "If you want
> to write an OS in Ada because you like Ada and want to help promote the use
> of Ada then including features/capabilities not found in other OS's would
> help make this hypothetical OS attractive to some segment of the market and
> help promote its use."
...

> --
> Marin David Condic
> Senior Software Engineer
> Pace Micro Technology Americas    www.pacemicro.com
> Enabling the digital revolution
> e-Mail:    marin.condic@pacemicro.com
> Web:      http://www.mcondic.com/
> 
> "Hyman Rosen" <hyrosen@mail.com> wrote in message
> news:3C8CCA42.4080506@mail.com...
> 
>>If you're just talking about adding some privacy
>>feature to an OS, I don't see much point in writing
>>a new OS from scratch just to get that feature. Why
>>not just add it to an existing OS?

To address security properly, ie. securely, requires that you
examine all aspects of the interaction with the O/S and exposed
interfaces. Otherwise, you are in the precarious position that
Microsoft is in now, trying to close up all the holes that keep
being discovered faster than they can close them. Security does
not work well in the "bag on the side" manner.
-- 
Warren W. Gay VE3WWG
http://home.cogeco.ca/~ve3wwg

Re: Ada Operating System

[Wes Groleau]

> > some associated package) failed to run when installed in a directory
> > whose pathname contained blanks. This is excatly the kind of issue
> > that leads to scripting problems - the piece of software fails to
> > distinguish between blanks which are part of a name and blanks ....
> 
> Where I have seen this, is not in Ada code. This was in the shell
> script (gate?) that was used for GtkAda. Let's blame the right
> software here, eh.

For that matter, there were _two_ problems.

1. The script writer did not consider the possibility
   of a directory name containing a space.

2. The person who named the directory was not aware
   of how painfully some operating systems punish you
   when you have spaces in your file names (or didn't care)

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Pam Kelly]

Hard to know which post to reply to but for what it's worth ...

Linux has a huge following because it's free, there are masses of GPL
applications available for it and there are increasingly a very significant
number of heavy duty commercial applications available.

So rewrite the Linux kernel in Ada. People might try it as a curiosity since
they'll be able to use all their existing applications. It won't be
frightening.

Pick some applications with a long track record of vulnerabilities such as
sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
reported and the Ada versions will be immune. Publicise it.

The security community will start taking an interest. Because it is
compatible with existing applications it's a viable os as it stands. They
start to recommend an incremental move to applications written in Ada.

Then you can substitute a new improved kernel.

We're working on it. If you're interested let me know.


"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C88E0D1.89161C16@despammed.com...
>
> A pessimist's viewpoint:
>
> To make an operating system successful, you need
> at least one of
>
>  - sufficient quantity of useful applications
>
>  - sufficient geek appeal to attract developers.
>
>  - sufficent robustness/power/features to overcome
>    the lack of either of the others.
>
> To get people to write applications that run on it,
> it must have enough users to make people want to
> write apps for it.  So you have a chicken/egg problem
> for the first point.
>
> For the second, you have the prevailing distaste for
> Ada against you ("Us" if you will).
>
> On the robustness/power/features point, obviously
> Ada has a significant advantage over C/C++/Java.
>
> BUT, when you multiply the language potential
> by the number of developers, Linux still comes
> out ahead.  In spite of its implementation language,
> Linux is quite robust, and it has hundreds of
> developers for every AdaOS developer to keep it
> that way.  Shift the target to features, and
> Microsoft will always have you beat there.
> They will always sacrifice security/robustness
> to beat any competition in features.  And their
> "if you can't beat 'em, steal 'em" technique
> also applies.
>
> So on the third point, you have say a hundred
> developers and a language that scores ten on
> some arbitrary rating scale vs. several thousand
> developers and a language that scores one or two.
>
> Do the math and Ada loses.  Unfortunately.
>
> --
> Wes Groleau
> http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Gary Scott]

If we want an open source version of an existing operating system, why
in the world pick an abomination like UNIX.  At least choose something
half-way decent like VMS or even VM.


Pam Kelly wrote:
> 
> Hard to know which post to reply to but for what it's worth ...
> 
> Linux has a huge following because it's free, there are masses of GPL
> applications available for it and there are increasingly a very significant
> number of heavy duty commercial applications available.
> 
> So rewrite the Linux kernel in Ada. People might try it as a curiosity since
> they'll be able to use all their existing applications. It won't be
> frightening.
> 
> Pick some applications with a long track record of vulnerabilities such as
> sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
> reported and the Ada versions will be immune. Publicise it.
> 
> The security community will start taking an interest. Because it is
> compatible with existing applications it's a viable os as it stands. They
> start to recommend an incremental move to applications written in Ada.
> 
> Then you can substitute a new improved kernel.
> 
> We're working on it. If you're interested let me know.
> 
> "Wes Groleau" <wesgroleau@despammed.com> wrote in message
> news:3C88E0D1.89161C16@despammed.com...
> >
> > A pessimist's viewpoint:
> >
> > To make an operating system successful, you need
> > at least one of
> >
> >  - sufficient quantity of useful applications
> >
> >  - sufficient geek appeal to attract developers.
> >
> >  - sufficent robustness/power/features to overcome
> >    the lack of either of the others.
> >
> > To get people to write applications that run on it,
> > it must have enough users to make people want to
> > write apps for it.  So you have a chicken/egg problem
> > for the first point.
> >
> > For the second, you have the prevailing distaste for
> > Ada against you ("Us" if you will).
> >
> > On the robustness/power/features point, obviously
> > Ada has a significant advantage over C/C++/Java.
> >
> > BUT, when you multiply the language potential
> > by the number of developers, Linux still comes
> > out ahead.  In spite of its implementation language,
> > Linux is quite robust, and it has hundreds of
> > developers for every AdaOS developer to keep it
> > that way.  Shift the target to features, and
> > Microsoft will always have you beat there.
> > They will always sacrifice security/robustness
> > to beat any competition in features.  And their
> > "if you can't beat 'em, steal 'em" technique
> > also applies.
> >
> > So on the third point, you have say a hundred
> > developers and a language that scores ten on
> > some arbitrary rating scale vs. several thousand
> > developers and a language that scores one or two.
> >
> > Do the math and Ada loses.  Unfortunately.
> >
> > --
> > Wes Groleau
> > http://freepages.rootsweb.com/~wgroleau


-- 

Gary Scott
mailto:scottg@flash.net

mailto:webmaster@fortranlib.com
http://www.fortranlib.com

Support the GNU Fortran G95 Project:  http://g95.sourceforge.net

Re: Ada Operating System

[James Ross]

Also, I found this (Actually it was in the thread "Device driver in
Ada95" a few days back)  Looks like the author quit working on it some
time back. 

http://web.wanadoo.be/rc.s/AdaOS/

It appears the GNAT runtime issue is addressed.  He makes the
statement "... boots gnatcode, with full support for elaboration ... "
Only a few files to look at actually!  I need to install GNAT on a
Linux box and try this out.  Might be good for jumpstarting the boot
process

>One of the things that came up in this thread was:
>
>http://marte.unican.es/
>
>It looks at first blush like it might constitute A Good Start. I'd suggest
>looking it over and seeing what ideas come from there.
>
>MDC

MaRTE might be an excellent resource for the real-time and scheduling
portions.  The OSKit Project looks interesting too! Another
technically interesting project is at http://www.menuetos.org/ It is
in Asm but some good reference type stuff in there.

Re: Ada Operating System

[David Starner]

On Wed, 13 Mar 2002 01:45:11 GMT, Gary Scott <scottg@flash.net> wrote:
> If we want an open source version of an existing operating system, why
> in the world pick an abomination like UNIX.  At least choose something
> half-way decent like VMS or even VM.

Because Unix runs most existing open source software. Because Unix is
the only ISO standardized OS. Because Unix has been available on what
Joe Hacker has used. And remember, Unix is snake oil. 

"Many Unix fans generously concede that VMS would probably be the
hacker's favorite commercial OS if Unix didn't exist; though true, this
makes VMS fans furious." -- The Jargon file

Seriously, I don't see any reason that would motivate me to clone any
existing operating system besides Windows, Macintosh or Unix. Surely, we
can do better a second time around, and they don't have the huge
software and brain base that would make bug compatibility with them
useful.

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably refering to the Internet)

Re: Ada Operating System

[Marin David Condic]

Rewriting the Linux kernel in Ada doesn't give anybody anything they don't
already have. Why would they glom onto an Ada OS that was identical to a C
OS if the C OS is more widespread, more mature, more supported, more
developed-for, more understood, etc. etc. etc. etc.? That's just Ada going
"Me too!!!" and there's no perceived advantage to grabbing a product that
lags behind something else. Nobody (well, almost nobody) is going to want an
OS or any other software just because it was written in some specific
language.

If an OS is going to be built in Ada, its worth taking a look at it from a
clean slate and doing it in such a way that it exploits the best features of
Ada. It also has to offer the computing world *something* it doesn't
otherwise get by using someone else's product or there will be near-zero
reasons to use it.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Pam Kelly" <al@adalinux.org> wrote in message
news:a6m8h5$rpb$1@helle.btinternet.com...
> Hard to know which post to reply to but for what it's worth ...
>
> Linux has a huge following because it's free, there are masses of GPL
> applications available for it and there are increasingly a very
significant
> number of heavy duty commercial applications available.
>
> So rewrite the Linux kernel in Ada. People might try it as a curiosity
since
> they'll be able to use all their existing applications. It won't be
> frightening.
>
> Pick some applications with a long track record of vulnerabilities such as
> sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
> reported and the Ada versions will be immune. Publicise it.
>
> The security community will start taking an interest. Because it is
> compatible with existing applications it's a viable os as it stands. They
> start to recommend an incremental move to applications written in Ada.
>
> Then you can substitute a new improved kernel.
>
> We're working on it. If you're interested let me know.
>

Re: Ada Operating System

[Marin David Condic]

Now modeling it on VMS would be something interesting - not an exact clone,
but at least something that looked similar. (IMHO, the only reason VMS
wasn't the winner of the OS wars was because DEC wouldn't/couldn't release
it for use on anything other than DEC hardware.) There were things about VMS
that I thought were bad ideas - notably a file system wherein all the DEC
experts would repeatedly warn you not to try to duplicate the "copy" utility
in your own source code. But there were still lots of wonderful things about
it.

Does anybody know of a OS textbook that uses VMS as its model for
discussion?

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Gary Scott" <scottg@flash.net> wrote in message
news:3C8EAFF7.B6CFF9A6@flash.net...
> If we want an open source version of an existing operating system, why
> in the world pick an abomination like UNIX.  At least choose something
> half-way decent like VMS or even VM.
>
>

Re: Ada Operating System

[Marin David Condic]

"David Starner" <dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote in message
news:a6mv1k$aik2@news.cis.okstate.edu...
>
> "Many Unix fans generously concede that VMS would probably be the
> hacker's favorite commercial OS if Unix didn't exist; though true, this
> makes VMS fans furious." -- The Jargon file
>
I *love* it!


> Seriously, I don't see any reason that would motivate me to clone any
> existing operating system besides Windows, Macintosh or Unix. Surely, we
> can do better a second time around, and they don't have the huge
> software and brain base that would make bug compatibility with them
> useful.
>
While it wouldn't be a bad idea to adopt an existing OS as a model for
understanding/implementing concepts, I heartily agree that there is
absolutely no point in duplicating what's already been done. Especially if
it hasn't been done all that well. (I'd put Unix in this category - much of
it looks too organically grown rather than properly engineered.) As any good
parent should strive to do, we should try to make a whole *new* set of
mistakes with our children rather than repeat the same old mistakes of our
own parents. :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Wes Groleau]

> So rewrite the Linux kernel in Ada. People might try it as a curiosity since

This is something I have been putting off doing for a long time.
Not sure it would work, but seemed like the approach that would
have less opposition than others.

> Pick some applications with a long track record of vulnerabilities such as
> sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
> reported and the Ada versions will be immune. Publicise it.

I wouldn't be so sure of that.  When the Ada version has had
ten hours of maintenance, it will be more secure than the C version
with fifty hours of maintenance.  BUT, the C version had _thousands_
of hours of maintenance so long ago that few remember how bad it was
back then.

> We're working on it. If you're interested let me know.

I am slightly interested.  As I said, I had thoughts
of doing that myself, but never got started.  At home
I now use Mach/BSD so I would only be willing to port
components that would work in that environment.  I don't
have a separate machine to devote just to experiments.

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Pam Kelly]

The Ada versions won't need maintenance :-)

Seriously, to prove the point that Ada is in itself superior the Ada
versions would have to stick so closely to the design of the C versions as
to be more or less straight translations - which would mean they would
benefit from past maintenance of the logic.

"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3C8F5D44.2E2E2EE7@despammed.com...
>
> > Pick some applications with a long track record of vulnerabilities such
as
> > sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
> > reported and the Ada versions will be immune. Publicise it.
>
> I wouldn't be so sure of that.  When the Ada version has had
> ten hours of maintenance, it will be more secure than the C version
> with fifty hours of maintenance.  BUT, the C version had _thousands_
> of hours of maintenance so long ago that few remember how bad it was
> back then.
>

Re: Ada Operating System

[Larry Kilgallen]

In article <a6oi8n$fhe$1@helle.btinternet.com>, "Pam Kelly" <al@adalinux.org> writes:
> The Ada versions won't need maintenance :-)
> 
> Seriously, to prove the point that Ada is in itself superior the Ada
> versions would have to stick so closely to the design of the C versions as
> to be more or less straight translations - which would mean they would
> benefit from past maintenance of the logic.

That would also mean they would have to be wart-for-wart compatible,
even when that means turning off Ada safety features.

So what is the point of having an Ada implmentation ?

Re: Ada Operating System

[Pam Kelly]

You're right. There has to be a market - even if the product is free.

Linux has a big market and a reputation amongst people who are knowledgeable
enough to evaluate a new variant.

Free Windows would attract a lot of attention from Bill's lawyers,  so it's
not really a contender, but I also think there'd be a lot of suspicion
amongst the general run of users. It's not the techie's choice.

ps. am I right in thinking that a VMS man was the brains behind NT ?


"David Starner" <dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote in message
news:a6mv1k$aik2@news.cis.okstate.edu...
> On Wed, 13 Mar 2002 01:45:11 GMT, Gary Scott <scottg@flash.net> wrote:
> > If we want an open source version of an existing operating system, why
> > in the world pick an abomination like UNIX.  At least choose something
> > half-way decent like VMS or even VM.
>
> Because Unix runs most existing open source software. Because Unix is
> the only ISO standardized OS. Because Unix has been available on what
> Joe Hacker has used. And remember, Unix is snake oil.
>
> "Many Unix fans generously concede that VMS would probably be the
> hacker's favorite commercial OS if Unix didn't exist; though true, this
> makes VMS fans furious." -- The Jargon file
>
> Seriously, I don't see any reason that would motivate me to clone any
> existing operating system besides Windows, Macintosh or Unix. Surely, we
> can do better a second time around, and they don't have the huge
> software and brain base that would make bug compatibility with them
> useful.
>
> --
> David Starner - starner@okstate.edu
> "It's not a habit; it's cool; I feel alive.
> If you don't have it you're on the other side."
> - K's Choice (probably refering to the Internet)

Re: Ada Operating System

[Pam Kelly]

Is it likely to be necessary to turn off Ada safety features in order to
'more or less' translate something like sendmail into Ada ?

Wouldn't it just depend ... if there had to be significant changes to parts
of an application (in order to retain naturally occurring safety features of
Ada) the point would still be made if a vulnerability occurred in a part of
the C version that was significantly the same as the Ada version.


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:oZVpXxHs63ZX@eisner.encompasserve.org...
> In article <a6oi8n$fhe$1@helle.btinternet.com>, "Pam Kelly"
<al@adalinux.org> writes:
> > The Ada versions won't need maintenance :-)
> >
> > Seriously, to prove the point that Ada is in itself superior the Ada
> > versions would have to stick so closely to the design of the C versions
as
> > to be more or less straight translations - which would mean they would
> > benefit from past maintenance of the logic.
>
> That would also mean they would have to be wart-for-wart compatible,
> even when that means turning off Ada safety features.
>
> So what is the point of having an Ada implmentation ?

Re: Ada Operating System

[Larry Kilgallen]

In article <a6okko$dvr$1@knossos.btinternet.com>, "Pam Kelly" <al@adalinux.org> writes:
> You're right. There has to be a market - even if the product is free.
> 
> Linux has a big market and a reputation amongst people who are knowledgeable
> enough to evaluate a new variant.
> 
> Free Windows would attract a lot of attention from Bill's lawyers,  so it's
> not really a contender, but I also think there'd be a lot of suspicion
> amongst the general run of users. It's not the techie's choice.
> 
> ps. am I right in thinking that a VMS man was the brains behind NT ?

Current discussion in comp.os.vms indicates that the kernel of
NT was found to be quite compatible with the Mica research
project David Cutler had worked on as his last job at DEC.
Even compatible down to the source code comments.  Rumors
have circulated for years regarding some out-of-court
settlement between DEC and Microsoft.

Cutler was the Project Leader for VMS V1, but after that
he worked on other software within DEC.  V1 was before
VMS had clusters.  The current version is V7.3.

Re: Ada Operating System

[Marin David Condic]

"Pam Kelly" <al@adalinux.org> wrote in message
news:a6okko$dvr$1@knossos.btinternet.com...
>
> Free Windows would attract a lot of attention from Bill's lawyers,  so
it's
> not really a contender, but I also think there'd be a lot of suspicion
> amongst the general run of users. It's not the techie's choice.
>
Now a Windows work-alike written in Ada would be interesting. However, the
way it would get crushed is by Microsoft constantly changing the "reference
copy" so that you're forever playing catch-up. Hence, I'd still think that
it would be wise to do something new. Using the basic architecture of
Windows as a model, wouldn't necessarily be a bad thing - you'd just have to
be ready and willing to go your own way when you find a better (or just more
Ada-ish) way of doing it.


> ps. am I right in thinking that a VMS man was the brains behind NT ?
>
My recollection was that the head dude was hired away from DEC and that he
brought with him key people from the VMS project. That's why underneat the
hood, Windows has a lot of similarity to VMS.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Adrian Knoth]

Pam Kelly <al@adalinux.org> wrote:

> Is it likely to be necessary to turn off Ada safety features in order to
> 'more or less' translate something like sendmail into Ada ?
							   ^ no space!

[Ada-sendmail]
> Ada) the point would still be made if a vulnerability occurred in a part of
> the C version that was significantly the same as the Ada version.

I guess there is no way (or even need) to recode all the software just
in another language. There are C-guys out on the net who more or less
know what they're doing.

The only way to improve Ada's influence in today's software-variety
is to start new (and relevant) projects right from the beginning in
Ada.

So to say, don't migrate from C to Ada within sendmail, just write a
full-featured MTA in Ada being that good to be sendmail'Succ.
Sure the world probably doesn't need another MTA (or even MUA), so
times are hard. But freshmeat.net has a lot of projects in competition 
to existing counterparts, so there's always a market for more Ada.

Show the world that software XYZ can work/make things easier and it'll
spread. Did the mass know Qt before KDE?

 
> "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
> news:oZVpXxHs63ZX@eisner.encompasserve.org...

And could you please fix your newsreader? These two lines are considered
an introduction-line. One line. The message-ID is completely irrelevant, 
it is already referenced in your header.

Next, why do you post your answer before the question? Do I want to
read your answer earlier than even knowing to what you're answering?

Do you start a telephone-call with a closing greeting?


-- 
mail: adi@thur.de  	http://adi.thur.de	PGP: v2-key via keyserver

Wie kommts das am Ende des Geldes noch soviel Monat �brig ist?

Re: Ada Operating System

[Pam Kelly]

Nobody is going to want an operating system because it was written in a
particular language, but they might become interested in a language if they
can see it in a context in which they are accustomed to only seeing C.

(I don't think that anyone would want an operating system written to exploit
the best features of a particular language either. People want an operating
system that does what they want it to do.)

I don't think there's a conflict here. It depends what the object is.

If the object is to write a better operating system then write a better
operating system.

If the object is to show that you can write a better operating system in Ada
than you can in any other language, fine.

If the object is to try to improve the quality of the kind of software used
every day for ordinary business activities, particularly on the internet, so
that there are fewer faults and most importantly fewer security breaches
then one way of going about it is to first demonstrate that Ada is a viable
alternative on C's own territory.


"Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> wrote in
message news:a6nld6$i4n$1@nh.pace.co.uk...
> Rewriting the Linux kernel in Ada doesn't give anybody anything they don't
> already have. Why would they glom onto an Ada OS that was identical to a C
> OS if the C OS is more widespread, more mature, more supported, more
> developed-for, more understood, etc. etc. etc. etc.? That's just Ada going
> "Me too!!!" and there's no perceived advantage to grabbing a product that
> lags behind something else. Nobody (well, almost nobody) is going to want
an
> OS or any other software just because it was written in some specific
> language.
>
> If an OS is going to be built in Ada, its worth taking a look at it from a
> clean slate and doing it in such a way that it exploits the best features
of
> Ada. It also has to offer the computing world *something* it doesn't
> otherwise get by using someone else's product or there will be near-zero
> reasons to use it.
>
> MDC
> --
> Marin David Condic
> Senior Software Engineer
> Pace Micro Technology Americas    www.pacemicro.com
> Enabling the digital revolution
> e-Mail:    marin.condic@pacemicro.com
> Web:      http://www.mcondic.com/
>
>
> "Pam Kelly" <al@adalinux.org> wrote in message
> news:a6m8h5$rpb$1@helle.btinternet.com...
> > Hard to know which post to reply to but for what it's worth ...
> >
> > Linux has a huge following because it's free, there are masses of GPL
> > applications available for it and there are increasingly a very
> significant
> > number of heavy duty commercial applications available.
> >
> > So rewrite the Linux kernel in Ada. People might try it as a curiosity
> since
> > they'll be able to use all their existing applications. It won't be
> > frightening.
> >
> > Pick some applications with a long track record of vulnerabilities such
as
> > sendmail or bind. Rewrite them in Ada. There'll be more vulnerabilities
> > reported and the Ada versions will be immune. Publicise it.
> >
> > The security community will start taking an interest. Because it is
> > compatible with existing applications it's a viable os as it stands.
They
> > start to recommend an incremental move to applications written in Ada.
> >
> > Then you can substitute a new improved kernel.
> >
> > We're working on it. If you're interested let me know.
> >
>
>
>

Re: Ada Operating System

[Larry Kilgallen]

In article <a6on6j$k69$1@paris.btinternet.com>, "Pam Kelly" <al@adalinux.org> writes:
> Is it likely to be necessary to turn off Ada safety features in order to
> 'more or less' translate something like sendmail into Ada ?

Considering the number of defects found in sendmail over the years,
I do not think any wart-for-wart copy of it would be significantly
better, regardless of the implementation language.

Re: Ada Operating System

[Larry Kilgallen]

In article <a6nlm3$ibs$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
> Now modeling it on VMS would be something interesting - not an exact clone,
> but at least something that looked similar. (IMHO, the only reason VMS
> wasn't the winner of the OS wars was because DEC wouldn't/couldn't release
> it for use on anything other than DEC hardware.) There were things about VMS
> that I thought were bad ideas - notably a file system wherein all the DEC
> experts would repeatedly warn you not to try to duplicate the "copy" utility
> in your own source code.

So is it trivial to write a Copy command for partitioned data sets
under MVS that works across an SNA network honoring wildcarded
specifications ?

I think even the current MacOS 9 "Copy" command is a bit complex,
certainly running in the background and possibly even parallelizing
its own work.

"Copy" is a capability whose desired effect is obvious to the
human but considerably difficult to implement in a pleasing fashion.

Re: Ada Operating System

[Alfred Hilscher]

David Starner wrote:
> Seriously, I don't see any reason that would motivate me to clone any
> existing operating system besides Windows, Macintosh or Unix. Surely, we

OS/2 ;-)

Re: Ada Operating System

[Alfred Hilscher]

Marin David Condic wrote:
> 
> Rewriting the Linux kernel in Ada doesn't give anybody anything they don't
> already have. Why would they glom onto an Ada OS that was identical to a C
> OS if the C OS is more widespread, more mature, more supported, more
> developed-for, more understood, etc. etc. etc. etc.? 

So choose OS/2 instead. It's not "widespread" ;-)

> Ada. It also has to offer the computing world *something* it doesn't
> otherwise get by using someone else's product or there will be near-zero
> reasons to use it.

Support. And a perspective to be continued.

Re: Ada Operating System

[Marin David Condic]

Never said there weren't other OS's that didn't also make it difficult. But
perhaps we're also talking about something entirely different. We once
approached the issue of being able to get an arbitrary filename, open the
file as a stream of bytes, suck up the file & be able to spit it back out
again & have it be the same thing it once was. (It was a databasey kind of
thing.) The DEC guys strongly recommended we didn't attempt to do that.
After attempting to do so, we also came to that conclusion.

For all the sins of Unix at least it was possible to treat a file as a
stream of raw bytes when it was convenient to do so. (Not that you couldn't
have the files contain all sorts of interesting attributes & organizations -
just that all information needed should be copyable using a byte stream.)
That would be a capability I'd build into an OS if I was building one from
scratch.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:YHWJXo5va3QF@eisner.encompasserve.org...
>
> So is it trivial to write a Copy command for partitioned data sets
> under MVS that works across an SNA network honoring wildcarded
> specifications ?
>
> I think even the current MacOS 9 "Copy" command is a bit complex,
> certainly running in the background and possibly even parallelizing
> its own work.
>
> "Copy" is a capability whose desired effect is obvious to the
> human but considerably difficult to implement in a pleasing fashion.

Re: Ada Operating System

[Marin David Condic]

"Alfred Hilscher" <Alfred.Hilscher@icn.siemens.de> wrote in message
news:3C90C94D.E8BB96F2@icn.siemens.de...
> Marin David Condic wrote:
> >
> > Rewriting the Linux kernel in Ada doesn't give anybody anything they
don't
> > already have. Why would they glom onto an Ada OS that was identical to a
C
> > OS if the C OS is more widespread, more mature, more supported, more
> > developed-for, more understood, etc. etc. etc. etc.?
>
> So choose OS/2 instead. It's not "widespread" ;-)
>
Well...., yeah, but its still going off and doing something that's already
been done before & I just don't see that creating a big incentive to want to
use *your* product versus using *IBM's* product. It might not be bad to look
at OS/2 and try to model what it does (or some of what it does) - but you
still need to break new ground in some way or its just the same old thing in
a different wrapper.


> > Ada. It also has to offer the computing world *something* it doesn't
> > otherwise get by using someone else's product or there will be near-zero
> > reasons to use it.
>
> Support. And a perspective to be continued.

O.K. If, for example, you were a die-hard OS/2 fan, you might conceivably
want an OS/2 clone only with available support. But a) there aren't that
many OS/2 fans relative to other operating systems, b) OS/2 *users* (as
opposed to fans) who were desparate for support would likely switch to some
other OS for which they could already get it c) duplicating something that
has already failed to capture a huge chunk of the market seems like throwing
good money after bad. If you're going to clone something - clone a *winner*.
:-)

My position would be that if an Ada OS is going to be built, it ought to
break some fundamental new ground and somehow or other be "different" from
the crowd or there wouldn't be a significant enough reason for people to go
after it.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/

Re: Ada Operating System

[Marin David Condic]

A straight copy of *anything* in some other language into Ada seems to be
mostly a waste of time & a bad idea. If it doesn't do things The Ada Way
utilizing what is best in Ada, then it won't benefit from Ada. That, and why
bother since the existing application already exists and a duplicate in
another language just seems like the work of the Department of Redundancy
Department.

A *redesign* of an existing application with implementation in Ada could
have some value. If you can identify the inputs and outputs of a given
program and make the innards from Ada, then you might be adding some
reliability. But I still think that it would need to go One Step Beyond the
existing app in terms of functionality or it isn't likely to offer any
significant reason to switch.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:DxCf+2YH8Vr+@eisner.encompasserve.org...
>
> Considering the number of defects found in sendmail over the years,
> I do not think any wart-for-wart copy of it would be significantly
> better, regardless of the implementation language.

Re: Ada Operating System

[Tarjei T. Jensen]

"Wes Groleau wrote:
> > So rewrite the Linux kernel in Ada. People might try it as a curiosity
since
>
> This is something I have been putting off doing for a long time.
> Not sure it would work, but seemed like the approach that would
> have less opposition than others.

This would be futile. It is better to write stuff that makes it easier to
create good Ada software or eases other peoples work.

greetings,

Re: Ada Operating System

[Aidan Skinner]

On Thu, 14 Mar 2002 12:49:10 -0500, Marin David Condic
<dont.bother.mcondic.auntie.spam@[acm.org> wrote in
<a6qnqo$3q9$1@nh.pace.co.uk>:

>  A *redesign* of an existing application with implementation in Ada could
>  have some value. If you can identify the inputs and outputs of a given
>  program and make the innards from Ada, then you might be adding some

Possibly the best target for this sort of thing would be the GNU HURD,
since it's microkernel based you could rewrite one of the servers in
Ada, or write a replacment for one of the stock servers that's better.

http://www.gnu.org/software/hurd/

Course, you'd need to port GNAT first for this to work...

- Aidan
-- 
aidan@velvet.net  http://www.velvet.net/~aidan/  aim:aidans42
finger for pgp key fingerprint: |-----------------------------
01AA 1594 2DB0 09E3 B850        | Money, honey. Cold cash is 
C2D0 9A2C 4CC9 3EC4 75E1        | all that counts

Re: Ada Operating System

[Mark Biggar]

Marin David Condic wrote:
> 
> A straight copy of *anything* in some other language into Ada seems to be
> mostly a waste of time & a bad idea. If it doesn't do things The Ada Way
> utilizing what is best in Ada, then it won't benefit from Ada. That, and why
> bother since the existing application already exists and a duplicate in
> another language just seems like the work of the Department of Redundancy
> Department.

I don't know, a reimplementation of Multix in Ada, suitably tweaked to
the
"Ada Way" would be a really interesting project.

--
Mark Biggar
mark.a.biggar@attbi.com

Re: Ada Operating System

[Marin David Condic]

But that's just it: Doing a rewrite of Multix to include The Ada Way(tm)
would of necessity mandate some redesign. Making a Multix look-alike might
be of some value to somebody, but you wouldn't want to just do a
translation.

I believe that The Ada Way would be much more than a tweak. How would you
get any Ada advantages from strong type checking, etc., unless you review
the design and decide that a bunch of things that used to be kept in raw
UI32's now need to be distinguished from each other? Then you've got to
start thinking about those types needing to be in packages with appropriate
operations on them. Pretty soon, you're tossing out whole chunks of existing
code in Ada-Bar (the original language of implementation) because they can't
be easily translated using the new types/packages. Before you know it,
you're about as far from a "translation" as you can get.

If there's no advantage to writing Ada-tran then there's no advantage to a
simple translation. A plug-compatible redesign of something *might* have
some reason to exist, but I still think if you're going to "get out of the
boat" you'd better go all the way. (Build somethig that serves the original
function, but design it to be new, innovative and go beyond the original
app.)

BTW: A friend of mine once told me of a JCL interpreter written to be a Unix
shell - "because you knew somebody had to do it!" So wierder things have
been done than reimplementing Multix. :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Mark Biggar" <mark.a.biggar@attbi.com> wrote in message
news:3C9249ED.45CD8ED2@attbi.com...
>
> I don't know, a reimplementation of Multix in Ada, suitably tweaked to
> the
> "Ada Way" would be a really interesting project.
>

Re: Ada Operating System

[James Ross]

>If an OS is going to be built in Ada, its worth taking a look at it from a
>clean slate and doing it in such a way that it exploits the best features of
>Ada.

I.e. doing it the "Ada way".  Not that I disagree,  but I am
re-learning Ada at this point and I don't quite think in Ada yet per
say :) Therefore, I question what specifically is there about Ada that
would make the _design_ of an OS better than one that was written in
say C++? (other than the inherent quality / bugs issues)

> It also has to offer the computing world *something* it doesn't
>otherwise get by using someone else's product or there will be near-zero
>reasons to use it.
>
>MDC

Something's better (or of my pet issues with current OS's):
  - Very easy to use / configure / maintain.  (Windows is getting
there with a way to go yet, Linux is still a plague)
  - More interoperability of applications facilitated by the OS (other
than just _Office_)
  - Install of anything should be a painless 1 or 2 click activity
  - Uninstall should leave nothing behind (nor side effects)
  - Impossible to break by accident, and very difficult to do so on
purpose.
  - Built in, native (very fast) SQL data base like file system 
  - Total separation of Data from Applications.  
  - Built in real-time functionality
  - Very High security 
  - Built in Programming / Debugging Tools
JR

Re: Ada Operating System

[DPH]

On Thu, 14 Mar 2002 12:43:30 -0500, "Marin David Condic"
<dont.bother.mcondic.auntie.spam@[acm.org> wrote:

>"Alfred Hilscher" <Alfred.Hilscher@icn.siemens.de> wrote in message
>news:3C90C94D.E8BB96F2@icn.siemens.de...
>> Marin David Condic wrote:
>> >
>> > Rewriting the Linux kernel in Ada doesn't give anybody anything they
>don't
>> > already have. Why would they glom onto an Ada OS that was identical to a
>C
>> > OS if the C OS is more widespread, more mature, more supported, more
>> > developed-for, more understood, etc. etc. etc. etc.?
>>
>> So choose OS/2 instead. It's not "widespread" ;-)
>>
>Well...., yeah, but its still going off and doing something that's already
>been done before & I just don't see that creating a big incentive to want to
>use *your* product versus using *IBM's* product. It might not be bad to look
>at OS/2 and try to model what it does (or some of what it does) - but you
>still need to break new ground in some way or its just the same old thing in
>a different wrapper.

If one is going to do an OS at all, it would seem that there is great
disadvantage to doing something that is really new unless it is
really, really great in some way.  Does anyone know how to do
something "really great"? 

If it isn't "really great", then I think it has to be something
Windows-y or Unix-y like Linux so that it can take advantage of the
pool of people that already know how to deal with these OSs.  That is,
no more training expense need be incurred to use "our" system.

Plus, if one were to do a Windows-y or Unix-y OS, and the
other-language-written OS is given to repeated faults that especially
Windows is famous for and especially in the area of security, and the
Ada OS was able to show many less such faults, then the value of Ada
as a language might thus be shown.  Plus, if an Ada OS could be built
with many less faults, that is a niche that is currently unfilled - an
OS that works like a popular OS but doesn't go belly up in the
presence of a few hackers...

Dave Head

Re: Ada Operating System

[Marin David Condic]

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:jgo59u49ro74pbfvoek6ruqqa5o790mq90@4ax.com...
>
> I.e. doing it the "Ada way".  Not that I disagree,  but I am
> re-learning Ada at this point and I don't quite think in Ada yet per
> say :) Therefore, I question what specifically is there about Ada that
> would make the _design_ of an OS better than one that was written in
> say C++? (other than the inherent quality / bugs issues)
>
Ada is Object Oriented - make the OS services look like the Ada OO model.
Ada has tasking - make the OS tasking look just like Ada.
Ada has protected types - make the OS provide services that look like it.
Ada has generics - make OS services generic where possible.
Ada supports strong typing - make the OS utilize strong typing. (Don't make
everything in it just be raw bytes & words.)
Ada supports runtime checking - make the OS do the same.

I'm sure with some thought a bunch of similar suggestions could be
developed. Look at the system services in your typical Unix implementation
and notice how much they tend to look "C-ish". The idea would be to build an
OS that looked "Ada-ish". That would be a *real* difference and make
programming for the OS lean towards Ada.


>
> Something's better (or of my pet issues with current OS's):
>   - Very easy to use / configure / maintain.  (Windows is getting
> there with a way to go yet, Linux is still a plague)
>   - More interoperability of applications facilitated by the OS (other
> than just _Office_)
>   - Install of anything should be a painless 1 or 2 click activity
>   - Uninstall should leave nothing behind (nor side effects)
>   - Impossible to break by accident, and very difficult to do so on
> purpose.
>   - Built in, native (very fast) SQL data base like file system
>   - Total separation of Data from Applications.
>   - Built in real-time functionality
>   - Very High security
>   - Built in Programming / Debugging Tools
> JR

There you go. Here's a bunch of ideas that might start addressing a need out
there that is not being filled well by other OS's.

I'd still suggest that the initial goal should be to get something minimal
built that works & start adding higher level ideas like these at a later
point. Make sure the underlying kernel provides the support needed to do
spiffy things like this, but concentrate on getting a new & unique kernel up
and running. (If you want things like easy installation, realtime
functionality, security, etc., you clearly need to consider that when
designing the kernel & device driver model, but try to stay "basic" and keep
the rest in your back pocket to pull out later.)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

Re: Ada Operating System

[Robert A Duff]

James Ross <rem.jr@rem.webross.com> writes:

>   - Uninstall should leave nothing behind (nor side effects)

Yeah.  Why do Unix and Windows applications scatter pieces of themselves
all over the place?  Why is "install" more complicated than a copy (with
maybe a decompression)?  Why is "uninstall" more complicated than
deleting a directory tree?

>   - Total separation of Data from Applications.  

Not sure what you mean by that.  Sounds like you're outlawing compilers,
linkers, and loaders, which treat programs as data.

- Bob

Re: Ada Operating System

[Rod Haper]

Marin David Condic wrote:
 >

> Ada is Object Oriented - make the OS services look like the Ada OO model.
> Ada has tasking - make the OS tasking look just like Ada.
> Ada has protected types - make the OS provide services that look like it.
> Ada has generics - make OS services generic where possible.
> Ada supports strong typing - make the OS utilize strong typing. (Don't make
> everything in it just be raw bytes & words.)
> Ada supports runtime checking - make the OS do the same.
> 
> I'm sure with some thought a bunch of similar suggestions could be
> developed. Look at the system services in your typical Unix implementation
> and notice how much they tend to look "C-ish". The idea would be to build an
> OS that looked "Ada-ish". That would be a *real* difference and make
> programming for the OS lean towards Ada.


Yes, I think you've hit upon an excellent idea!  These are some really 
good points that represent both new and inovative features and also 
define a coherent design philosopy.  I very much like this suggestion to 
design the OS to have a strong Ada-ish flavor and bias.

I've been following this Ada OS thread from the beginning it seemed like 
it was just drifting without a clear destination or sense of direction. 
  This sounds like a plan that could produce a unique OS with new 
features/functionality, with hopefully improved 
safety/security/reliabilty, and it would be very Ada friendly.
-- 
Rod

+----------------------------------+
| Rod Haper, rhaper@houston.rr.com |
|     There is a better way ...    |
|   LAP => Linux + Ada95 + Python  |
+----------------------------------+

Re: Ada Operating System

[Preben Randhol]

On Sat, 16 Mar 2002 20:07:45 GMT, Robert A Duff wrote:
> James Ross <rem.jr@rem.webross.com> writes:
> 
>>   - Uninstall should leave nothing behind (nor side effects)
> 
> Yeah.  Why do Unix and Windows applications scatter pieces of themselves
> all over the place?  Why is "install" more complicated than a copy (with
> maybe a decompression)?  Why is "uninstall" more complicated than
> deleting a directory tree?

Because it makes sense to share files between programs in many cases. I
mean you can have data/picture files that more than one program use.
Config files are needed to start daemons etc and it is nice to have them
at one place rather than all over the disc. Libraries are another reason
(and perhaps most important).

Of course the main point is dependancy. I mean wheter the program is
installed in one directory or in several doesn't matter much if your
system has dependancy inconsistencies. That is missing libraries,
data-files, needed external programs etc... The package system that
Debian Linux uses, called Deb (and also RPM which Redhat and others use),
takes care of this for you. 

If I want to install gtkada development package I say:

apt-get install libgtkada1-dev

and then the system checks to see if I'm missing (or if I need a newer
version) any of the packages this package is dependant on and installs
them too.

As for libgtkada1-dev it dependance is such:

Depends: libc6 (>= 2.2.4-4), libglade0, libglib1.2 (>= 1.2.0),
libgnat-3.13p-1 (>= 3.13p-5) | gnat (<< 3.14), libgnat-3.13p-1 (>=
3.13p-5) | gnat (>= 3.13p-5), libgtk1.2 (>= 1.2.10-2.1), libgtkada1 (>=
1.2.12-4), libgtkada1-glade (>= 1.2.12-4), xlibs (>> 4.1.0), libgtkada1
(= 1.2.12-4), libgtkada1-gnome (= 1.2.12-4), libgtkada1-glade (=
1.2.12-4), libc6-dev, libgtk1.2-dev, gnat (>= 3.13p-5), libgl-dev,
libgdk-pixbuf-dev, libglade-dev, libgnome-dev, libpopt-dev, rcs

If I want to remove libgtkada1-dev then I do:

apt-get remove libgtkada1-dev

and to upgrade my entire system I symply do:

apt-get update
apt-get upgrade
-- 
Preben Randhol         �For me, Ada95 puts back the joy in programming.�

Re: Ada Operating System

[James Ross]

>
>>   - Total separation of Data from Applications.  
>
>Not sure what you mean by that.  Sounds like you're outlawing compilers,
>linkers, and loaders, which treat programs as data.
>
>- Bob

Well, actually what I meant to say was a separation between Data and
Applications at a "User" level so to speak.  Not necessarily at a
system or development level.   I think Linux does a better job at this
than Windows but is far from ideal.  (I don't like the *nix way either
because there is no consistency, it's whatever / wherever the
developer decides) 

In the spirit of Ada enforcing rules about coding, then an Ada OS
should be just as anal about enforcing a single standard about where
everything goes, and how it works, in the system.  How things are
named, version info / files / libraries, how things are shared and
tracked should be a function of the OS and not left up to the
programmer to invent his own way every time.

The idea would be to only allow files (including configuration,
documents, data, etc) to be stored in defined areas.  The API's for
this would force a consistent method so there would no guessing where
*anything* was when it was time to back things up, upgrade,  wipe it
clean, etc...  There would be several ways of doing this that would
still allow the users to organize things in his/her way -- especially
if the native file system was based on a database structure as apposed
to a directory tree structure.

Here is what I am getting at:  I always seem to get stuck doing some
PC support / maintenance for other people that are pretty much
clueless.  I ask them, OK, I am going to clear your hard drive of
everything,  are you sure that all your files are backed up to the
network?  Oh yes, yes.  Then a day or two later I invariably find out
that no, they lost stuff because files were buried somewhere in the
"Program Files" directory tree or somewhere else.  Of course it
happens to my own computers when dealing with configuration settings.
I invariably loose most of my settings for the apps installed when
reinstalling Windows from scratch, not a fun thing :) 
JR

Re: Ada Operating System

[Darren New]

James Ross wrote:
> The API's for
> this would force a consistent method so there would no guessing where
> *anything* was when it was time to back things up, upgrade,  wipe it
> clean, etc... 

Windows already has this, as has UNIX. The problem is that the actual
mechanism etc has changed and improved over the years, and compatibility
has been maintained. 

The only way to make this work is either to get it perfect the first
time, or decide you'll live with the mistakes you make for the entire
lifetime of the OS and all its applications. Neither one really seems
commercially viable.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[James Ross]

On Sun, 17 Mar 2002 22:12:58 GMT, Darren New <dnew@san.rr.com> wrote:

>> The API's for
>> this would force a consistent method so there would no guessing where
>> *anything* was when it was time to back things up, upgrade,  wipe it
>> clean, etc... 
>
>Windows already has this, as has UNIX. 

I might be behind on the latest MS API's, but I am sure there is
nothing there that imposes the kind of thing I am suggesting.  For any
version of windows, including XP, and for any version of UNIX out
there, there is nothing in the OS to keep a user and/or program from
writing files anywhere and everywhere except for user permissions.  

For example: as an administrator I can just delete the GNAT\bin folder
without going through a proper uninstall.  Or I can move its entire
contents to WINDOWS\system32 folder if I want.  I can install EditPlus
inside either folder.  Sure, sounds like bizarre things to do -- but
the OS should not allow it regardless what reason I might have for
doing it.

These sort of restrictions would, in my opinion, go a long way to
making an OS more stable / secure / and easy to maintain. 

>The only way to make this work is to get it perfect the first
>time

Isn't this the "Ada Way"? Make sure your interfaces ( aka .ads's) are
near perfect to begin with... then design and redesign the rest down
the road.  As difficult, or impossible that might be, that would be
the goal!  

Of course were talking about a full fledged OS with a new native file
system (that I suggest be based on relational database model) and we
don't even have a kernel yet :)  Just pie in the sky thinking?!
JR

Re: Ada Operating System

[Darren New]

James Ross wrote:
> Isn't this the "Ada Way"? Make sure your interfaces ( aka .ads's) are
> near perfect to begin with... then design and redesign the rest down
> the road.  

Well, what I meant was that the ".ini" interface (for example) now
writes to the registry instead. So the interface was sufficiently good
when it was written that it could be carried over for a decade, but it
still doesn't put everything in one place/one file/etc.

> Of course were talking about a full fledged OS with a new native file
> system (that I suggest be based on relational database model) 

Well, if it's based on the relational database model, then you don't
even *have* files to worry about being in the wrong place. :)

I suggest perhaps you look at the PalmOS, which basically is a
database-oriented file system where you have to do the right things and
put the right configurations in the right places to make things work.
Just as one possible source of ideas.

Another possibility is to ditch the whole concept of persistant storage
being different from "RAM" storage, and (say) have various protected
objects serve as the file system. Then you can say things like "this
protected object maps application identifiers to configuration records"
and not worry about *how* the system maintains persistance thereof.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[James Ross]

On Mon, 18 Mar 2002 03:26:42 GMT, Darren New <dnew@san.rr.com> wrote:

>Well, if it's based on the relational database model, then you don't
>even *have* files to worry about being in the wrong place. :)

Ah, my point exactly! 

>I suggest perhaps you look at the PalmOS, which basically is a
>database-oriented file system ... 

I was not aware of any OS whose file system was basically a database.
I thought my idea was original... not! :) 

>Another possibility is to ditch the whole concept of persistant storage
>being different from "RAM" storage ...

Good suggestions.  Definitely something to look into / think about.
JR

Re: Ada Operating System

[Darren New]

James Ross wrote:
> 
> On Mon, 18 Mar 2002 03:26:42 GMT, Darren New <dnew@san.rr.com> wrote:
> 
> >Well, if it's based on the relational database model, then you don't
> >even *have* files to worry about being in the wrong place. :)
> 
> Ah, my point exactly!

You just have to worry about tables getting rearranged, or people
storing configuration information in application-specific databases. :-)

> I was not aware of any OS whose file system was basically a database.
> I thought my idea was original... not! :)

PalmOS is a modern one. Pr1me OS (yes, with a "1" there) is an older
one, as well as the predicessors to that whose names I forget. Many of
the mainframe systems had database-like file systems, which is why
people complained about complex structured file systems that UNIX did
away with.
 
> >Another possibility is to ditch the whole concept of persistant storage
> >being different from "RAM" storage ...
> Good suggestions.  Definitely something to look into / think about.

By which I mean, for example, something like making the entire quantity
of disk storage into swap space, for example. I mean, that's one way to
think about it. EROS does this, for example, altho I don't think EROS is
a particularly good example of how to do it right.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[James Ross]

On Mon, 18 Mar 2002 05:12:47 GMT, Darren New <dnew@san.rr.com> wrote:

>You just have to worry about tables getting rearranged, or people
>storing configuration information in application-specific databases. :-)

Well...  you have a point!  Of course system structures / tables would
be protected and access allowed through a well defined API.  But even
if you did break the rules, the disorderly data would still be secure
because all the objects whose "created_by" field contained "your_app"
would have been backed up just before I issued the command Uninstall
your_app! :)   
JR

Re: Ada Operating System

[Tucker Taft]

Marin David Condic wrote:
> 
> There's absolutely no getting around the fact that not all errors in
> programs are the kinds of things that can be caught by a compiler. Weak
> design, incorrect logic, failure to check all conditions, etc. are all
> things that no programming language can make up for with compile or runtime
> checks. I don't know that anyone here has ever contended that programming in
> Ada was going to result in error-free code.
> 
> I think the reasoning goes something like this: You can make logic errors in
> *any* programming language. You can make a whole slew of simple programming
> errors that are catchable by a compiler (what should we name these? "Coding
> Errors"? Let's call them that for the time being.) So if Ada allows you to
> make Logic Errors, but not Coding Errors and C/C++ lets you make both Logic
> *and* Coding errors, then it stands to reason that in general, programs will
> have fewer errors if written in Ada. I always use a spell-checker as an
> analogy. No spell-checker will stop me from saying stupid things - but it
> can help me catch the more mundane errors in what I write & thus reduce the
> overall error rate.

I would go a bit further than that.  It is true that Ada is very good at
catching "stupid" errors like leaving out a semicolon, misspelling an
identifier, swapping the order of parameters, indexing with the wrong
variable, etc.  But the surprising thing, once you start using a language
like Ada, is the number of "subtle" errors that it catches.  May favorite
analogy is that of trying to ski down a ski slope.  If there are is just
one (slalom) gate on the slope, pretty much anyone can make it through the gate
and down the slope.  However, once you get a few dozen gates, you have 
to have a "deeply correct" understanding of skiing to make it through
the "gauntlet," even though each gate is checking something "trivial."

In the same way, Ada throws up several "gates" on every line of the
program, both picky compile-time checks, and run-time checks for things
that can't be proved correct at compile-time.  The actual number of
checks performed compared to languages with a weaker typing model
is quite impressive when you start counting them.  All of these
checks are of the "mundane" variety, but the combination of them
all creates a gauntlet that, typically, only a "deeply consistent" 
program can satisfy.  The empirical evidence is that a program
written in a language with as many consistency checks as Ada, once
it makes it through the compile-time and run-time gauntlets, is much
closer to being "correct" than a program with a weaker gauntlet
to pass.  And the overall time required to bring a program to the
desired level of quality is significantly less because of these
gauntlets made up of "trivial" checks.

> BTW: I've had metrics on projects that bear this out. Its not just theory,
> but something measurable.
> 
> None of that means that an OS written in Ada is going to automagically be a
> better thing than Linux or Windows or anything else out on the market. It
> *can* be better, but it won't be just by virtue of the fact that it is
> written in Ada. I don't know that anyone here ever claimed it would be.
> 
> MDC
> 
> --
> Marin David Condic
> Senior Software Engineer
> Pace Micro Technology Americas    www.pacemicro.com
> Enabling the digital revolution
> e-Mail:    marin.condic@pacemicro.com
> Web:      http://www.mcondic.com/
> 
> "Hyman Rosen" <hyrosen@mail.com> wrote in message
> news:3C8C3C4E.9030703@mail.com...
> >
> > But this sort of thing isn't part of Ada, so if AdaOS will have it,
> > it will be because some decides to implement it. It then becomes
> > difficult to argue that the safety of AdaOS is due to the safety of
> > Ada.
> >

-- 
-Tucker Taft   stt@avercom.net   http://www.avercom.net
Chief Technology Officer, AverCom Corporation (A Titan Company) 
Bedford, MA  USA (AverCom was formerly the Commercial Division of AverStar:
http://www.averstar.com/~stt)

Re: Ada Operating System

[Marin David Condic]

Oh sure. I'd agree - I just don't want to get caught making extraordinary
claims that start the whole "Any Competent C Programmer Would..." kinds of
debate. While Ada is busy checking mundane kinds of things, its likely to
surface higher level mistakes. If you're tripping over all sorts of type
conflicts, for example, it probably means you need to revisit your design as
possibly needing some additional thinking. Or it might make it clear that
someone is misusing a structure in some way that reflects incorrect logic.
People often say "If you get an Ada program past the compiler, chances are
it is correct..." - but I just don't want to claim that this is, of
necessity, so.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"Tucker Taft" <stt@avercom.net> wrote in message
news:3C961352.DE25ADF1@avercom.net...
>
> I would go a bit further than that.  It is true that Ada is very good at
> catching "stupid" errors like leaving out a semicolon, misspelling an
> identifier, swapping the order of parameters, indexing with the wrong
> variable, etc.  But the surprising thing, once you start using a language
> like Ada, is the number of "subtle" errors that it catches.  May favorite
> analogy is that of trying to ski down a ski slope.  If there are is just
> one (slalom) gate on the slope, pretty much anyone can make it through the
gate
> and down the slope.  However, once you get a few dozen gates, you have
> to have a "deeply correct" understanding of skiing to make it through
> the "gauntlet," even though each gate is checking something "trivial."
>
> In the same way, Ada throws up several "gates" on every line of the
> program, both picky compile-time checks, and run-time checks for things
> that can't be proved correct at compile-time.  The actual number of
> checks performed compared to languages with a weaker typing model
> is quite impressive when you start counting them.  All of these
> checks are of the "mundane" variety, but the combination of them
> all creates a gauntlet that, typically, only a "deeply consistent"
> program can satisfy.  The empirical evidence is that a program
> written in a language with as many consistency checks as Ada, once
> it makes it through the compile-time and run-time gauntlets, is much
> closer to being "correct" than a program with a weaker gauntlet
> to pass.  And the overall time required to bring a program to the
> desired level of quality is significantly less because of these
> gauntlets made up of "trivial" checks.
>

Re: Ada Operating System

[David Starner]

On Sun, 17 Mar 2002 19:25:09 -0600, James Ross <rem.jr@rem.webross.com> wrote:
> Sure, sounds like bizarre things to do -- but
> the OS should not allow it regardless what reason I might have for
> doing it.

I _hate_ when gratitious restrictions stop me from doing what I need to
do. When the user really, really wants to shoot themselves in the foot,
the system's job is to provide the gun. (Note that Ada follows this rule
- machine insertions, unchecked deallocation, unchecked conversions.)
What if I'm about to rearrange some partitions, so I need to move some
data onto the only stable partition, which happens to be /usr/lib?

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably refering to the Internet)

Re: Ada Operating System

[David Starner]

On Sat, 16 Mar 2002 20:07:45 GMT, Robert A Duff <bobduff@shell01.TheWorld.com> wrote:
> Yeah.  Why do Unix and Windows applications scatter pieces of themselves
> all over the place?  

So PATH=/usr/bin, not
/usr/abcde/bin:/usr/acroread/bin:/usr/bc/bin:/usr/cat/bin ...? So that
shared libraries are actually shared, instead of having a million copies
on hard disk and in memory? 

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably refering to the Internet)

Re: Ada Operating System

[Mats Karlssohn]

Darren New wrote:
%<
> Another possibility is to ditch the whole concept of persistant storage
> being different from "RAM" storage, and (say) have various protected
> objects serve as the file system. Then you can say things like "this
> protected object maps application identifiers to configuration records"
> and not worry about *how* the system maintains persistance thereof.

This is a wery nice idea that I think is in the same spirit as the "intsalled images" in VMS. It also has a likeness with the ideas in Plan-9 but the other way around. In plan-9 everything is files, but the filesystem can be (and is) different for every process.

I think that I like the crossover idea of having every program that is runnable availible as VMS-like installed images, with the ability to
change the parts availible for each process. The two main parts that really need good design to build someting like this on is the security-subsystem and userlevel memory management.

I hope that this rather unstrutctured message makes some sence.
Also please feel free to correct any missunderstandings esp. reg VMS.

-- 
Mats Karlssohn, developer                         mailto:mats@mida.se  
Mida Systemutveckling AB                          http://www.mida.se
Box 64, S-732 22 ARBOGA, SWEDEN
Phone: +46-(0)589-89808   Fax: +46-(0)589-89809

Re: Ada Operating System

[James Ross]

On 19 Mar 2002 06:20:33 GMT, David Starner
<dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote:

>I _hate_ when gratitious restrictions stop me from doing what I need to
>do. When the user really, really wants to shoot themselves in the foot,
>the system's job is to provide the gun. 

I'll agree with that to a point.  A very secure OS should have an
"unsafe" mode where all bets are off -- do whatever you want. However,
under normal conditions no foot wounding allowed. 

By the way, It is the OS's responsibility to provide the means for
doing what needs to be done in a safe and secure way. 

The more I think about a "theoretical OS written in Ada", the more I
think that the strongest argument for it, that makes sense to me,  is
for security / stability.  (Not that I think that because it is
written in Ada, magically this would be the case.  It's just the
appropriate analogy in this case)
JR

Re: Ada Operating System

[Preben Randhol]

On Sat, 23 Mar 2002 00:06:58 -0600, James Ross wrote:
> I'll agree with that to a point.  A very secure OS should have an
> "unsafe" mode where all bets are off -- do whatever you want. However,
> under normal conditions no foot wounding allowed. 

How can it the be a secure OS if half the applications you are using are
unsecure? 

-- 
Preben Randhol         �For me, Ada95 puts back the joy in programming.�

Re: Ada Operating System

[David Starner]

On Sat, 23 Mar 2002 12:34:28 +0000 (UTC), Preben Randhol <randhol+abuse@pvv.org> wrote:
> On Sat, 23 Mar 2002 00:06:58 -0600, James Ross wrote:
>> I'll agree with that to a point.  A very secure OS should have an
>> "unsafe" mode where all bets are off -- do whatever you want. However,
>> under normal conditions no foot wounding allowed. 
> 
> How can it the be a secure OS if half the applications you are using are
> unsecure? 

It's not, but that's not something an OS vendor can completely dictate.
All the OS can do is make sure that there are ways for a secure program
to do what needs to be done. If everyone is going to write setuid (or
the equivelent) applications, you can't stop them.

-- 
David Starner - starner@okstate.edu
"It's not a habit; it's cool; I feel alive. 
If you don't have it you're on the other side." 
- K's Choice (probably referring to the Internet)

Re: Ada Operating System

[Larry Kilgallen]

In article <slrna9ou09.21m.randhol+abuse@kiuk0156.chembio.ntnu.no>, Preben Randhol <randhol+abuse@pvv.org> writes:
> On Sat, 23 Mar 2002 00:06:58 -0600, James Ross wrote:
>> I'll agree with that to a point.  A very secure OS should have an
>> "unsafe" mode where all bets are off -- do whatever you want. However,
>> under normal conditions no foot wounding allowed. 
> 
> How can it the be a secure OS if half the applications you are using are
> unsecure? 

That is why not all applications have access to inner mode.  They can
only harm data to which their own user has access.

Re: Ada Operating System

[Darren New]

Actually, I think there's another way that an Ada-based OS could be
superior to a C-based OS. Imagine an OS where all IPC is done via
message passing of messages between queues. That is, when you (for
example) want to read a file, you put the file handle and the "read"
token at the start of a buffer, pass the buffer to the queue for the
file system thread. The file system thread wakes up, fills up your
buffer, and passes it back to you. In a protected-memory system, it
would be helpful if said buffers were always allocated on page
boundaries, with at most one buffer per page, so simple memory-map
manipulations could move the buffer from one address space to the other. 

Would an Ada compiler (perhaps aided and abetted by the linker) be able
to take objects of the "Message" generic (or tagged) type and assure
they're always allocated on page boundaries, and objects of the "Queue"
generic (or tagged) type are always in pages together (so the OS knows
where they are perhaps) in a way that would be tedious and error-prone
in C? Could it be done with appropriate storage pools, so it doesn't
even really need to be specialized in the compiler?

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[Preben Randhol]

On 23 Mar 2002 21:47:25 -0600, Larry Kilgallen wrote:
> In article <slrna9ou09.21m.randhol+abuse@kiuk0156.chembio.ntnu.no>, Preben Randhol <randhol+abuse@pvv.org> writes:
>> On Sat, 23 Mar 2002 00:06:58 -0600, James Ross wrote:
>>> I'll agree with that to a point.  A very secure OS should have an
>>> "unsafe" mode where all bets are off -- do whatever you want. However,
>>> under normal conditions no foot wounding allowed. 
>> 
>> How can it the be a secure OS if half the applications you are using are
>> unsecure? 
> 
> That is why not all applications have access to inner mode.  They can
> only harm data to which their own user has access.

You mean like this: http://www.nsa.gov/selinux/ ?

But if the OS should have a built in unsafe mode?

-- 
Preben Randhol         �For me, Ada95 puts back the joy in programming.�

Re: Ada Operating System

[Marin David Condic]

What about an OS that deliberately let you set up Foot Shooting Mode?
Suppose it was ultra-secure in "Normal" processing, but if the Administrator
were to boot it up with some secret-handshake, etc., it enabled a process to
run with all capabilities? Your process could switch to Supervisor mode,
read & write directly to physical memory & devices, etc? When you need a
safe OS - it runs safe. When the OS gets in your way, move it aside. That
might make for a distinct product that would satisfy a whole class of needs.

And it would be within The Spirit Of Ada - normally we check to make sure
everything is kosher, but if you really want to circumvent the checks you
can.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"James Ross" <rem.jr@rem.webross.com> wrote in message
news:436o9uc7jg590rv5rb1l9v6be8vk49s278@4ax.com...
>
> I'll agree with that to a point.  A very secure OS should have an
> "unsafe" mode where all bets are off -- do whatever you want. However,
> under normal conditions no foot wounding allowed.
>
> By the way, It is the OS's responsibility to provide the means for
> doing what needs to be done in a safe and secure way.
>
> The more I think about a "theoretical OS written in Ada", the more I
> think that the strongest argument for it, that makes sense to me,  is
> for security / stability.  (Not that I think that because it is
> written in Ada, magically this would be the case.  It's just the
> appropriate analogy in this case)
> JR

Re: Ada Operating System

[Marin David Condic]

Here's another way for an Ada OS to distinguish itself while also making it
Ada-ish - encouraging development for it in Ada. If a spec was provided that
described a protected type for communicating between processes, wouldn't
that make for something kind of unique and easily connected-to only by Ada
programs? And if it was based on/associated with an Ada spec for generic or
tagged record based messages then again, you'd have something different that
was a natural encouragement to use Ada. Plus with a lot of built-in
intra/inter-process message passing capabilities, the OS would be naturally
suited to OOProgramming and networking.

It doesn't necessarily have to be big or address every possible need. It
only has to be different and useful.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"Darren New" <dnew@san.rr.com> wrote in message
news:3C9F5DBE.4B71408F@san.rr.com...
> Actually, I think there's another way that an Ada-based OS could be
> superior to a C-based OS. Imagine an OS where all IPC is done via
> message passing of messages between queues. That is, when you (for
> example) want to read a file, you put the file handle and the "read"
> token at the start of a buffer, pass the buffer to the queue for the
> file system thread. The file system thread wakes up, fills up your
> buffer, and passes it back to you. In a protected-memory system, it
> would be helpful if said buffers were always allocated on page
> boundaries, with at most one buffer per page, so simple memory-map
> manipulations could move the buffer from one address space to the other.
>
> Would an Ada compiler (perhaps aided and abetted by the linker) be able
> to take objects of the "Message" generic (or tagged) type and assure
> they're always allocated on page boundaries, and objects of the "Queue"
> generic (or tagged) type are always in pages together (so the OS knows
> where they are perhaps) in a way that would be tedious and error-prone
> in C? Could it be done with appropriate storage pools, so it doesn't
> even really need to be specialized in the compiler?
>
> --
> Darren New
> San Diego, CA, USA (PST). Cryptokeys on demand.
>       Remember, drive defensively if you drink.

Re: Ada Operating System

[James Ross]

On Mon, 25 Mar 2002 17:25:53 GMT, Darren New <dnew@san.rr.com> wrote:

>Imagine an OS where all IPC is done via
>message passing of messages between queues. 

This is an interesting concept. If "all" OS calls used such a
mechanism, then any application written is a target for a client /
server scenario without any redesign (or recompilation for that
matter).  The OS could simply forward the calls to the server running
the app.  It wouldn't matter if the box responding to the call was
your own or across the country!
JR

Re: Ada Operating System

[Marin David Condic]

It would be interesting to build an OS that used some sort of
command/response protocol with message passing and synchronous/asynchronous
operation. Clearly, with the underlying mechanism invisible to the
application, it wouldn't know if it made a request of the OS on the local
chip, another chip on the bus or a processor clear across the country.

The downside of it is that it would be inefficient for the cases where the
requests were local. Some analysis and modeling would need to be done to
determine if, given modern microprocessor speed, the penalty would be too
extreme to warrant its use.

It might be possible to provide *two* API's for the OS. One with standard,
synchronous  subprogram calls and another that duplicated the functionality
with command/response messaging. But utilizing it (from a compiler
perspective) might be way too complicated.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:gr10auodfr11kgtci3pvtj9dufc17r9tn5@4ax.com...
> On Mon, 25 Mar 2002 17:25:53 GMT, Darren New <dnew@san.rr.com> wrote:
>
> >Imagine an OS where all IPC is done via
> >message passing of messages between queues.
>
> This is an interesting concept. If "all" OS calls used such a
> mechanism, then any application written is a target for a client /
> server scenario without any redesign (or recompilation for that
> matter).  The OS could simply forward the calls to the server running
> the app.  It wouldn't matter if the box responding to the call was
> your own or across the country!
> JR
>

Re: Ada Operating System

[Wes Groleau]

> Suppose it was ultra-secure in "Normal" processing, but if the Administrator
> were to boot it up with some secret-handshake, etc., it enabled a process to

You probably didn't mean what it sounds like you mean,
but I'll argue anyway.  :-)

The change of modes idea is good, but it should be by
a documented process with security controls.

A "secret handshake" will not stay secret.  Even if it
is secret, some user somewhere some day will be lucky enough
or unlucky enough to do it by accident.

Murphy has spoken!

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[Darren New]

James Ross wrote:
> 
> On Mon, 25 Mar 2002 17:25:53 GMT, Darren New <dnew@san.rr.com> wrote:
> 
> >Imagine an OS where all IPC is done via
> >message passing of messages between queues.
> 
> This is an interesting concept. 

Right. It's called a "microkernel", btw. :-) I'm not making this stuff
up myself, just so ya know.

It's also how the Amiga worked, except the Amiga had no memory
protections.

It's also how the Ameoba OS works, except that the Ameoba generally
winds up sending things over the LAN instead of thru memory, but that
would be pretty easy to change.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[Marin David Condic]

I know that you believe you understood what you think I said, but I'm not
sure you realize that what you heard is not what I meant... :-)

What I meant was that there would be some process by which the OS could be
booted or otherwise put into a mode that said "I don't care about security -
let me have the machine" and still have some OS services available to you.
There would obviously be a need to make sure that the process by which this
was done would not be so simple as to let it happen accidentally, nor should
it be allowed without some confirmation that the person doing it was
actually allowed to do so. Hence the notion of a "Secret Handshake".

Clearly if this, as yet hypothetical OS, were to be produced and made
available in source code, someone would find any back doors & make them
known. That actually brings up another problem: Assuming you were billing
the system as "secure" by some definition, how would you be sure that any
given distribution *didn't* include a back door? That would be one heck of
an audit process, eh? :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"Wes Groleau" <wesgroleau@despammed.com> wrote in message
news:3CA0A0EA.F0CEEC89@despammed.com...
> You probably didn't mean what it sounds like you mean,
> but I'll argue anyway.  :-)
>
> The change of modes idea is good, but it should be by
> a documented process with security controls.
>
> A "secret handshake" will not stay secret.  Even if it
> is secret, some user somewhere some day will be lucky enough
> or unlucky enough to do it by accident.
>
> Murphy has spoken!
>

Re: Ada Operating System

[Larry Kilgallen]

In article <a7ntns$7hh$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
> What about an OS that deliberately let you set up Foot Shooting Mode?
> Suppose it was ultra-secure in "Normal" processing, but if the Administrator
> were to boot it up with some secret-handshake, etc., it enabled a process to
> run with all capabilities? Your process could switch to Supervisor mode,
> read & write directly to physical memory & devices, etc? When you need a
> safe OS - it runs safe. When the OS gets in your way, move it aside. That
> might make for a distinct product that would satisfy a whole class of needs.

I don't see why a special boot mode is required, compared to just
letting people have privileges (or not).

Today I can audit whether users log into their privileged accounts,
using out-of-the-box capabilities.  I can even control it if I need
to, from an account whose only "privilege" is the ability to turn
on and off the privileged logins.

I don't see any reason to send an Ada Operating System scurrying
in a different direction thaan that which is traditional in multiuser
operating systems.

Re: Ada Operating System

[Brian Catlin]

"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:cj7Hpn44BtRV@eisner.encompasserve.org...
> In article <a6okko$dvr$1@knossos.btinternet.com>, "Pam Kelly"
<al@adalinux.org> writes:
> Current discussion in comp.os.vms indicates that the kernel of
> NT was found to be quite compatible with the Mica research
> project David Cutler had worked on as his last job at DEC.
> Even compatible down to the source code comments.

I don't think it would be possible for the source comments to be the same,
because Mica was written in a language developed by Dave.  This language
resembled Algol, and he was very proud of it.  Bill G. and Steve B. mandated
that NT be written in C.  It would probably have been a much more robust
operating system had they chosen Ada

 -Brian

Re: Ada Operating System

[Wes Groleau]

> known. That actually brings up another problem: Assuming you were billing
> the system as "secure" by some definition, how would you be sure that any
> given distribution *didn't* include a back door? That would be one heck of
> an audit process, eh? :-)

How about a contractual provision that if a backdoor is ever discovered,
the vendor pays some humongous penalty to the purchaser?  :-)

-- 
Wes Groleau
http://freepages.rootsweb.com/~wgroleau

Re: Ada Operating System

[sk]

Hi,

>How about a contractual provision that if a backdoor is 
>ever discovered, the vendor pays some humongous penalty 
>to the purchaser?  :-)

I have thought for a while that if people stopped
trying to chase "script-kiddies" (with no possible
recovery of damage, and contrarily glorifies them) 
and sued the vendors of bad software for damage, the
movement towards secure software would accelerate
rapidly.

-- 
-------------------------------------
-- Merge vertically for real address
-------------------------------------
s n p @ t . o
 k i e k c c m
-------------------------------------

Re: Ada Operating System

[Darren New]

sk wrote:
> and sued the vendors of bad software for damage, the
> movement towards secure software would accelerate
> rapidly.

That's why most software vendors disclaim any warrantees.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.

Re: Ada Operating System

[Marin David Condic]

And the really tough question is: "What would *you* be willing to warrant
about the software *you* write?" While I know that I personally bend every
effort and utilize the best available technology to produce high quality
software, would I be willing to bet the company by claiming that the
software has no defects and that any defects found will be fixed or you have
some legal recourse to sue me for dammages? Probably not for anything beyond
the most trivial software.

The problem comes down to testing and validation. How can we demonstrate to
our lawyer's satisfaction that what we built is solid enough that they can
defend us in court against any and all warranty claims? Can we satisfy
ourselves that "It behaves as we claim it will in the manual" that we would
stand behind it with the company's bank account? What happens if it crashes
because of system services or other software colliding with it or it doesn't
run on some particular configuration or etc..... ??? Do we start writing
checks to all the customers who ever experience a crash while running our
software?

Personally, I might be willing to put out a given product and offer a money
back guarantee if not completely satisfied, while remaining unwilling to
warrant against all defects and all possible damages. I might have
sufficient confidence in an application that I'd believe the returns would
be small. The only problem with that scenario is how is a company to make
sure that the "dissatisfied customer" isn't keeping an "emergency offsite
backup" of the software they are returning? :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"Darren New" <dnew@san.rr.com> wrote in message
news:3CA1F5F8.A81D9EA1@san.rr.com...
>
> That's why most software vendors disclaim any warrantees.
>

Re: Ada Operating System

[John R. Strohm]

"David Starner" <dvdeug@x8b4e53cd.dhcp.okstate.edu> wrote in message
news:a76lbh$8i42@news.cis.okstate.edu...
> On Sun, 17 Mar 2002 19:25:09 -0600, James Ross <rem.jr@rem.webross.com>
wrote:
> > Sure, sounds like bizarre things to do -- but
> > the OS should not allow it regardless what reason I might have for
> > doing it.
>
> I _hate_ when gratitious restrictions stop me from doing what I need to
> do. When the user really, really wants to shoot themselves in the foot,
> the system's job is to provide the gun. (Note that Ada follows this rule
> - machine insertions, unchecked deallocation, unchecked conversions.)

The simplest answer, which is generally unpopular with the C crowd, is that,
to a first approximation, 99.997% of the time, the user does NOT want to
shoot himself in the foot, and reasonably expects the system to KEEP him
from shooting himself in the foot.  The appropriate answer is to provide
mechanisms that will catch him, 99,997 times out of 100,000.  At the same
time, you provide an override mechanism, preferably about three to five
levels deep in "Are you sure?"-equivalents, to WARN him that he is about to
lose toes and blood and ruin his shoes.

The C crowd seems to believe that, because you need this mechanism 3 times
out of 100,000 tries, it must be easily and immediately available 100% of
the time, and the user is required to look out for himself the other 99,997
times.  I have a problem with this approach.

> What if I'm about to rearrange some partitions, so I need to move some
> data onto the only stable partition, which happens to be /usr/lib?

I guess the obvious question has something to do with why your system has
partitions that require periodic manual rearrangement.

Remember, there is NO excuse for poor design.

Re: Ada Operating System

[James Ross]

On Tue, 26 Mar 2002 13:01:44 -0500, "Marin David Condic"
<dont.bother.mcondic.auntie.spam@[acm.org> wrote:

>There would obviously be a need to make sure that the process by which this
>was done would not be so simple as to let it happen accidentally, nor should
>it be allowed without some confirmation that the person doing it was
>actually allowed to do so. Hence the notion of a "Secret Handshake".

I agree that such a "shoot yourself in the foot" mode should be a big
inconvenience or you would have people using that mode to solve common
problems.  I.e. logging in as root on UNIX to do the stuff you have to
do. This mode should be for disaster recovery only and avoided
completely under any other conditions.  A version of the OS could be
compiled specifically for developing the OS itself that could switch
easily between the modes **IF** that were necessary, I am not to sure
that it would be.

One idea that came to mind is that during an install by the Admin a
key would be given to him.  Something equivalent of an OEM key for a
MS product. (aren't those a pain?).  Only on a cold reboot and with
that key would he be able to enter the unsafe mode.  That key would be
stored on the system encrypted using industrial strength encryption to
make it very difficult for hackers to break it even if they did get
passed the other security protecting where it is stored.

>Clearly if this, as yet hypothetical OS, were to be produced and made
>available in source code, someone would find any back doors & make them
>known. That actually brings up another problem: Assuming you were billing
>the system as "secure" by some definition, how would you be sure that any
>given distribution *didn't* include a back door? That would be one heck of
>an audit process, eh? :-)

You could come up with an Open Source License that stipulated that you
can use it for free but you can't publicly distribute it.  Also, any
changes made to the OS code base must be given back to the project.
Then maintain a single distribution point of releases.  Anyone wanting
to make sure they got the "secure" version would then make sure they
got it from that single distributor. This would definitely avoid the
distribution plethora like you have with Linux. 
JR

>
>MDC

Re: Ada Operating System

[tmoran]

>would I be willing to bet the company by claiming that the software has
>no defects and that any defects found will be fixed or you have some
>legal recourse to sue me for dammages?  Probably not for anything beyond
>the most trivial software.
  That's a straw man.  When you drive your car you don't expect to
be liable for accidents that no normal (non stunt) driver could
have prevented.  You are held to a "reasonable" standard, not one
of perfection.  Doctors don't give warranties that their patients
will not die, they only are liable if they do sloppy work.  If your
software company can show it used best practice to avoid bugs, you
will win your case.  If not, you, or your malpractice insurer,
must pay.  In the current state of affairs you (or Microsoft)
would practically need to maliciously insert bugs to lose a suit.

Re: Ada Operating System

[sk]

Hi,

Darren New: Wrote
>That's why most software vendors disclaim any warrantees.

MDC: wrote
>>And the really tough question is: "What would 
>  <snip>
>>backup" of the software they are returning? :-)

Warrantees were not my point.

My original point was that there would be a rapid 
acceleration towards secure software if you prosecuted 
vendors who did not show evidence of some care 
towards security (admittedly, establishing what "some
care" means is tricky).

The publically discussed vulnerabilities of a certain
vendor would suggest a lack of concern which would be
considered negligent in other professions.

-- 
-------------------------------------
-- Merge vertically for real address
-------------------------------------
s n p @ t . o
 k i e k c c m
-------------------------------------

Re: Ada Operating System

[Larry Kilgallen]

In article <a7su97$h6p$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:

> Personally, I might be willing to put out a given product and offer a money
> back guarantee if not completely satisfied, while remaining unwilling to
> warrant against all defects and all possible damages. I might have
> sufficient confidence in an application that I'd believe the returns would
> be small. The only problem with that scenario is how is a company to make
> sure that the "dissatisfied customer" isn't keeping an "emergency offsite
> backup" of the software they are returning? :-)

That is not an issue so long as your software is imperfect.
Once they claim a refund they don't receive updates to the
new better version.

By "imperfect" here, I mean to include "lacking features"
rather than just "containing defects".

Re: Ada Operating System

[Marin David Condic]

"James Ross" <rem.jr@rem.webross.com> wrote in message
news:6c55au8m539tmegu6u7rpli47ik51ssvrm@4ax.com...
>
> I agree that such a "shoot yourself in the foot" mode should be a big
> inconvenience or you would have people using that mode to solve common
> problems.  I.e. logging in as root on UNIX to do the stuff you have to
> do. This mode should be for disaster recovery only and avoided
> completely under any other conditions.  A version of the OS could be
> compiled specifically for developing the OS itself that could switch
> easily between the modes **IF** that were necessary, I am not to sure
> that it would be.
>
Yeah, that's my point. Most of the time it isn't necessary or desirable to
run in a mode that is either realtime or has direct access to the hardware.
I could imagine that three tiers of operation for the OS might exist. One is
"Normal" mode in which it works pretty much like a standard workstation OS.
THe next is "Realtime" mode in which you can create processes that utilize
otherwise undesirable scheduling algorithms. And the third might be
"Glorified Program Loader" mode in which the OS might provide things like
device drivers and interrupt handlers, but otherwise, you've got complete
control of the machine. Getting into anything but "Normal" mode should not
be easy or accidental.


> One idea that came to mind is that during an install by the Admin a
> key would be given to him.  Something equivalent of an OEM key for a
> MS product. (aren't those a pain?).  Only on a cold reboot and with
> that key would he be able to enter the unsafe mode.  That key would be
> stored on the system encrypted using industrial strength encryption to
> make it very difficult for hackers to break it even if they did get
> passed the other security protecting where it is stored.
>
That's one way. The exact hows and whens are a matter of design, but just
simply making it something that doesn't happen just because you've got
privileges is what's important.


>
> You could come up with an Open Source License that stipulated that you
> can use it for free but you can't publicly distribute it.  Also, any
> changes made to the OS code base must be given back to the project.
> Then maintain a single distribution point of releases.  Anyone wanting
> to make sure they got the "secure" version would then make sure they
> got it from that single distributor. This would definitely avoid the
> distribution plethora like you have with Linux.
> JR
>
Well regardless of license and distribution, it comes down to a question of
what sort of promises are you willing to make about a large body of code.
Most folks would say "Without any warranty... but we believe it is A Good
Thing" which amounts to nothing more than marketing because it has no teeth.
I personally wouldn't want to put my business and other assets on the line
over a body of code that I can't control. I *might* be willing to stand
behind a specific executable that I built and rigorously tested, but if I
hand out the source code with it and allow you to recompile it, all bets are
off.

MDC


--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

Re: Ada Operating System

[Marin David Condic]

<tmoran@acm.org> wrote in message
news:h9xo8.1545$bp1.1897680090@newssvr21.news.prodigy.com...
>   That's a straw man.  When you drive your car you don't expect to
> be liable for accidents that no normal (non stunt) driver could
> have prevented.  You are held to a "reasonable" standard, not one
> of perfection.  Doctors don't give warranties that their patients
> will not die, they only are liable if they do sloppy work.  If your
> software company can show it used best practice to avoid bugs, you
> will win your case.  If not, you, or your malpractice insurer,
> must pay.  In the current state of affairs you (or Microsoft)
> would practically need to maliciously insert bugs to lose a suit.

Well, if we were talking a concrete situation over some specific body of
code in a real business endeavor wherein we could consult lawyers in
developing a strategy for what we will or will not warrant, I might agree
with you. I just have a reluctance to rely on a strategy of "My lawyers can
beat up your lawyers!" Lawyers are, after all, a lot like atomic bombs. Once
you use them, everybody gets messed up. :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

Re: Ada Operating System

[Marin David Condic]

"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:vEY3SfNlq6Uc@eisner.encompasserve.org...
>
> That is not an issue so long as your software is imperfect.
> Once they claim a refund they don't receive updates to the
> new better version.
>
> By "imperfect" here, I mean to include "lacking features"
> rather than just "containing defects".

O.K. Maybe. But, for example, I'm running an old version of Microsoft Word
at home. I have no desire to upgrade it to the latest and greatest version
because it does what I want already. Also, I don't recall that Microsoft
offered to send me updates included in the purchase price of Word. If I want
an upgrade, I've got to pay for it.

Hypothetically, I go to CompUSA and buy Word. Install it on my machine.
Return it claiming I'm not happy with it for some reason. (or lots of them!
:-) I get my money back and still have Word doing what I need it to do. A
later version of Word comes on the market and I do the same thing. I keep
getting the upgrades and keep returning them with the story "I was hoping
that *this* version would be better..." What stops that scenario?

Not that I necessarily favor this scenario - but its worth at least
cogitating on: If there were some sort of key to a software package that
allows it to run on one machine and you were to return the software with the
key (so now you can't run it) then offering a money back guarantee is not
much of a problem. You buy it and like it - you use it indefinitely and can
move it from one machine to another. You buy it and don't like it, you
return it and can't use it anymore. That scenario might make vendors more
willing to stand behind their products because they only end up in trouble
if they produce a bad one.

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com

Re: Ada Operating System

[Larry Kilgallen]

In article <a7vb30$l56$1@nh.pace.co.uk>, "Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
> "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
> news:vEY3SfNlq6Uc@eisner.encompasserve.org...
>>
>> That is not an issue so long as your software is imperfect.
>> Once they claim a refund they don't receive updates to the
>> new better version.
>>
>> By "imperfect" here, I mean to include "lacking features"
>> rather than just "containing defects".
> 
> O.K. Maybe. But, for example, I'm running an old version of Microsoft Word
> at home. I have no desire to upgrade it to the latest and greatest version
> because it does what I want already. Also, I don't recall that Microsoft
> offered to send me updates included in the purchase price of Word. If I want
> an upgrade, I've got to pay for it.
> 
> Hypothetically, I go to CompUSA and buy Word. Install it on my machine.
> Return it claiming I'm not happy with it for some reason. (or lots of them!
> :-) I get my money back and still have Word doing what I need it to do. A
> later version of Word comes on the market and I do the same thing. I keep
> getting the upgrades and keep returning them with the story "I was hoping
> that *this* version would be better..." What stops that scenario?

They can charge previous returners a higher price for the next version.
Perhaps not in a CompUSA situation, but certainly when one buys directly
from the producer.  Those who get the second copy in the name of their
father-in-law just to get around the system would just be "leakage",
a common retail phenomenon.

Re: Ada Operating System

[Marin David Condic]

I would suspect that this would be, at minimum, a logistical nightmare. It
might work for someone who is dealing with a relatively small handful of
customers - some custom package being maintained for, say, a hundred or so
users. But for a serious commercial product with wide distribution I doubt
it would be practical.

Now in the case where you're giving the software away and selling support,
there's a whole different dynamic. "Here's your free distribution -
including a warranty - except that your soul remedy is full refund of your
purchase price - which was nothing." If you run the software and find it
buggy & unreliable and you don't want to use it - delete it. But of course,
if it *is* highly reliable, there's a lot less incentive for you to purchase
support, right? :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com


"Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message
news:Aj0bpsr17AIQ@eisner.encompasserve.org...
>
> They can charge previous returners a higher price for the next version.
> Perhaps not in a CompUSA situation, but certainly when one buys directly
> from the producer.  Those who get the second copy in the name of their
> father-in-law just to get around the system would just be "leakage",
> a common retail phenomenon.

Re: Ada Operating System

[Darren New]

Marin David Condic wrote:
> Most folks would say "Without any warranty... but we believe it is A Good
> Thing" which amounts to nothing more than marketing because it has no teeth.

I've only ever seen two kinds of shrink-wrap licenses. (Of course,
custom code that costs $10K+ comes with something different...)

One says "We don't warrant anything. We especially don't warrant that'll
work for you in particular, because you might try to make it do
something it wasn't designed to do."

The other says "We warrant it'll work substantially the way the
documentation says."

Of course, if it didn't, you'd take it back anyway, and "substantially"
generally means "modulo any bugs we might have left in it."

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
      Remember, drive defensively if you drink.