Project: FreeOS

Project: FreeOS

[Alfred Hilscher]

Hi all,

yesterdy - while surfing - I found an interessting link:
http://www.quasarbbs.com/daniel/freeosidx.html There are some people how
want write a free OS/2 clone. They decided to use C (what else :-( ).
But as the project is still in a very early phase, I think about whether
we should participate and try to revise the decision to use Ada instead.
This could be a chance to push Ada. What do you think about ?

Re: Project: FreeOS

[Ted Dennison]

In article <387C8CB3.1276637F@icn.siemens.de>,
  Alfred Hilscher <Alfred.Hilscher@icn.siemens.de> wrote:

> want write a free OS/2 clone. They decided to use C (what else :-( ).
> But as the project is still in a very early phase, I think about
> whether we should participate and try to revise the decision to use
> Ada instead. This could be a chance to push Ada. What do you think
> about ?

I think they'd be way better off using Ada for a project of that scale.
Bit I don't particularly pine for my old OS-2 system.

There is an Ada-based OS project of sorts underway on the AdaPower site.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[Gautier]

> This could be a chance to push Ada. What do you think about ?

Anyway with C they have chances to join the hundreds of
other genial but unfinished `free OSes'... Try dropping
the same in news:fr.comp.lang.ada - there is already a
project (maybe halted...)

-- 
Gautier

_____\\________________\_______\
http://members.xoom.com/gdemont/

Re: Project: FreeOS

[Juergen Pfeifer]

I don't believe that writing an OS will push Ada. Ok, an OS is a fascinating
piece of software, but on the other hand it is nothing more than a necessary
evil. An OS in itself has no value, it is just there to serve applications.

I believe a good way to attract attention for Ada is to write a popular
application or service in Ada. For example there are some projects on the
way to put together a comprehensive game package infrastructure. Games can
be much more complex and sophisiticated than an OS. And Ada is an excellent
choice for that.

J�rgen

"Alfred Hilscher" <Alfred.Hilscher@icn.siemens.de> schrieb im Newsbeitrag
news:<387C8CB3.1276637F@icn.siemens.de>...
> Hi all,
>
> yesterdy - while surfing - I found an interessting link:
> http://www.quasarbbs.com/daniel/freeosidx.html There are some people how
> want write a free OS/2 clone. They decided to use C (what else :-( ).
> But as the project is still in a very early phase, I think about whether
> we should participate and try to revise the decision to use Ada instead.
> This could be a chance to push Ada. What do you think about ?

Re: Project: FreeOS

[David Starner]

On Thu, 13 Jan 2000 01:44:38 +0100, Juergen Pfeifer <juergen.pfeifer@gmx.net> wrote:
>I don't believe that writing an OS will push Ada. Ok, an OS is a fascinating
>piece of software, but on the other hand it is nothing more than a necessary
>evil. An OS in itself has no value, it is just there to serve applications.
>
>I believe a good way to attract attention for Ada is to write a popular
>application or service in Ada. For example there are some projects on the
>way to put together a comprehensive game package infrastructure. Games can
>be much more complex and sophisiticated than an OS. And Ada is an excellent
>choice for that.

Nothing's going to be a quick fix. But an OS is much more primal than a 
game. Games come and go, get uninstalled and reinstalled, and are just
one of many. OS's come one or two to a machine, and changed at great
cost, and nudge all the programmers on that platform in that direction.
(Think about the people who would write C in Ada to avoid the level of
indirection libc would invoke. Okay, maybe this isn't such a great idea . . .)

Another nice thing about an OS, is that Linux has gone through at least 
three points where it was broken on different versions of the compiler 
(bad assembly, illegal aliasing, and abuse of label addressing/inlining). 
I don't see an Ada OS ever having a problem with abusing the standard like 
that.

-- 
David Starner - dstarner98@aasaa.ofe.org
If you wish to strive for peace of soul then believe; 
if you wish to be a devotee of truth, then inquire.
   -- Friedrich Nietzsche

Re: Project: FreeOS

[Larry Kilgallen]

In article <85jd6g$9qe2@news.cis.okstate.edu>, dvdeug@x8b4e53cd.dhcp.okstate.edu (David Starner) writes:

> Another nice thing about an OS, is that Linux has gone through at least 
> three points where it was broken on different versions of the compiler 
> (bad assembly, illegal aliasing, and abuse of label addressing/inlining). 
> I don't see an Ada OS ever having a problem with abusing the standard like 
> that.

That would be a theoretical OS, but since the discussion started with
a goal of something that would be popular it seems likely people could
not resist the temptation to make it backward compatible with something
else, bringing along that baggage.

I believe the most recently originated operating system that is popular
today is MacOS, dating from 1984.  Perhaps it is OS/400, depending on
when System 34 was created.  Certainly MVS, VMS and Unix are quite a
bit older than 1984.

Larry Kilgallen

Re: Project: FreeOS

[Ted Dennison]

In article <2000Jan13.081844.1@eisner>,
  Kilgallen@eisner.decus.org.nospam wrote:

> I believe the most recently originated operating system that is
> popular today is MacOS, dating from 1984.  Perhaps it is OS/400,
> depending on when System 34 was created.  Certainly MVS, VMS and Unix
> are quite a bit older than 1984.

NT isn't really backward-compatable with any previous OS, and was not
created until about 1994 or so I believe.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[David C. Hoos, Sr.]

Ted Dennison <dennison@telepath.com> wrote in message
news:85krc6$2mj$1@nnrp1.deja.com...
> In article <2000Jan13.081844.1@eisner>,
>   Kilgallen@eisner.decus.org.nospam wrote:
>
> > I believe the most recently originated operating system that is
> > popular today is MacOS, dating from 1984.  Perhaps it is OS/400,
> > depending on when System 34 was created.  Certainly MVS, VMS and Unix
> > are quite a bit older than 1984.
>
> NT isn't really backward-compatable with any previous OS, and was not
> created until about 1994 or so I believe.
>
The first betas of NT were released in 1993 (March if I remember correctly).

Re: Project: FreeOS

[Florian Weimer]

Ted Dennison <dennison@telepath.com> writes:

> NT isn't really backward-compatable with any previous OS, and was not
> created until about 1994 or so I believe.

Work on NT started a bit earlier, I think, although it was called OS/2
3.0 then.  This doesn't mean that the OS/2 3.0 actually shipped has
much in common with NT 3.1, the actual OS/2 3.0 is based on OS/2 2.0
technology.

On the other hand, NT includes quite a few backwards-compatibility
features (DOS, WoW, OS/2, POSIX).

Re: Project: FreeOS

[Ted Dennison]

In article <87vh4xwy8d.fsf@deneb.cygnus.argh.org>,
  Florian Weimer <fw@s.netic.de> wrote:

> On the other hand, NT includes quite a few backwards-compatibility
> features (DOS, WoW, OS/2, POSIX).
>

NT's "DOS" window never struck me as anything more than a command shell
that uses MS-DOS syntax for commands. And POSIX and OS/2 just
seemed to be tack-ons, not something that inflitrated the OS's
design in any way. I believe support for both has already been "phased
out".

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[Florian Weimer]

Ted Dennison <dennison@telepath.com> writes:

> > On the other hand, NT includes quite a few backwards-compatibility
> > features (DOS, WoW, OS/2, POSIX).
> 
> NT's "DOS" window never struck me as anything more than a command shell
> that uses MS-DOS syntax for commands. 

Come on, it's a bit more than that.

> And POSIX and OS/2 just seemed to be tack-ons, not something that
> inflitrated the OS's design in any way.

Well, NT was designed to run on a number of RISC platforms, and the
members of the ACE consortium certainly wanted compatibility with their
current systems.  In the beginning, I think, it was important that the
system could run, say, an Ultrix subsystem, although there was never
developed one.  In the end, hardly anyone cared about non-IA32 platforms,
and support was discontinued for one after the other.  The only remaining
one, the Alpha platform, is probably kept only to have a 64 bit system
to experiment with today, given the fact that Microsoft will have to
port NT to a 64 bit RISC platform very soon. ;)

> I believe support for both has already been "phased out".

Yes, of course.  The POSIX subsystem is nothing more than a bad joke, and
the OS/2 subsystem is 1.x only, an the PM add-on was never widely used.
On the other hand, the OS/2 subsystem runs Microsoft Multiplan very well
(no more struggle with EMS ;).

Re: Project: FreeOS

[Terry Sikes]

In article <87zou7hehz.fsf@deneb.cygnus.argh.org>,
Florian Weimer  <fw@s.netic.de> wrote:

>The only remaining one, the Alpha platform, is probably kept only to
>have a 64 bit system to experiment with today, given the fact that
>Microsoft will have to port NT to a 64 bit RISC platform very soon. ;)

Alpha/NT has been officially killed.  Rumor has it that Microsoft is
still building Win2K/64 on Alpha, though, because it only has one
Merced system in-house.  There is some chance it might be resurrected,
but I wouldn't bet on it.  Compaq seems more interested in Linux and
Tru64 Unix on Alpha...

Terry
--
tsikes@netcom.com

Re: Project: FreeOS

[Larry Kilgallen]

In article <85krc6$2mj$1@nnrp1.deja.com>, Ted Dennison <dennison@telepath.com> writes:
> In article <2000Jan13.081844.1@eisner>,
>   Kilgallen@eisner.decus.org.nospam wrote:
> 
>> I believe the most recently originated operating system that is
>> popular today is MacOS, dating from 1984.  Perhaps it is OS/400,
>> depending on when System 34 was created.  Certainly MVS, VMS and Unix
>> are quite a bit older than 1984.
> 
> NT isn't really backward-compatable with any previous OS, and was not
> created until about 1994 or so I believe.

The programming interfaces are compatible with previous Windows efforts
(or advertised as such) and thus it carries the baggage of providing that
backward compatibility.  There is even a way to get a "DOS window".

Larry Kilgallen

Re: Project: FreeOS

[Ted Dennison]

In article <2000Jan13.135331.1@eisner>,
  Kilgallen@eisner.decus.org.nospam wrote:
> In article <85krc6$2mj$1@nnrp1.deja.com>, Ted Dennison
<dennison@telepath.com> writes:
> > NT isn't really backward-compatable with any previous OS, and was
> > not created until about 1994 or so I believe.
>
> The programming interfaces are compatible with previous Windows
> efforts (or advertised as such) and thus it carries the baggage of
> providing that backward compatibility.  There is even a way to get a

They were advertiszed that way? They certianly aren't all that
compatable. Some of the GUI calls may be similar I suppose.

Now Win95 is different, and it uses many of the same calls as NT. And
the NT "DOS" command shell uses the same command syntax as DOS did (for
the most part). But to my knowledge NT not at all backward compatable
with DOS or Windows 3.1. Its internals certianly bear no resemblence.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[Preben Randhol]

"Juergen Pfeifer" <juergen.pfeifer@gmx.net> writes:

| I believe a good way to attract attention for Ada is to write a popular
| application or service in Ada. For example there are some projects on the
| way to put together a comprehensive game package infrastructure. Games can
| be much more complex and sophisiticated than an OS. And Ada is an excellent
| choice for that.

I think that a really good Open-Source (preferably GPL'ed) web-browser
that uses conservative amounts of memory and disc-space might be
better. Or a good groupware/projectmanager software. This is at least
missing in the Open-Source community.

-- 
Preben Randhol -- [randhol@pvv.org] -- [http://www.pvv.org/~randhol/]     
         "Det eneste trygge stedet i verden er inne i en fortelling." 
                                                      -- Athol Fugard

Re: Project: FreeOS

[Ted Dennison]

In article <m3iu0ycix4.fsf@kiuk0156.chembio.ntnu.no>,
  Preben Randhol <randhol@pvv.org> wrote:
> "Juergen Pfeifer" <juergen.pfeifer@gmx.net> writes:
>
> | way to put together a comprehensive game package infrastructure.
> | Games can be much more complex and sophisiticated than an OS. And
> | Ada is an excellent choice for that.
>
> I think that a really good Open-Source (preferably GPL'ed) web-browser
> that uses conservative amounts of memory and disc-space might be
> better. Or a good groupware/projectmanager software. This is at least
> missing in the Open-Source community.

Personally, I'd love to go off and write all these projects. And when
work peters out on OpenToken I probably will do one of them (if they are
still feasable at that point). But sadly even if I quit my job to work
on Open Source Ada projects full time I wouldn't have the time to do all
of the neat projects listed in this thread, much less al the others I'd
like to be able to do.

Its apparent that what's lacking is not good ideas for projects, but
good people willing to get them off the ground. Thus I would like to
personally challenge everyone who has contributed to this thread to put
their money where their mouth is. David Botton has a spot in the
AdaPower lab all warmed up and waiting for you.

For anyone seriously thinking about starting an OpenSource project, I
highly reccommend reading ESR's essays; available online at
http://www.tuxedo.org/~esr/writings/cathedral-bazaar/

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[Alfred Hilscher]

Hi Ted,

you are right. No one of us could do this all himself. The real
challenge is to find enough people to participate. If you once have
found them you can do your own job (see Linus Torvalds).


Ted Dennison wrote:
> Personally, I'd love to go off and write all these projects. And when
> work peters out on OpenToken I probably will do one of them (if they are
> still feasable at that point). But sadly even if I quit my job to work
> on Open Source Ada projects full time I wouldn't have the time to do all
> of the neat projects listed in this thread, much less al the others I'd
> like to be able to do.
> 
> Its apparent that what's lacking is not good ideas for projects, but
> good people willing to get them off the ground. Thus I would like to
> personally challenge everyone who has contributed to this thread to put
> their money where their mouth is. David Botton has a spot in the
> AdaPower lab all warmed up and waiting for you.
> 
> For anyone seriously thinking about starting an OpenSource project, I
> highly reccommend reading ESR's essays; available online at
> http://www.tuxedo.org/~esr/writings/cathedral-bazaar/

Re: Project: FreeOS

[Ted Dennison]

In article <387F0E96.76C0D0A3@icn.siemens.de>,
  Alfred Hilscher <Alfred.Hilscher@icn.siemens.de> wrote:
> you are right. No one of us could do this all himself. The real
> challenge is to find enough people to participate. If you once have
> found them you can do your own job (see Linus Torvalds).

Right. To share from my personal experience, just getting a useful or
promising system published is the key. If it is truly worthwhile people
are perfectly willing to help improve the project. At that point you
just have to try not to drive them away. :-)

But someone has to sit down and devote the time to get the project to
that initial state. And when I say someone, I generally mean someONE. I
don't think it can be effectively done by committe.

Not to pick on Nick et. al., but that is my impression of what is going
on in the "OS in Ada" project ( http://www.adapower.com/lab/adaos.html
). Its suffering from a massive case of what Alan Cox called the
"Town Council" effect. There has been tons of discussion, but all they
really have to show for it so far is a short list of very abstract
objectives. To be specific; it'll be Open Source software in
Object-Oriented Ada, perhaps using literate programming techniques.
I'm guessing that one of two things will happen with this project.
Either Nick will sit down for a month or so and pump out a basic kernel
of his own design for everyone to start hacking on, or it will dissapate
into a cloud of entropy within the next 4 months.

To put it another way, just talking about a project you'd like to see
done isn't particularly helpful. To quote Linus, "show me the source."

(reference-  http://slashdot.org/features/98/10/13/1423253.shtml )

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Project: FreeOS

[Alfred Hilscher]

I think the spread of an (popular) OS is wider than that of any
application (maybe except of Quake or Flight simulator). And this would
show to the common, the use (and usability) of Ada for great projects.
Ok, there maybe other large projects done in Ada, but most of them have
a great disadvantage - they are secret.

Juergen Pfeifer wrote:
> 
> I don't believe that writing an OS will push Ada. Ok, an OS is a fascinating
> piece of software, but on the other hand it is nothing more than a necessary
> evil. An OS in itself has no value, it is just there to serve applications.
> 
> I believe a good way to attract attention for Ada is to write a popular
> application or service in Ada. For example there are some projects on the
> way to put together a comprehensive game package infrastructure. Games can
> be much more complex and sophisiticated than an OS. And Ada is an excellent
> choice for that.
> 
> J�rgen
> 
> "Alfred Hilscher" <Alfred.Hilscher@icn.siemens.de> schrieb im Newsbeitrag
> news:<387C8CB3.1276637F@icn.siemens.de>...
> > Hi all,
> >
> > yesterdy - while surfing - I found an interessting link:
> > http://www.quasarbbs.com/daniel/freeosidx.html There are some people how
> > want write a free OS/2 clone. They decided to use C (what else :-( ).
> > But as the project is still in a very early phase, I think about whether
> > we should participate and try to revise the decision to use Ada instead.
> > This could be a chance to push Ada. What do you think about ?

Re: Project: FreeOS

[Preben Randhol]

Alfred Hilscher <Alfred.Hilscher@icn.siemens.de> writes:

| I think the spread of an (popular) OS is wider than that of any
| application (maybe except of Quake or Flight simulator). And this would
| show to the common, the use (and usability) of Ada for great projects.
| Ok, there maybe other large projects done in Ada, but most of them have
| a great disadvantage - they are secret.

Well how do you get the OS to become popular? Remember Linux was
started in 1991. I'm not so sure a new OS would be able to compete so
well against Linux. It would have to be able to run Linux apps and use
device drivers available for Linux. Otherwise not much hardware would
be supported. My point is that making an OS is a much bigger task than
making popular applications that people need and will use to do their
everyday activity. If they then want to change the app, they need to
learn Ada95 to do so, as the source code would be in Ada95. Thus one
could attract more people to Ada95.

-- 
Preben Randhol                   For me, Ada95 puts back
[randhol@pvv.org]                the joy in programming.
[http://www.pvv.org/~randhol/]   [http://www.gnuada.org/]

Re: Project: FreeOS

[Tarjei T. Jensen]

Preben Randhol wrote
>Well how do you get the OS to become popular? Remember Linux was
>started in 1991. I'm not so sure a new OS would be able to compete so
>well against Linux. It would have to be able to run Linux apps and use
>device drivers available for Linux. Otherwise not much hardware would
>be supported. My point is that making an OS is a much bigger task than
>making popular applications that people need and will use to do their
>everyday activity. If they then want to change the app, they need to
>learn Ada95 to do so, as the source code would be in Ada95. Thus one
>could attract more people to Ada95.


Linus Torvalds had a big advantage: FSG (GNU) and BSD. A lot of the things that
were neccessary for Linux to succeed were already available in C source form.
FSF has consentrated on providing tools so that there is something to run when
the OS finally become available. Among these are: gcc, awk, perl (not from
FSF), groff, etc.

So the message is: Get some good applications and tools written in Ada before
comitting to an operating system.


Alternative: Create an (real time) operating system (microkernel) for embedded
applications in assembly language (like the L3 and L4 microkernels). Then use
Ada to create services. Initially target should probably be pc/104 or some PC
chipset. Then create tools and systems to go with it (scheduler, tcp/ip stack,
terminal drivers, ethernet driver, printer driver, usb drivers, disk (ide,
scsi) drivers, firewire drivers, graphics subsystem, etc). BTW: VMS is not a
bad operating system to use as a model: It is _very_ modular.

Such a system could become successful. I know that rtems exists, but as far as
I understand that at best only supports Ada. And it seems to be mostly C.

When it is mature one could use it as a general computing platform. Which means
that one must create a Unix like environment for application developers and
users.


Greetings,

Re: Project: FreeOS

[Charles Hixson]

"Tarjei T. Jensen" wrote:

> /* snip */
> So the message is: Get some good applications and tools written in Ada before
> comitting to an operating system.
>
> /*snip */

I don't see why the tools need to be in Ada, merely because the OS is.  If the OS
supports the Linux interfaces (as I expect Debian's Hurd will) then the tools can
be freely shared among the two (three) operating systems.  Tools can be written in
Ada and run on Linux, so I don't see why all of the tools that run on the proposed
OS would need to be Ada.

Re: Project: FreeOS

[David Starner]

On Tue, 18 Jan 2000 18:19:41 GMT, Charles Hixson <charleshixsn@earthlink.net> wrote:
>"Tarjei T. Jensen" wrote:
>
>> /* snip */
>> So the message is: Get some good applications and tools written in Ada before
>> comitting to an operating system.
>>
>> /*snip */
>
>I don't see why the tools need to be in Ada, merely because the OS is.  If the OS
>supports the Linux interfaces
              ^^^^^ Posix interfaces. 
>(as I expect Debian's Hurd will) 
Actually the Hurd is continously having problems for funky features -
no pathname limit is a legal Posix option, but not one that Linux
supports, or any major Unix supports. The Hurd doesn't support Linux
interfaces; it supports Posix interfaces. Any new OS that has any plans
of being Unix-like should do the same.
>be freely shared among the two (three) operating systems.  


-- 
David Starner - dstarner98@aasaa.ofe.org
If you wish to strive for peace of soul then believe; 
if you wish to be a devotee of truth, then inquire.
   -- Friedrich Nietzsche

Re: Project: FreeOS

[Preben Randhol]

Charles Hixson <charleshixsn@earthlink.net> writes:

| "Tarjei T. Jensen" wrote:
| 
| > /* snip */
| > So the message is: Get some good applications and tools written in Ada before
| > comitting to an operating system.
| >
| > /*snip */
| 
| I don't see why the tools need to be in Ada, merely because the OS is.  If the OS

Huh? There is no OS written in Ada. The point Jensen was saying that
one should first make programs in Ada and then one could think of
making an OS later.

-- 
Preben Randhol -- [randhol@pvv.org] -- [http://www.pvv.org/~randhol/]     
         "Det eneste trygge stedet i verden er inne i en fortelling." 
                                                      -- Athol Fugard

Re: Project: FreeOS

[Tarjei T. Jensen]

Preben Randhol wrote
>Huh? There is no OS written in Ada. The point Jensen was saying that
>one should first make programs in Ada and then one could think of
>making an OS later.
>


There is at least one, possibly two OSes written in Ada. If my memory serves me
right one was called tunis. I believe the source is/was available somewhere.


Greetings,

Re: Project: FreeOS

[Mario Klebsch]

"Tarjei T. Jensen" <tarjei.jensen@kvaerner.com> writes:

>There is at least one, possibly two OSes written in Ada. If my memory serves me
>right one was called tunis. I believe the source is/was available somewhere.

Isn�t RTEMS (I hope it can be called an OS in this context) written in ADA?

73, Mario
-- 
Mario Klebsch						mario@klebsch.de

Re: Project: FreeOS

[Preben Randhol]

"Tarjei T. Jensen" <tarjei.jensen@kvaerner.com> writes:

| Preben Randhol wrote
| >Huh? There is no OS written in Ada. The point Jensen was saying that
| >one should first make programs in Ada and then one could think of
| >making an OS later.
| >
| 
| There is at least one, possibly two OSes written in Ada. If my memory serves me
| right one was called tunis. I believe the source is/was available somewhere.

OK, but beside the point (in that context) I think :-)

-- 
Preben Randhol -- [randhol@pvv.org] -- [http://www.pvv.org/~randhol/]     
         "Det eneste trygge stedet i verden er inne i en fortelling." 
                                                      -- Athol Fugard

Re: Project: FreeOS

[Bryce Bardin]

Preben Randhol wrote:
...
> Huh? There is no OS written in Ada. The point Jensen was saying that
> one should first make programs in Ada and then one could think of
> making an OS later.
> 
> --
> Preben Randhol -- [randhol@pvv.org] -- 
...

Existence proof:

The F22 software (written in Ada) runs on network of processors using an 
OS which is itself written in Ada.

Re: Project: FreeOS

[Chris Morgan]

Bryce Bardin <bbardin@home.com> writes:

> Existence proof:
> 
> The F22 software (written in Ada) runs on network of processors using an 
> OS which is itself written in Ada.

So does the UKRN submarine command system (where it hasn't been
upgraded to Ada and C++ on Solaris yet). Also the UK Type 23 Frigate
system. 

-- 
Chris Morgan <cm at mihalis.net>                  http://mihalis.net

Re: Project: FreeOS

[Larry Kilgallen]

In article <m3hfgg4w1w.fsf@kiuk0156.chembio.ntnu.no>, Preben Randhol <randhol@pvv.org> writes:

> Well how do you get the OS to become popular? Remember Linux was
> started in 1991. I'm not so sure a new OS would be able to compete so
> well against Linux. It would have to be able to run Linux apps and use
> device drivers available for Linux. Otherwise not much hardware would
> be supported.

The ability of Linux to corner the market is limited to that segment
of the market that wants a Unix-style operating system.  OS/2 marks
a different segment.

If you provide Linux compatibility you are bound by the legacy Unix
behavior, defeating the purpose of a "new" operating system.

Certainly providing useful programs on existing operating systems is
better.

Re: Project: FreeOS

[Preben Randhol]

kilgallen@eisner.decus.org (Larry Kilgallen) writes:

| The ability of Linux to corner the market is limited to that segment
| of the market that wants a Unix-style operating system.  OS/2 marks
| a different segment.

I was talking in more general terms, I wasn't referring to the 
particular effort of making an OS/2 OS :-)

| Certainly providing useful programs on existing operating systems is
| better.

That is what I think too.
-- 
Preben Randhol -- [randhol@pvv.org] -- [http://www.pvv.org/~randhol/]     
         "Det eneste trygge stedet i verden er inne i en fortelling." 
                                                      -- Athol Fugard

Re: Project: FreeOS

[Charles Hixson]

Larry Kilgallen wrote:

> In article <m3hfgg4w1w.fsf@kiuk0156.chembio.ntnu.no>, Preben Randhol <randhol@pvv.org> writes:
>
> /* snip */

> If you provide Linux compatibility you are bound by the legacy Unix
> behavior, defeating the purpose of a "new" operating system.
>
> Certainly providing useful programs on existing operating systems is
> better.

Providing useful programs on existing OS's is desireable, but different folk are interested in
different things.
Also, what limits does providing Linux compatibility impose?  One could, if one desired, restrict
the compatibility to running in a subshell of the OS rather as MSDos runs within Win95 (but
possibly with better isolation).  I suppose that this might result in an OS that was larger than
otherwise needed, but then most of the requirements of Linux would need to be met by any other OS
also, so it might not add that much.

Re: Project: FreeOS

[Alfred Hilscher]

In this special case they want do an OS/2 clone. I think there are still
a lot of OS/2 users who wish to continue with this system.

Preben Randhol wrote:
 Well how do you get the OS to become popular? Remember Linux was
> started in 1991. I'm not so sure a new OS would be able to compete so
> well against Linux. It would have to be able to run Linux apps and use

Re: Project: FreeOS

[Michael Garrett]

Here's my 1 cent.

What about an application like the Visual Basic IDE or Symantec's Visual
Cafe? The system could use a large "Component" library ( written in ADA ) as
well as existing Microsoft Componenet Technology. It could also support ADA
to Java Byte code, making available the use of existing Java Libraries, and
Java Beans.

Application wizards could be provided that not only create "Componenets" of
various types but also allow the plug in of well known "Patterns" from a
pattern library.
The user could even create "Patterns" that get added to his pattern library
for reuse.

Maybe even include code generation capability from UML .............

Of course it would have to run on Microsoft Platforms as well as Linux.....

Michael C. Garrett
Vice President of Research and Development
Medical Research Laboratories
michaelgarrett@csi.com

Re: Project: FreeOS

[Bill Greene]

Jan Kroken wrote:
> I disagree. Ada is a low level language, and should not be marketed
> as an alternative to Java, VB[0], Python, Perl, Lisp or other high
> level languages, ...
> 
> As a one line summary: The language we're trying to replace is C.

Some would characterize Ada as a "wide-spectrum language."  Certainly
you can program at the lowest level of abstraction in Ada.  This is one
of the uses for which Ada was designed, and it was designed well!

But Ada can also be used as a high-level language.  I certainly don't
see how Java or Visual Basic are high level languages and Ada is not.

IIRC, there was an Ada translator in the early 1980s (from
Carnegie-Mellon?) that generated LISP code as its intermediate
representation.
-- 
William R. Greene                              1100 Perimeter Park Drive
Ganymede Software, Inc.                                        Suite 104
http://www.ganymede.com                       Morrisville, NC  27560 USA
Phone: (919) 469-0997, ext. 280                      Fax: (919) 469-5553

Re: Project: FreeOS

[Brian Rogoff]

On Mon, 17 Jan 2000, Bill Greene wrote:
> Jan Kroken wrote:
> > I disagree. Ada is a low level language, and should not be marketed
> > as an alternative to Java, VB[0], Python, Perl, Lisp or other high
> > level languages, ...
> > 
> > As a one line summary: The language we're trying to replace is C.
> 
> Some would characterize Ada as a "wide-spectrum language."  Certainly
> you can program at the lowest level of abstraction in Ada.  This is one
> of the uses for which Ada was designed, and it was designed well!

Yes, Ada is an excellent wide-spectrum language, because it can do
low-level stuff well. I agree with Jan though that Ada is fairly low
level, especially when compared with Smalltalk, Common Lisp, Prolog,
Python, Icon, etc. 

Robert Dewar wrote an interesting opinion on why Java is a disappointment; 
in a nutshell, Java forces you to pay the price of a VHLL but doesn't give
you a much higher level view of programming than C or C++, except that you
get garbage collection. 

> But Ada can also be used as a high-level language.  I certainly don't
> see how Java or Visual Basic are high level languages and Ada is not.

OK, Java and VB were bad picks as representative HLLs. "Scripting language" 
is the right term for VB. Have you programmed in Scheme or OCaml? It is 
very different from Ada! 

> IIRC, there was an Ada translator in the early 1980s (from
> Carnegie-Mellon?) that generated LISP code as its intermediate
> representation.

A C to Ada translator is possible too, but its existence wouldn't imply
that C is a higher level language than Ada!

-- Brian

Re: Project: FreeOS

[Aidan Skinner]

On Mon, 17 Jan 2000 15:36:14 -0800, Brian Rogoff
<bpr@shell5.ba.best.com> wrote: 

>low-level stuff well. I agree with Jan though that Ada is fairly low
>level, especially when compared with Smalltalk, Common Lisp, Prolog,
>Python, Icon, etc. 

To an extent this can be solved with the judicious use of libraries,
although I'm not entirley sure how one would implement prolog-esque
features in a "nice" manner.

- Aidan
-- 
Little Willy was a chemist, Little Willy is no more,
What he thought was H2O, Was H2SO4.
http://www.skinner.demon.co.uk/aidan/

Re: Project: FreeOS

[Jon S Anthony]

Aidan Skinner wrote:
> 
> On Mon, 17 Jan 2000 15:36:14 -0800, Brian Rogoff
> <bpr@shell5.ba.best.com> wrote:
> 
> >low-level stuff well. I agree with Jan though that Ada is fairly low
> >level, especially when compared with Smalltalk, Common Lisp, Prolog,
> >Python, Icon, etc.

Of these, Common Lisp is probably both the highest level and yet the
most efficient.  Highest because you can basically create other
languages _inside_ it (note the difference from "with it").
 
> To an extent this can be solved with the judicious use of libraries,
> although I'm not entirley sure how one would implement prolog-esque
> features in a "nice" manner.

Actually a "prolog-library" wouldn't be that bad.  Now, try creating
closures, the idea of the reader/writer, and Lisp style macros in a
library.  This stuff is what gives you _really_ high level capabilities,
yet very efficient ones.

Doable?  Yes, but you would basically have to write a huge chunk of CL
implementation to do it.


/Jon

-- 
Jon Anthony
Synquiry Technologies, Ltd. Belmont, MA 02478, 617.484.3383
"Nightmares - Ha!  The way my life's been going lately,
 Who'd notice?"  -- Londo Mollari

Re: Project: FreeOS

[Jon S Anthony]

Brian Rogoff wrote:
> 
> Robert Dewar wrote an interesting opinion on why Java is a disappointment;
> in a nutshell, Java forces you to pay the price of a VHLL but doesn't give
> you a much higher level view of programming than C or C++, except that you
> get garbage collection.

I think that is a pretty good description of one aspect of Java
"failure".  I've tended to say that Java can't make up it's mind as to
whether it is really a VHLL or just another statically typed compiled
language.  In many ways, _as a language_, it ends up being the worst of
all possible worlds.  In practice it definitely has some advantages...

/Jon

-- 
Jon Anthony
Synquiry Technologies, Ltd. Belmont, MA 02478, 617.484.3383
"Nightmares - Ha!  The way my life's been going lately,
 Who'd notice?"  -- Londo Mollari