Ada OS

Ada OS

[Peter Hermann]

Randy Brukardt mentioned the idea of an operating system all in Ada
which he ranked as unrealistic.
As a starting point it is conceivable to have a tool
written in Ada to master all concerns of
all boot records including MBR and partitions, erroneous too,
in a reliable cautious preserving dialog with a (sysman) user.

btw: I am in desperate need of such a beast :-)
 (i.e. you may rank this posting as selfish ;-) 

-- 
--Peter.Hermann@ihr.uni-stuttgart.de        (+49)0711-685-872-44(Fax79)
--Nobelstr.19 Raum 0.030, D-70569 Stuttgart IHR Hoechstleistungsrechnen
--http://www.ihr.uni-stuttgart.de/

Re: Ada OS

[tmoran]

> written in Ada to master all concerns of
> all boot records including MBR and partitions, erroneous too,
> in a reliable cautious preserving dialog with a (sysman) user.
  I have parts of that but they are in Ada 83 to run under DOS.  An
NTFS disk got corrupted and was totally inaccessible under Windows,
but this program was able to find and copy many files.  Direct port IO
and Int 16#13# calls from the DOS program did the trick.

Re: Ada OS

[Peter Hermann]

tmoran@acm.org wrote:
> > written in Ada to master all concerns of
> > all boot records including MBR and partitions, erroneous too,
> > in a reliable cautious preserving dialog with a (sysman) user.
>   I have parts of that but they are in Ada 83 to run under DOS.  An
> NTFS disk got corrupted and was totally inaccessible under Windows,
> but this program was able to find and copy many files.  Direct port IO
> and Int 16#13# calls from the DOS program did the trick.

surely another precious brick in the mosaic for Ada OS aficionados,
at least in terms of know how.

The dependency on DOS or Linux basic tools had to be retracted
more and more until Ada boots itself ... a far goal, of course.

This proverb comes to my mind:

____________
 How do you
  want it?

   Soon
   Good
   Cheap

 Pick two!
____________

Re: Ada OS

[Lucretia]

On Nov 19, 1:14 pm, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
wrote:
> Randy Brukardt mentioned the idea of an operating system all in Ada
> which he ranked as unrealistic.

Where did he say this and what were the reasons?

Thanks,
Luke.

Re: Ada OS

[Peter Hermann]

Lucretia <lucretia9@lycos.co.uk> wrote:
> On Nov 19, 1:14 pm, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
> wrote:
> > Randy Brukardt mentioned the idea of an operating system all in Ada
> > which he ranked as unrealistic.
> 
> Where did he say this and what were the reasons?
> 
> Thanks,
> Luke.

voila:

>From: "Randy Brukardt" <randy@rrsoftware.com>
>Newsgroups: comp.lang.ada
>Subject: Re: ada compiler?
>Date: Fri, 16 Nov 2007 19:37:27 -0600
>able to make an OS mechanism work "right".. (My holy grail would be an
>all-Ada system, but I realize that is not practical in the real world for
>the vast majority of applications - including mine.)

Re: Ada OS

[Lucretia]

On Nov 22, 10:13 am, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
wrote:

> >From: "Randy Brukardt" <ra...@rrsoftware.com>
> >Newsgroups: comp.lang.ada
> >Subject: Re: ada compiler?
> >Date: Fri, 16 Nov 2007 19:37:27 -0600
> >able to make an OS mechanism work "right".. (My holy grail would be an
> >all-Ada system, but I realize that is not practical in the real world for
> >the vast majority of applications - including mine.)

Thought it was this bit. I wonder why he thinks it's not all that
viable.

Luke.

Re: Ada OS

[anon]

The problem is the scope of the project. Most people who hear about 
an "Ada OS" they think it will be similar to "Microsoft Windows" 
or Mac, Linux, Posix, Unix design, that's a big job. The following 
will give you some idea, but know that I am skipping a lot of steps 
between each of the stages.

Stage 0: A plan or purpose for the OS. Microsoft's or the Unix type 
of OS are a general purpose. But you can have specific version such 
as Real-Time OS. This needs to be written in detail before any other 
stages are started. And writing documentation is dull, most people 
stop before finishing this stage. 

        NOTE: A simple design may take a week something like 
              Windows design could take a year. 

        NOTE: Where do you stop using Ada only coding and 
              allow projects in other languages. Or do you. 

Stage 1: create a team of programmers. A simple kernel may require one 
programmer, but a DOS like OS will take a team of 10 to 25 programmers 
or more. Plus, some of these programmers may leave the project before 
the OS is created, so, you have to have others available. Then you have 
to deal with the in fighting. Most will want to create the kernel, but 
how many will want to create the first command line simple text editor 
like 'e' or a full blown 'vi'.

Stage 2: Creating the kernel initially based on Stage 0 design. 
Plus, if the kernel is for multiple classes of cpus (IA_32, IA_32_64, 
IA_64, others) then you will need a kernel for each class. At the 
movement Ada does not contain conditional compiling statements. And 
I hope that it never does.

        NOTE: Now most Ada OS's stop here. That's because the 
              author's reason for the OS is done or the author 
              has no more time for their project.
              But the kernel is not an OS, it just the 
              beginning or core of an OS.


Stage 3: Is the Ada compiler and its libraries for this kernel. 

        NOTE: Here also multiple compilers are needed because 
              Ada does not have conditional compiling. 


STAGE NOTE: Starting with Stage Four the system should be hardware 
independent enough to allow standardize applications. Only need to 
compile source for each cpu class.

Stage 4: OS support applications. Such as hardware configuration and 
simple command shell.  These hardware applications, sets up the 
networks or Time and Date functions, etc. This also includes the File 
Systems support applications such as FDISK and FORMAT.

        Note: At this point the OS should have the functionability 
              of a DOS. With one major application the Ada 
              compiler system. 

        Note: In today environment, DOS in not really useable. 
              So, the OS will need a GUI, a browser for the 
              internet, ect.

Stage 5: GUI and the windows like applications. But this requires the 
kernel has some kind of graphic video driver for each type of video 
card supported. And the compiler having a graphics library packages.  

        Note: Linux uses X-Windows, there are a couple of
              GUIs interfaces out there like KDE or GOME to 
              name two. And these took years to write with each 
              having their own applications. And each project 
              is not finished either.

And the list goes on!

TIME TABLE: Except for the simple kernel which a single student 
programmer might write within a semester. Each Stage may require a 
year or more. So, an Ada only type of OS might be 5 years in the 
future. 

FINAL WORD: In 5 year, how many people will want to use Ada or an 
Ada OS. Plus, the time required for upgrading the system because of 
hardware. So, most programmer just do not want to spend that amount 
of time on a non-paying project that most people will never use. 

FINAL NOTE: Linux started off as a free replacement to the high 
dollar UNIX OS. That gave the initial programmers a reason to join 
and stay with the Linux team. Later came other reasons.

FINAL NOTE: This is not all but it does give you an idea just how 
big this project is.

In <bfc38822-0a92-47fd-ad74-853ea6b66013@v4g2000hsf.googlegroups.com>, Lucretia <lucretia9@lycos.co.uk> writes:
>On Nov 22, 10:13 am, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
>wrote:
>
>> >From: "Randy Brukardt" <ra...@rrsoftware.com>
>> >Newsgroups: comp.lang.ada
>> >Subject: Re: ada compiler?
>> >Date: Fri, 16 Nov 2007 19:37:27 -0600
>> >able to make an OS mechanism work "right".. (My holy grail would be an
>> >all-Ada system, but I realize that is not practical in the real world for
>> >the vast majority of applications - including mine.)
>
>Thought it was this bit. I wonder why he thinks it's not all that
>viable.
>
>Luke.

Re: Ada OS

[Peter Hermann]

It is self-evident that an OS from scratch would be 
a gigantic costly endeavor when written from scratch.
OTOH Linus Torvalds is wise enough to identify Ada
as a useful tool to replace pieces of C-code
bit by bit towards temporary coexistence of both languages.
The biggest advantage is the deep knowledge of
weak points (if any) of the architecture of Linux.

ph

Re: Ada OS

[Maciej Sobczak]

On 23 Lis, 10:46, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
wrote:

> OTOH Linus Torvalds is wise enough to identify Ada
> as a useful tool to replace pieces of C-code

Could you please post a reference to some article where Torvalds says
this?

--
Maciej Sobczak * www.msobczak.com * www.inspirel.com

Re: Ada OS

[Brian May]

>>>>> "Peter" == Peter Hermann <ica2ph@lucky.ihr.uni-stuttgart.de> writes:

    Peter> It is self-evident that an OS from scratch would be a
    Peter> gigantic costly endeavor when written from scratch.  OTOH
    Peter> Linus Torvalds is wise enough to identify Ada as a useful
    Peter> tool to replace pieces of C-code bit by bit towards
    Peter> temporary coexistence of both languages.  The biggest
    Peter> advantage is the deep knowledge of weak points (if any) of
    Peter> the architecture of Linux.

A better approach might be to rewrite modules from The Hurd in Ada, as
in The Hurd most things are run as independent user level
processes. So you wouldn't have to redo everything from scratch.
-- 
Brian May <bam@snoopy.apana.org.au>

Re: Ada OS

[Randy Brukardt]

"Lucretia" <lucretia9@lycos.co.uk> wrote in message
news:bfc38822-0a92-47fd-ad74-853ea6b66013@v4g2000hsf.googlegroups.com...
> On Nov 22, 10:13 am, Peter Hermann <ica...@lucky.ihr.uni-stuttgart.de>
> wrote:
>
> > >From: "Randy Brukardt" <ra...@rrsoftware.com>
> > >Newsgroups: comp.lang.ada
> > >Subject: Re: ada compiler?
> > >Date: Fri, 16 Nov 2007 19:37:27 -0600
> > >able to make an OS mechanism work "right".. (My holy grail would be an
> > >all-Ada system, but I realize that is not practical in the real world
for
> > >the vast majority of applications - including mine.)
>
> Thought it was this bit. I wonder why he thinks it's not all that
> viable.

Anon has it right: it is too large of a project to attempt (absent a
substantial community willing to work on it, and that would require having
real problems that need to be solved). Just running my all-Ada mail and web
servers would require a file system, TCP/IP (sockets) implementation,
bootstrap loader, keyboard and screen drivers, and task manager. (The latter
could be "borrowed" from the Janus/Ada runtime, but the others would need to
be built more-or-less from scratch.)

I once (when I was still in college, so this was a long time ago) created a
mini-OS in Ada (built on top of CP/M device drivers), but I never got much
further than the file system. And I probably had a lot more time and energy
then than I do now....

                      Randy.

Re: Ada OS

[I. Levashew]

Peter Hermann пїЅпїЅпїЅпїЅпїЅ:
> Randy Brukardt mentioned the idea of an operating system all in Ada
> which he ranked as unrealistic.
Well, it isn't. I'm a bit more optimistic about it.

What if we choose a bit another way of accomplishing this non-easy goal? 
What I mean: not to do it from scratch, but reengineer existing system. 
Infect Open Source.

It's about motivation. How much effort one must spend to get any result.
I think the one willing to see something rewritten in Ada should 
concentrate on providing easy way to infect C programs. That is, it 
should be easy to get any open source tarball-distributed utility and 
put a bit of Ada source into it. It should be benefittable to write 
addons in Ada, not in C or C++. Once the society realize that many 
tarballs contain Ada code, Ada may become more common PL than ever. 
Currently we can obtain (for example) Leopard double-layer(!) install 
DVD. Install DVD always contain developer tools. There is Perl, Python, 
Ruby, Tcl/Tk, etc inside. But on the whole double-layer DVD there wasn't 
any GNAT despite being part of main GCC distribution. Ada lock-ins will 
ensure higher availability of GNAT.

There were times when Perl was the only scripting PL widely used in *NIX 
("Swiss Army Chainsaw" of *NIX). Now we can see Perl and Python sharing 
this niche. What if the same thing to some extent is applicable to Ada?

I'd like to start infecting projects that are not developed anymore, but 
still used. That way :
a) original authors shouldn't care about inclusion of Ada in sources
b) Ada programs wouldn't be rare aliens.

ShakesPeer, a Direct Connect client for Mac OS X have all the repertoire 
of bugs specific to (Objective)C. Will it be rewritten in Ada, it'll 
become better, thus raising reputation of Ada.
(For several reasons writing something in Ada is insanely difficult on 
Mac OS X, so I can't even get started doing it.)

If there was many successful experiences of C(++) project converted to 
Ada, idea of rewritting C(++)->Ada would be more popular. Thus we could 
have much more coder hands willing to rewrite.

It's just a wild idea.

Re: Ada OS

[I. Levashew]

Peter Hermann пїЅпїЅпїЅпїЅпїЅ:
> Randy Brukardt mentioned the idea of an operating system all in Ada
> which he ranked as unrealistic.
Well, it isn't. I'm a bit more optimistic about it.

What if we choose a bit another way of accomplishing this non-easy goal? 
What I mean: not to do it from scratch, but reengineer existing system. 
Infect Open Source.

It's about motivation. How much effort one must spend to get any result.
I think the one willing to see something rewritten in Ada should 
concentrate on providing easy way to infect C programs. That is, it 
should be easy to get any open source tarball-distributed utility and 
put a bit of Ada source into it. It should be benefittable to write 
addons in Ada, not in C or C++. Once the society realize that many 
tarballs contain Ada code, Ada may become more common PL than ever. 
Currently we can obtain (for example) Leopard double-layer(!) install 
DVD. Install DVD always contain developer tools. There is Perl, Python, 
Ruby, Tcl/Tk, etc inside. But on the whole double-layer DVD there wasn't 
any GNAT despite being part of main GCC distribution. Ada lock-ins will 
ensure higher availability of GNAT.

There were times when Perl was the only scripting PL widely used in *NIX 
("Swiss Army Chainsaw" of *NIX). Now we can see Perl and Python sharing 
this niche. What if the same thing to some extent is applicable to Ada?

I'd like to start infecting projects that are not developed anymore, but 
still used. That way :
a) original authors shouldn't care about inclusion of Ada in sources
b) Ada programs wouldn't be rare aliens.

ShakesPeer, a Direct Connect client for Mac OS X have all the repertoire 
of bugs specific to (Objective)C. Will it be rewritten in Ada, it'll 
become better, thus raising reputation of Ada.
(For several reasons writing something in Ada is insanely difficult on 
Mac OS X, so I can't even get started doing it.)

If there was many successful experiences of C(++) project converted to 
Ada, idea of rewritting C(++)->Ada would be more popular. Thus we could 
have much more coder hands willing to rewrite.

It's just a wild idea.

Re: Ada OS

[anon]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4528 bytes --]

People might be interested in Ada OS but most do not want to spend the 
time to build an Ada OS or convert an existing dead one. Now in saying 
this you can see this in the following project. Just expand this concept 
into the world of an OS. You can see that an Ada OS will never get done 
other than the projects that are in the works today. And they kind of limit 
the Ada OS to the kernel only based on a Posix interface.

The example is the "c2Ada" project. Which is now hosted by sourceforge. 
Simonjwright with help from jcream unzip the last 2007 Linux based file 
and placed the source files into SVN tree area. Why?

        If they were into Ada like they suggest they are. They would have:

                1. Placed the original zip file into the download 
                   area stating that this was the original code and 
                   is stored for archiving purposes only.

                2. Documenting the following steps ( 3 .. 5 ).

                3. Compile the program.

                4. Using the c2Ada to convert the original "c2ada" 
                   C code to a standard Ada 95 code.

                5. Editing the 20% of the code that the program 
                   bypasses.

                6. Placing this new "ADA ONLY" version into the 
                   SVN tree as well as an archived version in the 
                   downloading area. 

                7. Insert the documented transition (from step 2) 
                   into the web site.

        Time Table: A couple of weekends. Basically while watching 
                    TV on Saturday and Sunday afternoon. One weekend 
                    for converting the program and a second for 
                    correcting the documentation and uploading.

        Then later:
                8. Update the program for Ada 2005.


NOTE: c2Ada -- The original converts C to Ada 95 only.


        In doing this they would prove the program works and how hard it 
is to use (Documentation).

        But has this happen. No, Why not? That's the question! Both 
"c2ada" project admins post in this newsgroup along with the other two 
admins, so you could ask them.

        Their answers might give you more of the true reason why a Ada 
OS is unrealistic.


In <fkd741$r2f$1@registered.motzarella.org>, "I. Levashew" <octagram@bluebottle.com> writes:
>Peter Hermann �����:
>> Randy Brukardt mentioned the idea of an operating system all in Ada
>> which he ranked as unrealistic.
>Well, it isn't. I'm a bit more optimistic about it.
>
>What if we choose a bit another way of accomplishing this non-easy goal? 
>What I mean: not to do it from scratch, but reengineer existing system. 
>Infect Open Source.
>
>It's about motivation. How much effort one must spend to get any result.
>I think the one willing to see something rewritten in Ada should 
>concentrate on providing easy way to infect C programs. That is, it 
>should be easy to get any open source tarball-distributed utility and 
>put a bit of Ada source into it. It should be benefittable to write 
>addons in Ada, not in C or C++. Once the society realize that many 
>tarballs contain Ada code, Ada may become more common PL than ever. 
>Currently we can obtain (for example) Leopard double-layer(!) install 
>DVD. Install DVD always contain developer tools. There is Perl, Python, 
>Ruby, Tcl/Tk, etc inside. But on the whole double-layer DVD there wasn't 
>any GNAT despite being part of main GCC distribution. Ada lock-ins will 
>ensure higher availability of GNAT.
>
>There were times when Perl was the only scripting PL widely used in *NIX 
>("Swiss Army Chainsaw" of *NIX). Now we can see Perl and Python sharing 
>this niche. What if the same thing to some extent is applicable to Ada?
>
>I'd like to start infecting projects that are not developed anymore, but 
>still used. That way :
>a) original authors shouldn't care about inclusion of Ada in sources
>b) Ada programs wouldn't be rare aliens.
>
>ShakesPeer, a Direct Connect client for Mac OS X have all the repertoire 
>of bugs specific to (Objective)C. Will it be rewritten in Ada, it'll 
>become better, thus raising reputation of Ada.
>(For several reasons writing something in Ada is insanely difficult on 
>Mac OS X, so I can't even get started doing it.)
>
>If there was many successful experiences of C(++) project converted to 
>Ada, idea of rewritting C(++)->Ada would be more popular. Thus we could 
>have much more coder hands willing to rewrite.
>
>It's just a wild idea.

Re: Ada OS

[Paul]

Funny that while this was written, there is a thread on OpenBSD's 
mailing list about Ada.

What I would like to know though, are there many Ada developers who use 
OpenBSD?

OpenBSD values security, correctness and standardization, all of which I 
see in Ada.

Paul

OpenBSD and Ada (was: Re: Ada OS)

[Tero Koskinen]

On Mon, 24 Dec 2007 05:41:54 GMT Paul wrote:
> What I would like to know though, are there many Ada developers who use 
> OpenBSD?

I use almost(1) exclusively OpenBSD and Ada and I think I have seen
something like 4 other people using OpenBSD and Ada more or less
actively.

Hardest part of that combination is to get fully featured GNAT built
and running. GCC in the OpenBSD ports tree has Ada support but lacks
tasking and some other features. It can be used for bootstrapping, but
for better support you need to use unofficial ports (2,3).


(1) My laptop dual boots XP (with ObjectAda Special Ed.) and OpenBSD (GNAT).
(2) http://marc.info/?l=openbsd-ports&m=117679065130354&w=2
(3) http://marc.info/?l=openbsd-ports&m=119851255408925&w=2
-- 
Tero Koskinen - http://iki.fi/tero.koskinen/

Re: Ada OS

[I. Levashew]

Paul wrote:

> OpenBSD values security, correctness and standardization, all of which I 
> see in Ada.

Well, security is currently something fashionable. Nobody wants to be 
the one not concious about security. But Ada isn't being mentioned more 
often. I had impression that it's a good show business to be 
security-concious. Cyclone, Fortify Source, Singularity. Cyclone 
pretends to fill "an empty niche". Fortify Source makes money on finding 
security issues in a source code, but without any reference to Ada. And 
Singularity pretends to be "absolutely secure" OS, mostly in modified 
.NET with tiny pieces in ... guess what? "verified C++ source code". 
Microsoft .NET implementation had critical vulnerabilities like any 
other OS component. How can one believe that "this time they'll maid 
something real".


In these circumstances one argument remains unbeaten: programs must not 
do bad things not because they were not allowed to do them. Most people 
forget one simple thing: virtually no any developer ever wrote

if Shell_Codes.Argument_Is_Shell_Code(Argument) then
    Shell_Codes.Run_Shell_Code(Argument);
end if;

in their programs. Security measures such as SELinux, chroot, 
Capabilities, etc. can only be secondary layers of prevention, not the 
primary one.

Returning to the original topic, there are many independent decisions to 
be maid when designing OS distro. OK, let's write everything we can in 
Ada. What's next? Currently there are many identified problems in modern 
OSes. One of them is extremelly high complexity of OS. Good example is 
Windows. After using it for a long time you are losing control over it. 
Different programs can alter some OS files, registry keys. Three years 
past installation it can be difficult to remember all the changes maid 
to the system. The only good way to backup such a system -- to dump the 
whole system partition. Next generation OS must be managed in a 
functional way. One property of a function is exploited here: it's 
deterministic. It must produce the same results wherever and whenever it 
was invoked. It must be simple to build up any program for any 
architecture with any configure options. Building crosscompilers, 
managing compilation process must be done via underlying subsystem. Nix 
package management is far from being perfect. But it is aimed right. In 
Nix OS functional way of doing things is also applied to system 
configuration, not just package management. If one could express the 
whole system with set of expressions in a functional language, it'll be 
easy to backup the whole system, manage multiple identical(or not so 
identical) workstations and servers and add a new ones. Ideally, without 
long hours of installation process. Thouthands of files present in 
modern OSes would be just cache of functional expression computation 
results. And provided that functions are really deterministic one can 
allow workstations to exchange their cached expressions.

SELinux is about managing rights in a functional way.

As you can see, both Nix and SELinux have nothing to do with Ada. These 
are independent design choices. And there are many other independent 
design decisions to be maid. Fortunatelly, there was enough time to 
identify common problems.

IIUC, Nix and SELinux are not currently integrated. I think their 
combination can be good step towards next-generation OS. Like SEBSD, 
SEDarwin are SELinux way of managing rights applied to BSD and Darwin, 
let's call it Senix. Sounds good, no?

God in the details. Each detail matters. That's why next-generation OS 
must be written in Ada, but it's neither the least nor the most 
meaningful detail.

IMHO.

Re: Ada OS

[Dmitry A. Kazakov]

On Thu, 27 Dec 2007 02:56:04 +0600, I. Levashew wrote:

> Paul wrote:
> 
>> OpenBSD values security, correctness and standardization, all of which I 
>> see in Ada.
> 
> Well, security is currently something fashionable. Nobody wants to be 
> the one not concious about security. But Ada isn't being mentioned more 
> often. I had impression that it's a good show business to be 
> security-concious.

Under "security" a marketing guy understands making secure something which
is not. They sell solutions for problems. The software which is secure by
its nature has no market, obviously. You can't sell a medical ventilator to
people who can breathe. First they must be ill, badly ill...

> in their programs. Security measures such as SELinux, chroot, 
> Capabilities, etc. can only be secondary layers of prevention, not the 
> primary one.

Right, a lid to seal the compost-bin...

> God in the details. Each detail matters. That's why next-generation OS 
> must be written in Ada, but it's neither the least nor the most 
> meaningful detail.

I think that Ada still lacks some important features. Concerning security
and OS design the problem with Ada is that it has a trusted model of
interaction between components. You can do a lot of unchecked stuff. The
model of access rights is very primitive (visible - private).

In short, there will be needed a level of OS API *outside* the language to
protect objects and methods of. That wouldn't be a truly modern OS, as Unix
was when it started to use a higher-level language in its design. To be
modern now means that the very language should be able to deliver
protection (and some other things, like distribution, persistence,
heterogeneity).

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de

Re: Ada OS

[Georg Bauhaus]

On Thu, 2007-12-27 at 02:56 +0600, I. Levashew wrote:
>  Next generation OS must be managed in a 
> functional way. One property of a function is exploited here: it's 
> deterministic.

You mean, the mathematical notion of a function, the one that
---absent monads or a similar feature---asks its programmers
to not consider the operating system issues: time and space?

>  It must produce the same results wherever and whenever it 
> was invoked.

How would a function like that implement the random numbers device?

Re: Ada OS

[Dmitry A. Kazakov]

On Wed, 26 Dec 2007 23:48:12 +0100, Georg Bauhaus wrote:

> How would a function like that implement the random numbers device?

(:-))

As it does in mathematics, by returning the random variable (whole
distribution) instead of its individual values (realizations). BTW, in
earlier days one indeed used this way. Remember printed tables of random
numbers?

http://en.wikipedia.org/wiki/Random_number_table

Theoretically you can always get rid of identity and so mutability, by
turning to greater cardinalities. This applies even to the real-time clock.
Just create a universe each 1�s and here you are... (:-))

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de

Re: Ada OS

[I. Levashew]

Georg Bauhaus wrote:
> asks its programmers
> to not consider the operating system issues: time and space?
No, not programmers. Nix is used for compilation and configuration. So 
the one who cares about these processes are maintainers and administrators.

> How would a function like that implement the random numbers device?
It's not a task being solved here.

Ada lacks support greatly (was: OpenBSD and Ada)

[Rico Secada]

On Wed, 26 Dec 2007 21:49:15 +0200
Tero Koskinen <tero.koskinen@iki.fi> wrote:

> On Mon, 24 Dec 2007 05:41:54 GMT Paul wrote:
> > What I would like to know though, are there many Ada developers who
> > use OpenBSD?
> 
> I use almost(1) exclusively OpenBSD and Ada and I think I have seen
> something like 4 other people using OpenBSD and Ada more or less
> actively.
> 
> Hardest part of that combination is to get fully featured GNAT built
> and running. GCC in the OpenBSD ports tree has Ada support but lacks
> tasking and some other features. It can be used for bootstrapping, but
> for better support you need to use unofficial ports (2,3).

I think that is the main problem with Ada nowadays. It lacks proper
support on multiple platforms. Unless something is going to be done
about this, it will never reach a wider usage in the Open Source
community.
 
> 
> (1) My laptop dual boots XP (with ObjectAda Special Ed.) and OpenBSD
> (GNAT).
> (2) http://marc.info/?l=openbsd-ports&m=117679065130354&w=2
> (3) http://marc.info/?l=openbsd-ports&m=119851255408925&w=2
> -- 
> Tero Koskinen - http://iki.fi/tero.koskinen/

Re: OpenBSD and Ada

[Samuel Tardieu]

>>>>> "Tero" == Tero Koskinen <tero.koskinen@iki.fi> writes:

Tero> Hardest part of that combination is to get fully featured GNAT
Tero> built and running. GCC in the OpenBSD ports tree has Ada support
Tero> but lacks tasking and some other features. It can be used for
Tero> bootstrapping, but for better support you need to use unofficial
Tero> ports.

Have you checked whether those patches have been included in the
development version of GCC? If not, you should submit them using
Bugzilla (http://gcc.gnu.org/bugzilla) so that they get included
before GCC 4.3.0 is released.

  Sam
-- 
Samuel Tardieu -- sam@rfc1149.net -- http://www.rfc1149.net/

Re: Ada lacks support greatly

[Florian Weimer]

* Rico Secada:

> I think that is the main problem with Ada nowadays. It lacks proper
> support on multiple platforms. Unless something is going to be done
> about this, it will never reach a wider usage in the Open Source
> community.

I don't think portability is a major concern for open source projects.
If it runs on GNU/Linux (i386 and maybe amd64), it's usually good
enough.

It's generally very hard to tell what makes a project successful.  I
used to think that the cathedral-like development style favored by
AdaCore was part of the problem, but there still are some similarly
structured projects which have got quite a following.

Re: Ada OS

[Thomas Preymesser]

On Dec 21 2007, 12:29 pm, a...@anon.org (anon) wrote:
>
>                 4. Using the c2Ada to convert the original "c2ada"
>                    C code to a standardAda95 code.

c2ada cannot be used because it seems to be broken (Segmentation fault
under Ubuntu Linux 7.10)

Re: Ada OS

[anon]

That's one problem that I was also hinting at!

If a person manages an Ada project they should insure that each version 
of the project works. And use documentation to info the user how to 
make it work and states any known problems. Instead of just finding an 
old project and dumping it into a CVS/SVN server.

"Simon Wright" should of archived the last Linux "c2ada" version 
instead of posting the zip file contents, because it is easy to alter a 
file and then post that altered file which can cause the project to 
become non-functional. 

Plus, just dumping a broken project without the original package cause 
the newbees to HATE Ada, because they can not use a project.  And all 
projects that are written in other languages should be only stored as 
archived and re-written in Ada before posting the project. Because any 
project that is written in standard Ada only language can be compiled 
and executed by all. But if more than two languages are use most people 
will just skip the project and move on.

        Now, for the c2Ada, you should ask "Simon Wright" why he did 
not insure that the c2ada project is full functional before he posted it. Or 
have full documentation of how to build it.


In <1c50901a-7e4c-4495-ac1c-000ffc054a86@21g2000hsj.googlegroups.com>, Thomas Preymesser <thopre@gmail.com> writes:
>On Dec 21 2007, 12:29 pm, a...@anon.org (anon) wrote:
>>
>>                 4. Using the c2Ada to convert the original "c2ada"
>>                    C code to a standardAda95 code.
>
>c2ada cannot be used because it seems to be broken (Segmentation fault
>under Ubuntu Linux 7.10)
>

Re: Ada OS

[Georg Bauhaus]

anon instructed:

>         Now, for the c2Ada, you should ask "Simon Wright" why he did 
> not insure that the c2ada project is full functional before he posted it.

I can't say whether the air is thin up there but I know
a full language translator for C (with CPP) is not trivial
and AFAIK has always had flaws.

What do you know about c2ada?

Will you be doing anyhting about it?

Re: Ada OS

[Simon Wright]

anon@anon.org (anon) writes:

> That's one problem that I was also hinting at!
>
> If a person manages an Ada project they should insure that each
> version of the project works. And use documentation to info the user
> how to make it work and states any known problems. Instead of just
> finding an old project and dumping it into a CVS/SVN server.

There isn't a released C2Ada on sf.net.

> "Simon Wright" should of archived the last Linux "c2ada" version 
> instead of posting the zip file contents, because it is easy to alter a 
> file and then post that altered file which can cause the project to 
> become non-functional. 

"should *have*", please.

And it wasn't me, though it might have been. I'm a co-developer (& as
it happens admin) on the project, special interest in making it
portable (to Mac OS X, specifically).

Not sure where the last c2ada archive came from, only glad jcreem took
the initiative to make it visible so that interested parties can
contribute.

> Plus, just dumping a broken project without the original package cause 
> the newbees to HATE Ada, because they can not use a project.  And all 
> projects that are written in other languages should be only stored as 
> archived and re-written in Ada before posting the project. Because any 
> project that is written in standard Ada only language can be compiled 
> and executed by all. But if more than two languages are use most people 
> will just skip the project and move on.

If I was rewriting this project I would might not use Ada, but I would
certainly insist on a proper set of design documentation and code
standards. It's very difficult to gather what most of the subprograms
are for, what the pre- and post-conditions are etc, which makes it
very hard to change. Part of that's down to C and C habits, of course.

I think I would also abandon any attempt to make a generalised
translator, instead let's just (!) go for an interface generator.

>         Now, for the c2Ada, you should ask "Simon Wright" why he did
> not insure that the c2ada project is full functional before he
> posted it. Or have full documentation of how to build it.

You don't seem to understand what SourceForge is for!

Re: Ada OS

[Simon Wright]

Thomas Preymesser <thopre@gmail.com> writes:

> On Dec 21 2007, 12:29 pm, a...@anon.org (anon) wrote:
>>
>>                 4. Using the c2Ada to convert the original "c2ada"
>>                    C code to a standardAda95 code.
>
> c2ada cannot be used because it seems to be broken (Segmentation fault
> under Ubuntu Linux 7.10)

Not just under Linux! Your report (thanks) is in fact a duplicate of
bug 1797851, and the same workround applies: PYTHONPATH must point to
the directory containing Symbol.py and the other Python files supplied
with c2ada.

There's a function which looks as though it ought to sort this but
since nothing ever calls it I don't see how!

I have yet to grasp what Python is in there for :-(

Re: Ada OS

[Simon Wright]

Simon Wright <simon.j.wright@mac.com> writes:

> Thomas Preymesser <thopre@gmail.com> writes:
>
>> On Dec 21 2007, 12:29 pm, a...@anon.org (anon) wrote:
>>>
>>>                 4. Using the c2Ada to convert the original "c2ada"
>>>                    C code to a standardAda95 code.
>>
>> c2ada cannot be used because it seems to be broken (Segmentation fault
>> under Ubuntu Linux 7.10)
>
> Not just under Linux! Your report (thanks) is in fact a duplicate of
> bug 1797851, and the same workround applies: PYTHONPATH must point to
> the directory containing Symbol.py and the other Python files supplied
> with c2ada.
>
> There's a function which looks as though it ought to sort this but
> since nothing ever calls it I don't see how!

Now fixed in SVN.

> I have yet to grasp what Python is in there for :-(

But that one isn't fixed ..

Re: Ada OS

[anon]

Really, you say: No release!

http://c2ada.wiki.sourceforge.net/

This web page ("wiki") should contain the history and status 
of the project. Location of the main documentation and links 
to the files.

No files in the Download section but there are in the SVN area!

http://c2ada.svn.sourceforge.net/viewvc/c2ada/

"jcreem" may be the one that inserted the files but as primary 
manager, "Simon Wright" is responsible for the project.

A unmodified version of the file c2ada_linux_august_2007.zip 
should be in the download area as an archived version. 

You did not see my other post on Ada OS where I stated that 
to proved the usage and to documentation the operation of 
the "c2ada" the program should be used on itself. No rewrite!
Just using the program on itself to prove that it works and 
then posting this new version.

That is:

                1. Placed the original zip file into the download 
                   area stating that this was the original code and 
                   is stored for archiving purposes only.

                2. Documenting the following steps ( 3 .. 5 ).

                3. Compile the program.

                4. Using the c2Ada to convert the original "c2ada" 
                   C code to a standard Ada 95 code.

                5. Editing the 20% of the code that the program 
                   bypasses.

                6. Placing this new "ADA ONLY" version into the 
                   SVN tree as well as an archived version in the 
                   downloading area. 

                7. Insert the documented transition (from step 2) 
                   into the web site.

        Time Table: A couple of weekends. Basically while watching 
                    TV on Saturday and Sunday afternoon. One weekend 
                    for converting the program and a second for 
                    correcting the documentation and uploading.

        Then later:
                8. Update the program for Ada 2005. 


Of course, you could leave step 8 for others but the documentation 
on the main web page should say this. And for editing the program 
could be done during commercials. 

As for SourceForce.Net, it is for open source projects, not for a set 
of broken projects that hurt the OSI because they do not work. It is 
one thing, if the project becomes outdated, but projects like "c2ada" 
do not become outdated. Well maybe if "C" becomes outdated, but that 
is not going to happen any time soon.

And I would NEVER create an empty (more than a day or two) project or 
have a project that is broken, because that hurts the OSI. But if you 
look (search SourceForce) there are 100s of abandon projects at 
SourceForce that is hurting SourceForce as well as the OSI.


In <m2abngf30h.fsf@mac.com>, Simon Wright <simon.j.wright@mac.com> writes:
>anon@anon.org (anon) writes:
>
>> That's one problem that I was also hinting at!
>>
>> If a person manages an Ada project they should insure that each
>> version of the project works. And use documentation to info the user
>> how to make it work and states any known problems. Instead of just
>> finding an old project and dumping it into a CVS/SVN server.
>
>There isn't a released C2Ada on sf.net.
>
>> "Simon Wright" should of archived the last Linux "c2ada" version 
>> instead of posting the zip file contents, because it is easy to alter a 
>> file and then post that altered file which can cause the project to 
>> become non-functional. 
>
>"should *have*", please.
>
>And it wasn't me, though it might have been. I'm a co-developer (& as
>it happens admin) on the project, special interest in making it
>portable (to Mac OS X, specifically).
>
>Not sure where the last c2ada archive came from, only glad jcreem took
>the initiative to make it visible so that interested parties can
>contribute.
>
>> Plus, just dumping a broken project without the original package cause 
>> the newbees to HATE Ada, because they can not use a project.  And all 
>> projects that are written in other languages should be only stored as 
>> archived and re-written in Ada before posting the project. Because any 
>> project that is written in standard Ada only language can be compiled 
>> and executed by all. But if more than two languages are use most people 
>> will just skip the project and move on.
>
>If I was rewriting this project I would might not use Ada, but I would
>certainly insist on a proper set of design documentation and code
>standards. It's very difficult to gather what most of the subprograms
>are for, what the pre- and post-conditions are etc, which makes it
>very hard to change. Part of that's down to C and C habits, of course.
>
>I think I would also abandon any attempt to make a generalised
>translator, instead let's just (!) go for an interface generator.
>
>>         Now, for the c2Ada, you should ask "Simon Wright" why he did
>> not insure that the c2ada project is full functional before he
>> posted it. Or have full documentation of how to build it.
>
>You don't seem to understand what SourceForge is for!

Re: Ada OS

[Jeffrey Creem]

anon wrote:
> Really, you say: No release!
> 
> http://c2ada.wiki.sourceforge.net/
> 
> This web page ("wiki") should contain the history and status 
> of the project. Location of the main documentation and links 
> to the files.
> 
> No files in the Download section but there are in the SVN area!
> 
> http://c2ada.svn.sourceforge.net/viewvc/c2ada/
> 
> "jcreem" may be the one that inserted the files but as primary 
> manager, "Simon Wright" is responsible for the project.
> 
> A unmodified version of the file c2ada_linux_august_2007.zip 
> should be in the download area as an archived version. 
> 
> You did not see my other post on Ada OS where I stated that 
> to proved the usage and to documentation the operation of 
> the "c2ada" the program should be used on itself. No rewrite!
> Just using the program on itself to prove that it works and 
> then posting this new version.
> 
> That is:
> 
>                 1. Placed the original zip file into the download 
>                    area stating that this was the original code and 
>                    is stored for archiving purposes only.
> 
>                 2. Documenting the following steps ( 3 .. 5 ).
> 
>                 3. Compile the program.
> 
>                 4. Using the c2Ada to convert the original "c2ada" 
>                    C code to a standard Ada 95 code.
> 
>                 5. Editing the 20% of the code that the program 
>                    bypasses.
> 
>                 6. Placing this new "ADA ONLY" version into the 
>                    SVN tree as well as an archived version in the 
>                    downloading area. 
> 
>                 7. Insert the documented transition (from step 2) 
>                    into the web site.
> 
>         Time Table: A couple of weekends. Basically while watching 
>                     TV on Saturday and Sunday afternoon. One weekend 
>                     for converting the program and a second for 
>                     correcting the documentation and uploading.
> 
>         Then later:
>                 8. Update the program for Ada 2005. 
> 

Sounds great. Please make the edits and contribute the updates. That is 
the whole idea of going to SF and having a wiki is that instead of 
complaining about what could be done you can do it.

Re: Ada OS

[Ole-Hjalmar Kristensen]

>>>>> "SW" == Simon Wright <simon.j.wright@mac.com> writes:

    SW> I have yet to grasp what Python is in there for :-(

Last time I looked, it seemed to me that someone used it to do some
rudimentary text analysis. Nothing that could not have been done in C.

-- 
   C++: The power, elegance and simplicity of a hand grenade.

Re: Ada OS

[anon]

That's your job!  You or "Simon Wright" have the orginal 
c2ada_linux_august_2007.zip, I do not.

Now, after 3+ months you should have the complete "c2ada" working. And 
using itself on itself along with the 20% of editing that the program 
requires (document states). You should by now have the Ada-only version 
that you could post.  So why not post that one.


In <c4hd55-is1.ln1@newserver.thecreems.com>, Jeffrey Creem <jeff@thecreems.com> writes:
>anon wrote:
>> Really, you say: No release!
>> 
>> http://c2ada.wiki.sourceforge.net/
>> 
>> This web page ("wiki") should contain the history and status 
>> of the project. Location of the main documentation and links 
>> to the files.
>> 
>> No files in the Download section but there are in the SVN area!
>> 
>> http://c2ada.svn.sourceforge.net/viewvc/c2ada/
>> 
>> "jcreem" may be the one that inserted the files but as primary 
>> manager, "Simon Wright" is responsible for the project.
>> 
>> A unmodified version of the file c2ada_linux_august_2007.zip 
>> should be in the download area as an archived version. 
>> 
>> You did not see my other post on Ada OS where I stated that 
>> to proved the usage and to documentation the operation of 
>> the "c2ada" the program should be used on itself. No rewrite!
>> Just using the program on itself to prove that it works and 
>> then posting this new version.
>> 
>> That is:
>> 
>>                 1. Placed the original zip file into the download 
>>                    area stating that this was the original code and 
>>                    is stored for archiving purposes only.
>> 
>>                 2. Documenting the following steps ( 3 .. 5 ).
>> 
>>                 3. Compile the program.
>> 
>>                 4. Using the c2Ada to convert the original "c2ada" 
>>                    C code to a standard Ada 95 code.
>> 
>>                 5. Editing the 20% of the code that the program 
>>                    bypasses.
>> 
>>                 6. Placing this new "ADA ONLY" version into the 
>>                    SVN tree as well as an archived version in the 
>>                    downloading area. 
>> 
>>                 7. Insert the documented transition (from step 2) 
>>                    into the web site.
>> 
>>         Time Table: A couple of weekends. Basically while watching 
>>                     TV on Saturday and Sunday afternoon. One weekend 
>>                     for converting the program and a second for 
>>                     correcting the documentation and uploading.
>> 
>>         Then later:
>>                 8. Update the program for Ada 2005. 
>> 
>
>Sounds great. Please make the edits and contribute the updates. That is 
>the whole idea of going to SF and having a wiki is that instead of 
>complaining about what could be done you can do it.

Re: Ada OS

[Georg Bauhaus]

On Wed, 2008-01-09 at 19:32 +0000, anon wrote:

> Now, after 3+ months you should have the complete "c2ada" working.

Uhm, is there a secret that tells how to be much more capable
than two teams at Rational (cbind) and then Intermetrics
(C2Ada) have been? Would you mind sharing it?

I'm impressed with how easy it is for some to assess
the c2ada source state and the quickly outline the
corresponding HR requirements.

What kind of job does one get with this qualification?

Re: Ada OS

[Simon Wright]

anon@anon.org (anon) writes:

> Really, you say: No release!
>
> http://c2ada.wiki.sourceforge.net/
>
> This web page ("wiki") should contain the history and status 
> of the project. Location of the main documentation and links 
> to the files.
>
> No files in the Download section but there are in the SVN area!
>
> http://c2ada.svn.sourceforge.net/viewvc/c2ada/

If there is nothing in the download section then there is no release.

> "jcreem" may be the one that inserted the files but as primary 
> manager, "Simon Wright" is responsible for the project.

Why am I in quotes? Do you think that's not my name?

Where do you get the notion that I'm primary manager? (if nominated I
will not run, ...).

> A unmodified version of the file c2ada_linux_august_2007.zip 
> should be in the download area as an archived version. 

Not a completely unreasonable point, I have to admit rather against my
will.

Actually, I'd like the one before as well .. if possible the whole
history!

> You did not see my other post on Ada OS where I stated that 
> to proved the usage and to documentation the operation of 
> the "c2ada" the program should be used on itself. No rewrite!
> Just using the program on itself to prove that it works and 
> then posting this new version.

I don't normally read your posts, cos you're in my kill file.

> That is:
>
>                 1. Placed the original zip file into the download 
>                    area stating that this was the original code and 
>                    is stored for archiving purposes only.
>
>                 2. Documenting the following steps ( 3 .. 5 ).
>
>                 3. Compile the program.
>
>                 4. Using the c2Ada to convert the original "c2ada" 
>                    C code to a standard Ada 95 code.

c2ada is nowhere near capable of this.

Personally I wouldn't want to *convert* a whole C program to Ada,
though I've no objection at all to using an interface converter (such
as I took c2ada to be) to front-end an existing implementation.

>                 5. Editing the 20% of the code that the program 
>                    bypasses.
>
>                 6. Placing this new "ADA ONLY" version into the 
>                    SVN tree as well as an archived version in the 
>                    downloading area. 
>
>                 7. Insert the documented transition (from step 2) 
>                    into the web site.
>
>         Time Table: A couple of weekends. Basically while watching 
>                     TV on Saturday and Sunday afternoon. One weekend 
>                     for converting the program and a second for 
>                     correcting the documentation and uploading.

Hello, Planet Earth calling anon, it took that long to get the thing
to compile on Mac OS X (mainly, convert regexp.h usage to regex.h).

>         Then later:
>                 8. Update the program for Ada 2005. 
>
> Of course, you could leave step 8 for others but the documentation 
> on the main web page should say this. And for editing the program 
> could be done during commercials.

Personally I prefer to get up and make tea/stretch legs.

> As for SourceForce.Net, it is for open source projects, not for a set 
> of broken projects that hurt the OSI because they do not work. It is 
> one thing, if the project becomes outdated, but projects like "c2ada" 
> do not become outdated. Well maybe if "C" becomes outdated, but that 
> is not going to happen any time soon.

I don't think c2ada would be very interesting if there was no more
Ada.

I don't think that c2ada is broken exactly, it just doesn't work
yet. Why it doesn't work when (presumably) it used to -- probably
before someone tried to make it translate function bodies as well as
specs -- is a different matter.