Is Ada preferred over C/C++ for the realtime domain?

Is Ada preferred over C/C++ for the realtime domain?

[Peter Hermann]

[ Article reposted from comp.realtime ]
[ Author was Stefanos Togkoulidis <stog@intranet.gr> ]
[ Posted on Wed, 17 Apr 2002 09:42:25 +0300 ]

Hello there,
I myself, came recently across with Ada as a language proposed for realtime
and embedded systems design. I'd like to ask if nowadays Ada is preferred
for the realtime domain, or C (and C++) is still the language of choice?
Thank you.

Stefanos Togoulidis




-- 
--Peter Hermann(49)0711-685-3611 fax3758 ica2ph@csv.ica.uni-stuttgart.de
--Pfaffenwaldring 27 Raum 114, D-70569 Stuttgart Uni Computeranwendungen
--http://www.csv.ica.uni-stuttgart.de/homes/ph/
--Team Ada: "C'mon people let the world begin" (Paul McCartney)

Re: Is Ada preferred over C/C++ for the realtime domain?

[Stephen Leake]

Peter Hermann <ica2ph@iris16.csv.ica.uni-stuttgart.de> writes:

> I myself, came recently across with Ada as a language proposed for realtime
> and embedded systems design. I'd like to ask if nowadays Ada is preferred
> for the realtime domain, or C (and C++) is still the language of choice?

I have _always_ prefered Ada for realtime; I learned it before C.

I'm not clear who else you are talking about :).

-- 
-- Stephe

Re: Is Ada preferred over C/C++ for the realtime domain?

[Peter Hermann]

Stephen Leake <stephen.a.leake.1@gsfc.nasa.gov> wrote:
> Peter Hermann <ica2ph@iris16.csv.ica.uni-stuttgart.de> writes:
[wrong!]

Stephe, for the record: I did not write the original question.
I simply crossposted from comp.realtime for awareness to AdaTeamers.
martin.m.dowie posted there a good reply.

-- 
--Peter Hermann(49)0711-685-3611 fax3758 ica2ph@csv.ica.uni-stuttgart.de
--Pfaffenwaldring 27 Raum 114, D-70569 Stuttgart Uni Computeranwendungen
--http://www.csv.ica.uni-stuttgart.de/homes/ph/
--Team Ada: "C'mon people let the world begin" (Paul McCartney)

Re: Is Ada preferred over C/C++ for the realtime domain?

[Ted Dennison]

Peter Hermann <ica2ph@iris16.csv.ica.uni-stuttgart.de> wrote in message news:<a9m13p$f0i$1@news.uni-stuttgart.de>...
> Stephen Leake <stephen.a.leake.1@gsfc.nasa.gov> wrote:
> > Peter Hermann <ica2ph@iris16.csv.ica.uni-stuttgart.de> writes:
> [wrong!]
> 
> Stephe, for the record: I did not write the original question.
> I simply crossposted from comp.realtime for awareness to AdaTeamers.
> martin.m.dowie posted there a good reply.

What I've seen work pretty well in the past is just quoting the
relavant parts of the initial message, adding something short but
relevant, and adding c.l.a on the list of newsgroups it goes to.


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

Re: Is Ada preferred over C/C++ for the realtime domain?

[Pat Rogers]

"Donald Gillies" <gillies@cs.ubc.ca> wrote in message
news:a9n4sa$dfj$1@cascade.cs.ubc.ca...
> "Pat Rogers" <rogers@gnat.com> writes:
>
> >This is a troll, right?  Does anyone really still believe that Ada (either
> >version) was designed by a committee?  Both were designed by small design
> >teams, each with a very strong leader.  Oh, and each team had a great deal
> >of experience developing large systems, as well as designing languages.
>
> Oh Pat, stop trolling yourself.  Pat ignores the committee-based
> history of Ada, which started out with the "Strawman" requirements
> specification, then "Woodenman", "Tinman", "Ironman", and "Steelman"
> and as a result, ADA.

OK, first things first:  It is "Ada", the name of a person.  If you don't know
that -- the very first thing that anybody knows about the language -- then you
don't know zilch about it.  It is a trivial issue, but it is a very good
tripwire.

Second, are you saying you cannot distinguish between phases of development?
Certainly the requirements was done by a commitee -- who said they weren't?  The
language design was done by a different group -- a team. The revision design was
also done by a team.  Those are both facts.  You don't have to like it.

> The language ADA was defined in the 1970's,
> when nobody have a clear idea of what was essential in a language, and
> what could be deferred to a software library.
>
> So, one of the most unique things about the ADA language is : It's the
> only language that has had a sex change.

Neither of us have any idea what you're talking about.


> The history of ADA is riddled with blunders and excessive
> overcomplexity.  Rabid Adaphiles try to hide all of the embarassing
> details.  Here are just a few.  Since it was designed in total vacuum,
> nobody even bothered to test the grammer and it was found not to be
> LALR(1).  This one mistake contributed to the extremely embarassing
> fact that the first compiler for the full language took 4 years to
> develop and it was written in ML and compiled 1 line per minute.
>
>
> I could go on but its too much like shooting fish in a barrel.

Why?  You haven't said anything right yet.  It wasn't four years, it wasn't ML
(it was SETL), and it wasn't 1 line per minute.  I used Ada compilers well
before the one you're thinking of.

<snip>
> Another thing to care about is that the syntax and semantics of the
> language does not force a person into a certain programming style, and

Ada does not force anything.  What nonsense.

> does not discourage the use of fundamental arithmatic operations (such
> as shifts and pointer increment/decrement) that exist on all modern
> computers.

What are you talking about?

> In this respect, and for modern systems programming, C
> still kicks the ass of C++ and ADA.

Unsupported rubbish.

You're information is both wrong and out of date.

Re: Is Ada preferred over C/C++ for the realtime domain?

[Wes Groleau]

> > The history of ADA is riddled with blunders and excessive
> > overcomplexity.  Rabid Adaphiles try to hide all of the embarassing
> > details.  Here are just a few.  Since it was designed in total vacuum,
> > nobody even bothered to test the grammer and it was found not to be
> > LALR(1).  This one mistake contributed to the extremely embarassing
> > fact that the first compiler for the full language took 4 years to
> > develop and it was written in ML and compiled 1 line per minute.

Rabid Adaphobes are so scared of Ada that they
post lies to keep people from trying it.

:-)

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

Re: Is Ada preferred over C/C++ for the realtime domain?

[Marin David Condic]

Just to clarify a bit on "The Mandate". It wasn't dropped because of a lack
of Ada expertise. It was dropped because the DoD commissioned a study on DoD
software requirements that concluded that Ada was fine for mission critical
software but shouldn't be a requirement for other software & recommended a
process by which projects should justify language choice based on life cycle
costs. The decision ended up going further than the study recommended by not
requiring Ada for anything and insisting projects justify language choice
based on life cycle costs.

The idea was not to discourage the use of Ada but to enable project managers
to look at the application at hand and make an intelligent decision about
what made sense for that particular situation. (For example, an application
that might be really graphics intensive or perhaps have a large AI
requirement would then be allowed to use some language or development
environment that had really good support for those features, rather than
require Ada, which is a more general purpose language.)

It did tend to get interpreted as "The DoD is ditching Ada!!!", which is
understandable, but incorrect. There are still lots of DoD software projects
that use Ada because it is well suited to a wide range of their problems.
There are also lots of commercial Ada projects as well - especially where
high reliability, long project life and real-time/embedded considerations
are a big deal.

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


"Pat Kohli" <kohli@ameritel.net> wrote in message
news:3CBF7A7A.A67B7D92@ameritel.net...
>
> the embedded telecom code is C/C++.  Years ago Ada was mandated as a
> non-functional requirement for US military applications.  At some point
the
> brass realized there was a shortfall of Ada expertise, given the US
military's
> SW requirements.  The requirement went away as a mandate; and Ada lives on
as a
> language which is particularly well suited to RT applications:  cruise
missiles,
> fly-by-wire, etc.  You don't see Ada used in writing clients for
relational
> databases, much, though, w/ the lapse in the military mandate.
>
>
>
>

Re: Is Ada preferred over C/C++ for the realtime domain?

[Stefan Skoglund]

Marin David Condic wrote:
> It did tend to get interpreted as "The DoD is ditching Ada!!!", which is
> understandable, but incorrect. There are still lots of DoD software projects

Also it is that interpretations has a habit of supporting other
intrests of the interpreter.

If you are a manager or a programmer, you look on a project and if you
know
what you arent gonna work with this project for ever you will
try to import your own views in the project like a preference
for Jave or C# .
Why ? because in 6 months you are gonna start looking for a new
job and at that time it is easier to get it if you are
knowledgeable in the current fad.

Re: Is Ada preferred over C/C++ for the realtime domain?

[Matt Majka]

Marin David Condic  wrote:

>
>There are also lots of commercial Ada projects as well - especially where
>high reliability, long project life and real-time/embedded considerations
>are a big deal.

The Boeing 777, for example.
Ada is required by Boeing for DO-178B Levels A and B.  For lower
levels it is still preferred, but you can try to convince them to allow
you to use something else.  They shy away from C++ altogether.

Re: Is Ada preferred over C/C++ for the realtime domain?

[Antonio López]

> It did tend to get interpreted as "The DoD is ditching Ada!!!", which is
> understandable, but incorrect. There are still lots of DoD software projects
> that use Ada because it is well suited to a wide range of their problems.
> There are also lots of commercial Ada projects as well - especially where
> high reliability, long project life and real-time/embedded considerations
> are a big deal.

Also, I think that bindings will increase the use of the Ada language
in areas where is not posible to perform the whole developpement in
Ada.
The most critical components of an application can be written in Ada
and compiled as a Dll for his use, while the rest of the application
can be developped in C++, Java, ...

Antonio

Re: Is Ada preferred over C/C++ for the realtime domain?

[Marin David Condic]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1223 bytes --]

Yes, except that mixed language development can be really painful. This is
especially true for embedded projects where you can have really important
issues with compilers & tools.

There are certainly a large number of things that can readily be done
entirely in Ada. Its also relatively easy to connect Ada to other languages
where you might have large existing libraries you want to leverage. You
might be able to identify equivalent libraries in Ada if it is something
general enough. Deciding to mix languages has its problems, but might be
worth doing depending on what you want to achieve.

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


"Antonio L�pez" <alm@gtd.es> wrote in message
news:47759943.0204250000.5c80a803@posting.google.com...
>
> Also, I think that bindings will increase the use of the Ada language
> in areas where is not posible to perform the whole developpement in
> Ada.
> The most critical components of an application can be written in Ada
> and compiled as a Dll for his use, while the rest of the application
> can be developped in C++, Java, ...
>
> Antonio