Embedded Ada Development Tools

Embedded Ada Development Tools

[Jason Gerry]

I am looking to do embedded developement using Ada.  I realize the
value in working with Ada in a saftey critical environment (as
compared to C).  There seems to be a terrible problem in doing a
design in Ada in an embedded environment.  From what I have seen,
there are very little tools, besides Green Hills, and limited targets
for embedded Ada development.  My project doesn't require and embedded
OS, and I would like to feel that there were more targets, Compiler
vendors, out there before commiting to serious design/developement
work.  I do give gredit for Greens Hills to get there Ada environments
to market (www.ghs.com).

Thanks to All...

Jason Gerry

Re: Embedded Ada Development Tools

[Jim Rogers]

Jason Gerry wrote:

> I am looking to do embedded developement using Ada.  I realize the
> value in working with Ada in a saftey critical environment (as
> compared to C).  There seems to be a terrible problem in doing a
> design in Ada in an embedded environment.  From what I have seen,
> there are very little tools, besides Green Hills, and limited targets
> for embedded Ada development.  


There are additional embedded development environment including
GNAT, Aonix, and DDCI.

Jim Rogers

Re: Embedded Ada Development Tools

[Marin David Condic]

We might also mention that a good place to start looking is:
http://www.adapower.com/ - in the "Links" section you will find links to a
variety of Ada vendors as well as other things. There are lots of Ada
resources to be found by starting at this web page.

Its a shame that Ada, having been invented to serve in the embedded arena,
doesn't meet with more success competing against other languages for
development projects. Its well suited to the task, but suffers from lack of
interest by embedded programmers & as a result tends to lack the compilers
and tools targeted to a lot of embedded machines. There are a number of
machines for which you *can* get Ada & related tools, but its not nearly the
number one sees for C compilers.

If you're using one of the more common processors, you should be able to
find what you need. If its a bit off the beaten path, you may have problems.

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

"Jim Rogers" <jimmaureenrogers@worldnet.att.net> wrote in message
news:3CFBC246.7020201@worldnet.att.net...
>
>
> There are additional embedded development environment including
> GNAT, Aonix, and DDCI.
>

Re: Embedded Ada Development Tools

[Greg Bek]

Rational Software also provide Ada compilers for embedded systems.
See http://www.adaic.org/compilers/cpl/lists/CPLbase.html for the current
list of certified compilers.

Embedded targets supported: 68K Family, RH32, PowerPC, MIPS, Intel.
Embedded Operating Systems: Tornado, LynxOS, and Rational Exec (bare board).
(Not all combinations of RTOS/processor supported).

Greg Bek
Product Manager - Rational Apex




"Jim Rogers" <jimmaureenrogers@worldnet.att.net> wrote in message
news:3CFBC246.7020201@worldnet.att.net...
> Jason Gerry wrote:
>
> > I am looking to do embedded developement using Ada.  I realize the
> > value in working with Ada in a saftey critical environment (as
> > compared to C).  There seems to be a terrible problem in doing a
> > design in Ada in an embedded environment.  From what I have seen,
> > there are very little tools, besides Green Hills, and limited targets
> > for embedded Ada development.
>
>
> There are additional embedded development environment including
> GNAT, Aonix, and DDCI.
>
> Jim Rogers
>
>

Re: Embedded Ada Development Tools

[John Kern]

Marin David Condic wrote:
> 

> If you're using one of the more common processors, you should be able to
> find what you need. If its a bit off the beaten path, you may have problems.
> 

What embedded processors have sold more than the Motorola HC12 family?

I suspect a lot of embedded projects start off small, (8/16-bit) and
eventually graduate to the 32-bitters where switching to Ada would
involve abandoning proven legacy C code.  Ada seems to be mostly for
markets which have always though big (like aerospace) as opposed to the
industries started small (like automotive.)

One of these days, I'm going to buy my son the Lego Mindstorms kit and
get that Ada compiler working.

Re: Embedded Ada Development Tools

[Marin David Condic]

I've seen Ada for 16-bit processors, but you're point is taken. Most Ada
implementations target 32-bit or bigger machines and while this is nice for
some of the more grandiose projects, its not where most of the little
commercial things live. And of course, even if it were possible to get an
Ada compiler to target some small 8-bit processor, we'd still be stuck with
that chicken-and-egg problem. Developers have no Ada compiler for these
targets from which to choose, so their interests go elsewhere to what they
can get. (Generally, C)Compiler-writers won't make something for that target
because its a lot of work and they don't see any interested developers to
justify the work.  Such lack of compilers leads the developers to conclude
that it isn't possible to get Ada targeted to their processors and fuels a
lot of the misconceptions about Ada (too slow, too big, etc.)

Maybe the correct answer is to do what seems to be the current direction -
get Ada used in more PC/Workstation apps where the targeting job is a bit
easier and as it gets more popular in this domain, it starts making it
easier to justify targeting more embedded targets because of programmer
familiarity, development of tools, etc.

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


"John Kern" <jkern3@NOSPAM.visteon.com> wrote in message
news:3CFE29BB.A7D22E77@NOSPAM.visteon.com...
>
> I suspect a lot of embedded projects start off small, (8/16-bit) and
> eventually graduate to the 32-bitters where switching to Ada would
> involve abandoning proven legacy C code.  Ada seems to be mostly for
> markets which have always though big (like aerospace) as opposed to the
> industries started small (like automotive.)
>

Re: Embedded Ada Development Tools

[Greg Bek]

John Kern <jkern3@NOSPAM.visteon.com> wrote in message news:<3CFE29BB.A7D22E77@NOSPAM.visteon.com>...
> 
> One of these days, I'm going to buy my son the Lego Mindstorms kit and
> get that Ada compiler working.

The best way to do this is to start with Barry Fagin's ada2nqc program.
It allows you to write pure Ada code and then translate it into NQC
for compilation and download to the RCX.

I took Barry's code and added tasking and a couple of other things
to it, although I think Barry has since done it as well.

There is some slight mismatch between Ada tasking semantics and
the Lego RCX tasking, but it doesn't stop you writing pure Ada
code.  The code might not work if ported to another target, but
that isn't a real problem in this domain.

Here is some sample code in Ada:
with Lego;
use Lego;
procedure scanbot is
  Angle : constant Sensor_Port := Sensor_1;
  Eye : constant Sensor_Port := Sensor_3;
  
  Left : constant Output_Port := Output_A;
  Head : constant Output_Port := Output_B;
  Right : constant Output_Port := Output_C;
  
  Slop : constant Integer := 2;
  
  Center : constant Integer := 53;
  Sweep : constant Integer := 48;
  
  Align_Time : constant Integer := 300; -- 3 seconds
  Turn_Time : constant Integer := 50; -- 1/2 second
  
  Threshold : constant Integer := 65;
  
  procedure Align is
     Current : Integer;
  begin
    Output_Reverse (Head);
    Clear_Sensor(Angle);
    Current := Get_Sensor_Value(Angle);
    Output_On(Head);
    Wait (100);
    while ( Current /= Get_Sensor_Value (Angle )) loop
       Current := Get_Sensor_Value(Angle);
       Wait (200);
    end loop;
   
    Output_Off(Head);
    Output_Forward (Head);
    Clear_Sensor(Angle);
  end Align;
  
  task Steer is
     entry Start;
     entry Stop;
  end Steer;
  
  task body Steer is
  begin
     accept Start;
     While (True) loop
      if (Eye > Threshold) then
         Output_Off(Head);
         Play_Sound(Click);
         if (Angle < Center - Slop) then
            Output_Reverse(Left);
            Wait(Turn_Time);
            Output_Forward(Left);
         else if (Angle > Center + Slop ) then
               Output_Reverse(Right);
               Wait(Turn_Time);
               Output_Forward(Right);
            end if;
         end if;
         Output_On(Head);

      end if;     
     end loop;
  end Steer;

begin
  Config_Sensor(Eye, Config_Light);
  Config_Sensor(Angle, Config_Rotation);

  Align;

  Output_on ( Left + Right + Head );

  Steer.Start;

  while (True) loop
    if (Angle < (Center - Sweep)) then
      Output_Forward (Head);
    else
       if (Angle > (Center + Sweep)) then
          Output_Reverse (Head);
       end if;
    end if;   
  end loop;  

end;

Re: Embedded Ada Development Tools

[John English]

Greg Bek wrote:
> 
> John Kern <jkern3@NOSPAM.visteon.com> wrote in message news:<3CFE29BB.A7D22E77@NOSPAM.visteon.com>...
> >
> > One of these days, I'm going to buy my son the Lego Mindstorms kit and
> > get that Ada compiler working.
> 
> The best way to do this is to start with Barry Fagin's ada2nqc program.
> It allows you to write pure Ada code and then translate it into NQC
> for compilation and download to the RCX.

Do you have a URL for this?
 
-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.it.bton.ac.uk/staff/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: Embedded Ada Development Tools

[Jason Gerry]

I agree with most of the information posted.... and.... not to start a
flame war or anything,  but could you imagine doing embedded Ada
development within a small project team, and the ada developer was to
leave the company..... Yikes! ... I understand Ada is easier to
maintain, but you still need someone off the street who is adept at
embedded systems programming and Ada coding.... It would keep me up at
night.


Jason Gerry

RE: Embedded Ada Development Tools

[Robert C. Leif]

From: Bob Leif
To: Marin Condic et al.
The simplest way to port Ada to a small processor is to use a J code
compiler. This is absolutely not a new idea. It goes back to at least
UCSD Pascal. 

The best approach would be to a have a native J code processor which
included an ADC, DAC, watch-dog timer, and USB port. 

-----Original Message-----
From: comp.lang.ada-admin@ada.eu.org
[mailto:comp.lang.ada-admin@ada.eu.org] On Behalf Of Marin David Condic
Sent: Wednesday, June 05, 2002 9:55 AM
To: comp.lang.ada@ada.eu.org
Subject: Re: Embedded Ada Development Tools

I've seen Ada for 16-bit processors, but you're point is taken. Most Ada
implementations target 32-bit or bigger machines and while this is nice
for
some of the more grandiose projects, its not where most of the little
commercial things live. And of course, even if it were possible to get
an
Ada compiler to target some small 8-bit processor, we'd still be stuck
with
that chicken-and-egg problem. Developers have no Ada compiler for these
targets from which to choose, so their interests go elsewhere to what
they
can get. (Generally, C)Compiler-writers won't make something for that
target
because its a lot of work and they don't see any interested developers
to
justify the work.  Such lack of compilers leads the developers to
conclude
that it isn't possible to get Ada targeted to their processors and fuels
a
lot of the misconceptions about Ada (too slow, too big, etc.)

Maybe the correct answer is to do what seems to be the current direction
-
get Ada used in more PC/Workstation apps where the targeting job is a
bit
easier and as it gets more popular in this domain, it starts making it
easier to justify targeting more embedded targets because of programmer
familiarity, development of tools, etc.

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


"John Kern" <jkern3@NOSPAM.visteon.com> wrote in message
news:3CFE29BB.A7D22E77@NOSPAM.visteon.com...
>
> I suspect a lot of embedded projects start off small, (8/16-bit) and
> eventually graduate to the 32-bitters where switching to Ada would
> involve abandoning proven legacy C code.  Ada seems to be mostly for
> markets which have always though big (like aerospace) as opposed to
the
> industries started small (like automotive.)
>

Re: Embedded Ada Development Tools

[Jerry Petrey]

John English wrote:

>
> > The best way to do this is to start with Barry Fagin's ada2nqc program.
> > It allows you to write pure Ada code and then translate it into NQC
> > for compilation and download to the RCX.
>
> Do you have a URL for this?
>
>

John, here are some links to Barry's work on Ada Mindstorms:

http://www.faginfamily.net/barry/Papers/AdaLetters.htm
http://www.usafa.af.mil/dfcs/Ada_Mindstorms_manual.htm
http://www.faginfamily.net/barry/Papers/IEEERA.htm
http://www.acm.org/sigada/conf/sigada2001/private/SIGAda2001-CDROM/SIGAda2001-Proceedings/Fagin-Paper.pdf


Jerry
--
---------------------------------------------------------------------------------
-- Jerry Petrey
-- Senior Principal Systems Engineer - Navigation (GPS/INS), Guidance, & Control
-- Raytheon Missile Systems          - Member Team Ada & Team Forth
-- NOTE: please remove <NOSPAM> in email address to reply
---------------------------------------------------------------------------------

RE: Embedded Ada Development Tools

[Steven Deller]

Jason,
Emenu has over 20 engineers with EXACTLY that expertise -- embedded
systems and Ada.  Each has over 14 years experience.  We cover areas
from Southern CA to New England, Texas, Arizona, Minnesota, Maryland and
more (our engineers are willing to travel).  And we all can train a
project's own personnel in Ada, embedded systems, and numerous
application areas for embedded systems.  We do C++ and other languages,
but ALL of us would prefer Ada -- it gives us the best way to express
our quality engineering approaches.  We use OO, OO tools that generate
Ada and pay attention to *architecture* as much as implementation.  We
cover classes to bits and everything in between.

And Emenu is NOT, by far, the only source of excellent expertise in Ada
and embedded systems.

If you look, you will find...

Regards,
Steve Deller
Emenu Incorporated
Steve.Deller@Emenu-Inc.com
410 757 6924

> -----Original Message-----
> [mailto:comp.lang.ada-admin@ada.eu.org] On Behalf Of Jason Gerry
> Sent: Thursday, June 06, 2002 8:19 AM
> To: comp.lang.ada@ada.eu.org
> Subject: Re: Embedded Ada Development Tools
> 
> I agree with most of the information posted.... and.... not 
> to start a flame war or anything,  but could you imagine 
> doing embedded Ada development within a small project team, 
> and the ada developer was to leave the company..... Yikes! 
> ... I understand Ada is easier to maintain, but you still 
> need someone off the street who is adept at embedded systems 
> programming and Ada coding.... It would keep me up at night.

Re: Embedded Ada Development Tools

[John R. Strohm]

Jason, every company that has actually TRIED training programmers to use Ada
has learned that it is not at all difficult, and that a good programmer very
quickly becomes productive in Ada.

When Boeing mandated Ada for the 777, every subcontractor in the country
screamed bloody murder.  Boeing insisted.  None of them wanted to lose out
on the biggest contract they would see for a long time, with the biggest
airplane manufacturer in the world, that carried the certain knowledge that
whoever took this job WOULD get the follow-on work for the NEXT airplane,
and the rules would probably be the same next time around.  They all toed
the line.

They ALL came in on time and in budget.  Despite training costs.  Despite,
in at least one case, having to scrap a bunch of already-completed work and
start over from scratch.

Furthermore, when the companies really, seriously, did the metrics
collection and looked at the results, they learned that Ada saved them time
and money.  The best example is the Pratt & Whitney experience, for which
Marin David Condic is the knowledgeable individual who was in the trenches
at the time.

"Jason Gerry" <jgerry11@yahoo.com> wrote in message
news:e8a5e1a4.0206060519.ce12ba@posting.google.com...
> I agree with most of the information posted.... and.... not to start a
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company..... Yikes! ... I understand Ada is easier to
> maintain, but you still need someone off the street who is adept at
> embedded systems programming and Ada coding.... It would keep me up at
> night.
>
>
> Jason Gerry

Re: Embedded Ada Development Tools

[Jeffrey Carter]

Jason Gerry wrote:
> 
> I agree with most of the information posted.... and.... not to start a
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company..... Yikes! ... I understand Ada is easier to
> maintain, but you still need someone off the street who is adept at
> embedded systems programming and Ada coding.... It would keep me up at
> night.

This is FUD, pure and simple.

-- 
Jeff Carter
"I fart in your general direction."
Monty Python & the Holy Grail

Re: Embedded Ada Development Tools

[Jeffrey Carter]

Marin David Condic wrote:
> 
> I've seen Ada for 16-bit processors, but you're point is taken. Most Ada
> implementations target 32-bit or bigger machines and while this is nice for
> some of the more grandiose projects, its not where most of the little
> commercial things live. And of course, even if it were possible to get an
> Ada compiler to target some small 8-bit processor, we'd still be stuck with
> that chicken-and-egg problem. Developers have no Ada compiler for these
> targets from which to choose, so their interests go elsewhere to what they
> can get. (Generally, C)Compiler-writers won't make something for that target
> because its a lot of work and they don't see any interested developers to
> justify the work.  Such lack of compilers leads the developers to conclude
> that it isn't possible to get Ada targeted to their processors and fuels a
> lot of the misconceptions about Ada (too slow, too big, etc.)

Anything that is targeted by an ANSI C compiler is also targeted by
Averstar's Ada-to-ANSI-C compiler.

-- 
Jeff Carter
"I fart in your general direction."
Monty Python & the Holy Grail

Re: Embedded Ada Development Tools

[Bill Tate]

jgerry11@yahoo.com (Jason Gerry) wrote in message news:<e8a5e1a4.0206060519.ce12ba@posting.google.com>...
> I agree with most of the information posted.... and.... not to start a
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company..... Yikes! ... I understand Ada is easier to
> maintain, but you still need someone off the street who is adept at
> embedded systems programming and Ada coding.... It would keep me up at
> night.
> 
> 
> Jason Gerry

"...small project team..." and "...who is adept..."  Ahhh - an image
is emerging in my mind :>)

Re: Embedded Ada Development Tools

[Ted Dennison]

jgerry11@yahoo.com (Jason Gerry) wrote in message news:<e8a5e1a4.0206060519.ce12ba@posting.google.com>...
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company..... Yikes! ... I understand Ada is easier to
> maintain, but you still need someone off the street who is adept at
> embedded systems programming and Ada coding.... It would keep me up at
> night.

Presumably you are comparing this with a similar situation using C. 

Actually, I know of a couple of studies that found that novice Ada
programmers who knew C still had more productivity and less bugs when
they used Ada instead of C. Plus, Ada's generally considered easier to
read and understand than C, which will be a big help once the
replacement developer is hired.

So actually, I think I'd be way more worried about my embedded
programmer for a C system leaving than for an Ada one.


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

Re: Embedded Ada Development Tools

[Richard Riehle]

Marin David Condic wrote:

> I've seen Ada for 16-bit processors, but you're point is taken. Most Ada
> implementations target 32-bit or bigger machines and while this is nice for
> some of the more grandiose projects, its not where most of the little
> commercial things live. And of course, even if it were possible to get an
> Ada compiler to target some small 8-bit processor, we'd still be stuck with
> that chicken-and-egg problem.

Maybe it's time for RR Software to resurrect that 8-bit Ada compiler
they developed back in the Eighties.

Richard Riehle

Re: Embedded Ada Development Tools

[Richard Riehle]

"John R. Strohm" wrote:

> Furthermore, when the companies really, seriously, did the metrics
> collection and looked at the results, they learned that Ada saved them time
> and money.  The best example is the Pratt & Whitney experience, for which
> Marin David Condic is the knowledgeable individual who was in the trenches
> at the time.

Just read my latest edition of Crosstalk.  Yet another article
that names every programming language ever invented
and completely ignores any mention of Ada.   And here
we have an author whose credibility in implying that
Ada is irrelevant is reinforced by having been at AJPO.

Even as we educate people about the benefits of Ada, former DoD
officials with their focus on consulting dollars throw unnecessary
obstacles in the path of wider acceptance.

Richard Riehle

Re: Embedded Ada Development Tools

[Richard Riehle]

Ted Dennison wrote:

> Actually, I know of a couple of studies that found that novice Ada
> programmers who knew C still had more productivity and less bugs when
> they used Ada instead of C. Plus, Ada's generally considered easier to
> read and understand than C, which will be a big help once the
> replacement developer is hired.

I just took a quick tour through a new book from Microsoft Press
titled, Writing Secure Code.   It would be hilarious if it were not
so pathetic.    Over and over the authors demonstrate how to avoid
the pitfalls of writing code in C and C++.   Over and over I find
myself reacting with, "This problem simply would not occur if
they were writing in Ada."

The authors are well-intentioned and intelligent.  They even admit,
early in the book, that they are writing about C code because that
is the more widely used language even though it is clearly less
secure and less safe than many other languages.  Of course, they
never mention Ada since they probably no nothing about it.

One of my favorite monthly columns in C++ Report was called
"Obfuscated C++."    It was also a source of amusement because
the obfuscations described simply could not have happened in
Ada.   Still, these "professionals" continue to sell their  souls
to the Devil when it would be so much easier to harmonize
with the angels.

Richard Riehle

Re: Embedded Ada Development Tools

[Richard Riehle]

Greg Bek wrote:

> Rational Software also provide Ada compilers for embedded systems.
> See http://www.adaic.org/compilers/cpl/lists/CPLbase.html for the current
> list of certified compilers.
>
> Embedded targets supported: 68K Family, RH32, PowerPC, MIPS, Intel.
> Embedded Operating Systems: Tornado, LynxOS, and Rational Exec (bare board).
> (Not all combinations of RTOS/processor supported).

HmmmmmMMMMmmmmmmm!

Maybe Rational should make the old RR Software Eight Bit compiler
part of their compiler portfolio.   That would deal with the problem
of eight bit compilers.

Richard Riehle

Re: Embedded Ada Development Tools

[Michael Bode]

Richard Riehle <richard@adaworks.com> writes:

> Maybe Rational should make the old RR Software Eight Bit compiler
> part of their compiler portfolio.   That would deal with the problem
> of eight bit compilers.

GCC can target the Atmel AVR 8-Bit controllers. But I fear that does
not mean one can pick some compiler switches during GCC build and get
Gnat for AVR. If someone would find a way to make even a subset of Ada
available on the AVR through GCC that would be a nice start.

Re: Embedded Ada Development Tools

[John English]

Jason Gerry wrote:
> 
> I agree with most of the information posted.... and.... not to start a
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company.....

>THE< Ada developer? What sort of a team is that, with only one developer
who understands the language being used? That's a VERY small project
team...

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.it.bton.ac.uk/staff/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: Embedded Ada Development Tools

[Marin David Condic]

A couple of things: It isn't *that* hard to find people who know Ada &
embedded systems. It isn't *that* hard to get someone up to speed enough on
Ada to make them productive. All languages have to start from somewhere &
obviously people succeed in learning them as they emerge and needs arise to
use them, so I don't really see the problem. (Note that things like C# and
.Net are emerging and suddenly everyone wants to find people with 20 years
of experience working with them. Too bad. You'll just have to *train* them,
eh? :-)

By your argument, embedded systems should never have gone outside of
assembler because someone might learn the "new" language and quit and you
wouldn't be able to find people who knew the new language. Yet it has been
done over and over again. Fear not - if the need arises, intelligent
programmers will learn whatever language is needed to get the job done. :-)

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


"Jason Gerry" <jgerry11@yahoo.com> wrote in message
news:e8a5e1a4.0206060519.ce12ba@posting.google.com...
> I agree with most of the information posted.... and.... not to start a
> flame war or anything,  but could you imagine doing embedded Ada
> development within a small project team, and the ada developer was to
> leave the company..... Yikes! ... I understand Ada is easier to
> maintain, but you still need someone off the street who is adept at
> embedded systems programming and Ada coding.... It would keep me up at
> night.
>

Re: Embedded Ada Development Tools

[Marin David Condic]

True, but its a lousy answer to someone with some small embedded target.
"Let me get this straight: I've got to buy *two* compilers and figure out
procedures to integrate all that stuff into a development environment? Why
don't I just get my little two-month, 5000 line job done with the C compiler
that came with my embedded board and just be done with it?"

Supposing for a minute that Ada can be made to fit some small SBC - it
probably can in most cases. The only way your average developer is going to
be willing to go with Ada is if Ada is one of the languages that comes with
the development kit for the SBC. Otherwise, there's just way too much work,
hassle and risk with trying to cobble something else together and since most
really small computers are used for really small jobs, there just isn't
going to be enough food value with Ada to justify going to all the extra
effort.

I respect the fact that Averstar has *an* answer and I'm sure there are
times when this makes sense, but you have to look at the realities of the
small-computer market and remember that the object of the game isn't to use
Ada, but to get your SBC project built and out the door quickly and at the
lowest possible cost.

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


"Jeffrey Carter" <jrcarter@acm.org> wrote in message
news:3CFFB7A3.178C0549@acm.org...
>
> Anything that is targeted by an ANSI C compiler is also targeted by
> Averstar's Ada-to-ANSI-C compiler.
>

Re: Embedded Ada Development Tools

[Larry Kilgallen]

In article <3D00A230.F02C0CF4@brighton.ac.uk>, John English <je@brighton.ac.uk> writes:
> Jason Gerry wrote:
>> 
>> I agree with most of the information posted.... and.... not to start a
>> flame war or anything,  but could you imagine doing embedded Ada
>> development within a small project team, and the ada developer was to
>> leave the company.....
> 
>>THE< Ada developer? What sort of a team is that, with only one developer
> who understands the language being used? That's a VERY small project
> team...

What sort of a team is that with other programmers who are incapable
of learning the language being used ?

What sort of a management is that, who pay the other programmers to do
nothing ?

Re: Embedded Ada Development Tools

[Jeffrey Carter]

Marin David Condic wrote:
> 
> True, but its a lousy answer to someone with some small embedded target.
> "Let me get this straight: I've got to buy *two* compilers and figure out
> procedures to integrate all that stuff into a development environment? Why
> don't I just get my little two-month, 5000 line job done with the C compiler
> that came with my embedded board and just be done with it?"

Since your board came with a C compiler, you don't have to buy 2
compilers. You only have to buy the Ada-to-C translator once, after
which it will work with many such boards.

I agree that this approach is not going to win any converts to Ada. But
to those who have to develop on such systems, and who would rather use
Ada, it may be an attractive alternative. I may recall incorrectly (is
that even possible? :), but I think that was the OP's position.

-- 
Jeff Carter
"Your mother was a hamster and your father smelt of elderberries."
Monty Python & the Holy Grail

Re: Embedded Ada Development Tools

[Marin David Condic]

Arguing that since you already have the C compiler, its a sunk cost might
get around that part of the expenditure, but you still had to pay for it. I
don't think it makes a convincing argument for why someone would want to go
to the *extra* effort and cost of buying an Ada-to-C translator, figuring
out how to connect everything up into a development kit, dealing with tool
compatibility problems (symbolic debuggers, for example), etc.

You've *really* got to believe that Ada is going to buy you back all the
added cost and risk in some way to be convinced this is what you should do.
For lots of small SBC projects, I just don't see it. "I bought a development
board and it came with this C compiler, loader, debugger, monitor, etc. It
cost me $500. My project is likely to amount to 5000 lines of code and two
months of development. I'm going to spend *how much* to buy an Ada to C
translator and I've got to invest *how long* to figure out how to connect
this up with my $500 development kit???? When is it that I realize the
benefit to justify the cost?"

As far as I can see, if Ada wants to play in the small SBC market, it needs
to come along with that $500 development board as an already integrated item
that provides every bit of functionality that the C compiler does, or it
just doesn't stand a chance of winning any converts.

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


"Jeffrey Carter" <jrcarter@acm.org> wrote in message
news:3D0176F9.CBF39FB1@acm.org...
>
> Since your board came with a C compiler, you don't have to buy 2
> compilers. You only have to buy the Ada-to-C translator once, after
> which it will work with many such boards.
>
> I agree that this approach is not going to win any converts to Ada. But
> to those who have to develop on such systems, and who would rather use
> Ada, it may be an attractive alternative. I may recall incorrectly (is
> that even possible? :), but I think that was the OP's position.
>

Re: Embedded Ada Development Tools

[Michael Bode]

"Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:

> As far as I can see, if Ada wants to play in the small SBC market, it needs
> to come along with that $500 development board as an already integrated item
> that provides every bit of functionality that the C compiler does, or it
> just doesn't stand a chance of winning any converts.

So is there any $500 (or less) Ada compiler for any small embedded system?

Re: Embedded Ada Development Tools

[Randy Brukardt]

Michael Bode wrote in message ...
>"Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org> writes:
>
>> As far as I can see, if Ada wants to play in the small SBC market, it
needs
>> to come along with that $500 development board as an already
integrated item
>> that provides every bit of functionality that the C compiler does, or
it
>> just doesn't stand a chance of winning any converts.
>
>So is there any $500 (or less) Ada compiler for any small embedded
system?

Janus/Ada 83 for MS-DOS is still $500. It has been used for embedded
186-based systems using an embedded MS-DOS. (The full bare-machine
version costs a lot more than $500, though).

Don't know if this is small enough for you, though.

         Randy Brukardt.