Ada used in General Aviation (GA) applications?

Ada used in General Aviation (GA) applications?

[Kai Glaesner]

Hello,

there is a (silent) revolution going on in General Aviation these days. Long
after taking over in the cockpit of military and air-transport category
aircraft the computer now found its way into the GA airplane cockpit.

With companies like Garmin bringing state-of-the-art avionics down to
affordable levels it wont take long until a "glass-cockpit" becomes standard
in GA.

Now I wonder about what role Ada plays and will play in this business
segment. I hope it's a big one but anyone out there who knows (or can make
serious guess)?

Thanks in advane for answers

Kai

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

My guess is that like most embedded systems and most commercial/military 
avionics, it will be done in just about anything but Ada. The Joint 
Strike Fighter is pretty much entirely C++ with the exception being my 
engine control and I'm losing the battle on that one.

Avionics developers don't purchase hundreds of thousands of compilers, 
so they don't drive the market. They've got to get adequate tools, 
talent and support and don't have megabucks available to just go 
home-grow it, so they start shopping around for what they can get on the 
open market and that ends up being mostly C++ these days.

If Ada wants to be a player in General Aviation avionics it needs to do 
something to either a) target & capture some more "general" market that 
will provide the $$$ that will build the tools/talent/support or b) zero 
in on the GA market and invest heavily in building everything that 
market could possibly want. The "B" option is not practical because, as 
I've observed, we avionics builders don't buy huge volumes of compilers. 
The "A" option seems to meet with a general lack of interest on the part 
of those in a position to do something about it.

My guess is that Ada isn't going to make any real headway into the 
General Aviation market because the big commercial/military aviation 
market is ignoring it in mass numbers. But perhaps if some Ada 
enthusiasts were to try to come up with some marketable General Aviation 
product that happened to be programmed in Ada it might have some 
encouraging results. You just don't see a lot of Ada entrepreneurs out 
there though. :-)

MDC


Kai Glaesner wrote:
> Hello,
> 
> there is a (silent) revolution going on in General Aviation these days. Long
> after taking over in the cockpit of military and air-transport category
> aircraft the computer now found its way into the GA airplane cockpit.
> 
> With companies like Garmin bringing state-of-the-art avionics down to
> affordable levels it wont take long until a "glass-cockpit" becomes standard
> in GA.
> 
> Now I wonder about what role Ada plays and will play in this business
> segment. I hope it's a big one but anyone out there who knows (or can make
> serious guess)?
> 
> Thanks in advane for answers
> 
> Kai
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Marin David Condic wrote:

> My guess is that like most embedded systems and most commercial/military 
> avionics, it will be done in just about anything but Ada. The Joint 
> Strike Fighter is pretty much entirely C++ with the exception being my 
> engine control and I'm losing the battle on that one.

For US gov't projects, the longer they take and the more they cost, the 
greater the profit for the contractor. Since Ada makes projects shorter 
and cheaper, there are strong market forces against the use of Ada on 
these projects.

For civil aviation products, there are a lot of forces in action. Many 
civil products are based on military products, so one would expect them 
to use the same language as the military product.

For new products, the reasons are different. Language choice is a 
software engineering decision that is usually not made by software 
engineers (usually by managers or coders). They are not going to make 
the correct choice very often. This is no different than a manager 
committing to a schedule and budget before requirements definition takes 
place.

-- 
Jeff Carter
"Blessed are they who convert their neighbors'
oxen, for they shall inhibit their girth."
Monty Python's Life of Brian
83

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

Marin David Condic <nobody@noplace.com> wrote in
news:409F69CB.8020604@noplace.com: 

 
> If Ada wants to be a player in General Aviation avionics it needs to do
> something to either a) target & capture some more "general" market that
> will provide the $$$ that will build the tools/talent/support


There is a wide range of embedded applications on 8- and 16-bit processors 
like 8051 and HC11. But - no (Ada-)compiler vendor offers a compiler for 
such targets.

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

"Bernd Specht" <Bernd.Specht@gmx.com> wrote in message
news:Xns94E5D0503CC17BerndSpechtgmxcom@151.189.20.10...
> Marin David Condic <nobody@noplace.com> wrote in
> news:409F69CB.8020604@noplace.com:
>
>
> > If Ada wants to be a player in General Aviation avionics it needs to do
> > something to either a) target & capture some more "general" market that
> > will provide the $$$ that will build the tools/talent/support
>
>
> There is a wide range of embedded applications on 8- and 16-bit processors
> like 8051 and HC11. But - no (Ada-)compiler vendor offers a compiler for
> such targets.

There are GNAT ports to AVR microcontrollers (see
http://sourceforge.net/projects/avr-ada).

But compiler vendors aren't going to support targets for which they see no
demand - so if you seriously want an Ada compiler for such things ASK! :-)

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

I happen to work in that particular market.  Barco makes cockpit
displays.  The vast majority of the software is written in Ada; it is
my understanding that Ada is still mandatory for DO-178B
certification, but I may be wrong on this.  Now the general aviation
market is only a part of the markets that Barco addresses, and usually
GA aircraft tend to be on the higher end of the spectrum (i.e. at
least turboprop, not piston-engined).

-- 
Ludovic Brenta.

Re: Ada used in General Aviation (GA) applications?

[Dmitry A. Kazakov]

On Mon, 10 May 2004 20:10:38 +0000 (UTC), "Martin Dowie"
<martin.dowie@btopenworld.com> wrote:

>But compiler vendors aren't going to support targets for which they see no
>demand - so if you seriously want an Ada compiler for such things ASK! :-)

I doubt that customers would be happy to pay for compiler targeting.
Mine will definitely not. They simply take C. The compiler vendors
should foresee and invest into new platforms. There cannot be any
demand in a compiler that does not exist!

--
Regards,
Dmitry Kazakov
www.dmitry-kazakov.de

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

Dmitry A. Kazakov <mailbox@dmitry-kazakov.de> wrote in
news:r701a0dgctn5coihlk6ugti2vuni40m4e8@4ax.com: 

> On Mon, 10 May 2004 20:10:38 +0000 (UTC), "Martin Dowie"
> <martin.dowie@btopenworld.com> wrote:
> 
>>But compiler vendors aren't going to support targets for which they see
>>no demand - so if you seriously want an Ada compiler for such things
>>ASK! :-) 
> 
> I doubt that customers would be happy to pay for compiler targeting.
> Mine will definitely not. They simply take C. The compiler vendors
> should foresee and invest into new platforms. There cannot be any
> demand in a compiler that does not exist!
> 
> --
> Regards,
> Dmitry Kazakov
> www.dmitry-kazakov.de
> 


Full ACK!

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

"Martin Dowie" <martin.dowie@btopenworld.com> wrote in
news:c7onju$smf$1@titan.btinternet.com: 

> There are GNAT ports to AVR microcontrollers (see
> http://sourceforge.net/projects/avr-ada).

1. We work on 8051 and HC11, there is _no_ reason to switch to other 
processors (especially the programming language is no reason).

2. For professional development, it's not enough to have a compiler. You 
need a complete tool-chain (e.g. configurator, debugger ...)

> But compiler vendors aren't going to support targets for which they see
> no demand - so if you seriously want an Ada compiler for such things
> ASK! :-) 

No, thats wrong. Would you go to GM and tell them: "Please build me a car 
whith five wheels,60 inch wheels, ..."? NO, you go to your local dealer, see 
what he has and then buy it or leave it.

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

Ludovic Brenta <ludovic.brenta@insalien.org> wrote in message news:<87hduoynoq.fsf@insalien.org>...
> I happen to work in that particular market.  Barco makes cockpit
> displays.  The vast majority of the software is written in Ada; it is
> my understanding that Ada is still mandatory for DO-178B
> certification, but I may be wrong on this.  Now the general aviation
> market is only a part of the markets that Barco addresses, and usually
> GA aircraft tend to be on the higher end of the spectrum (i.e. at
> least turboprop, not piston-engined).

DO-178B has never mandated Ada - but has 'highly recommended' it.

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

That's a bit of a cynical perspective and one that I don't find evidence 
for in my dealings with DoD contractors. DoD contractors are under 
considerable pressure to reduce costs too - not just those in the 
general market.

It usually has more to do with "infrastructure" issues and at times the 
"follow the herd" syndrome - which is not always totally without 
rationality. It costs a lot to spit into the wind and if someone is 
going to do so, they need to be able to justify it with something 
quantifiable. Its easy to enumerate the costs of going against the tide. 
Its hard to enumerate the savings of using a superior technology that is 
not "mainstream". (Think: "Nobody ever got fired for picking IBM.")

I agree that Ada has technical superiority and - if one can get to that 
elusive "All other things being equal" state - one might make the case 
that can justify it. But things are never equal and Ada is behind C++ on 
most infrastructure issues. Its extremely expensive to use Ada in 
comparison to C++ on most of these sorts of projects and unless Ada can 
go demonstrate some really concrete superiority in cost savings 
elsewhere, its going to lose the competition.

I'll stand by my position that Ada would be better off trying to 
optimize its capabilities for a given market segment that is large 
enough to fund the vendor's work and from that position see about going 
after the smaller segment of embedded/avionics development. The two 
might go hand-in-hand if, for example, Ada were to start concentrating 
on providing superior capability in something like math functions or 
communications.

MDC



Jeffrey Carter wrote:
> 
> For US gov't projects, the longer they take and the more they cost, the 
> greater the profit for the contractor. Since Ada makes projects shorter 
> and cheaper, there are strong market forces against the use of Ada on 
> these projects.
> 
> For civil aviation products, there are a lot of forces in action. Many 
> civil products are based on military products, so one would expect them 
> to use the same language as the military product.
> 
> For new products, the reasons are different. Language choice is a 
> software engineering decision that is usually not made by software 
> engineers (usually by managers or coders). They are not going to make 
> the correct choice very often. This is no different than a manager 
> committing to a schedule and budget before requirements definition takes 
> place.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

There's a good example of why Ada loses out in lots of embedded markets. 
Its enormously expensive to try to address the embedded market and the 
payback can be releatively small. Maybe someone sells a million HC11 
processor based units, but they only need *one* compiler to build the 
code. Who's going to go develop that compiler?

Better to try to get a bigger target audience and see if the embedded 
stuff falls out from there.

MDC

Bernd Specht wrote:
> 
> There is a wide range of embedded applications on 8- and 16-bit processors 
> like 8051 and HC11. But - no (Ada-)compiler vendor offers a compiler for 
> such targets.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Bernd Specht wrote:
> 
> 1. We work on 8051 and HC11, there is _no_ reason to switch to other 
> processors (especially the programming language is no reason).
> 
> 2. For professional development, it's not enough to have a compiler. You 
> need a complete tool-chain (e.g. configurator, debugger ...)
> 
Amen! I've said this before, but it seems to get dismissed. But then, so 
does Ada in the embedded world. People buy a whole development kit and 
that isn't cheap to build. It is often basically given away when you buy 
a development board because it is the hope of the vendor to sell you 
lots of boards - not lots of compilers. So if Ada wants to play in that 
arena, it needs someone to go out and develop a superior board-level 
product that just happens to have an Ada compiler targeting it.


> 
> No, thats wrong. Would you go to GM and tell them: "Please build me a car 
> whith five wheels,60 inch wheels, ..."? NO, you go to your local dealer, see 
> what he has and then buy it or leave it.
> 
> 
You could ask, but the answer is "No - but we'll let you pay to develop 
it and you can wait a year to get it." You're absolutely right on this 
score. *NOBODY* is going to pay a vendor and wait for the privilege of 
using Ada. They're going to select their board/chip and they're going to 
  use whatever compilers/development environments are available for that 
hardware. If Ada isn't sitting on the shelf waiting to be bought, then 
Ada doesn't play.

Its a tough nut to crack and I think Ada would be better off trying to 
target something else that might end up spilling over into the embedded 
world rather than try a direct assault.

MDC



-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Britt Snodgrass]

Ludovic Brenta <ludovic.brenta@insalien.org> wrote in message news:<87hduoynoq.fsf@insalien.org>...
> I happen to work in that particular market.  Barco makes cockpit
> displays.  The vast majority of the software is written in Ada; it is
> my understanding that Ada is still mandatory for DO-178B
> certification, but I may be wrong on this.  Now the general aviation
> market is only a part of the markets that Barco addresses, and usually
> GA aircraft tend to be on the higher end of the spectrum (i.e. at
> least turboprop, not piston-engined).

DO-178B (see section 4.4.2) does not specify or recommend any
particular software language, regardless of level. Some C language
applications such as RTOS's have been certified to DO-178B level A. 
However the ARINC 653-1 standard(last updated in Oct 2003) does
recommend Ada (see section 1.4.2).

Britt

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

Marin David Condic <nobody@noplace.com> wrote in
news:40A0BBE0.9060209@noplace.com: 

> There's a good example of why Ada loses out in lots of embedded
> markets. Its enormously expensive to try to address the embedded market
> and the payback can be releatively small. Maybe someone sells a million
> HC11 processor based units, but they only need *one* compiler to build
> the code. Who's going to go develop that compiler?

Well, if there were someone who sells *millions* on  HC11/8051 products and 
needs only *one* compiler, then I would agree. But - in this range it is 
often the case, that a small company sells a few hundrets or thousands 
units, but *every* developer has its own develop environment. So to sell 
*millions* of units, there would be a lot of companies, who need a lot of 
development tools (_not only_ compiler, but the _whole_ toolchain). But - 
toolset must not be expensive (if one sells a few thousand boards, they 
won't be willing to pay $1000_000 for the tools).

> 
> Better to try to get a bigger target audience and see if the embedded 
> stuff falls out from there.
> 
> MDC
> 
> Bernd Specht wrote:
>> 
>> There is a wide range of embedded applications on 8- and 16-bit
>> processors like 8051 and HC11. But - no (Ada-)compiler vendor offers a
>> compiler for such targets.
>> 
> 
> 

Re: Ada used in General Aviation (GA) applications?

[Bernd Trog]

Bernd.Specht@gmx.com (Bernd Specht) wrote in message 

> 1. We work on 8051 and HC11, there is _no_ reason to switch to other 
> processors (especially the programming language is no reason).

Try http://gel.sourceforge.net/ada_example.php

It's a cross compiler for 68HC11 and 68HC12.

Use an Ada-to-C compiler for 8051.

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

"Martin Dowie" <martin.dowie@btopenworld.com> wrote in message
news:af783afe.0405110329.2f2352a5@posting.google.com...
> Ludovic Brenta <ludovic.brenta@insalien.org> wrote in message
news:<87hduoynoq.fsf@insalien.org>...
> > I happen to work in that particular market.  Barco makes cockpit
> > displays.  The vast majority of the software is written in Ada; it is
> > my understanding that Ada is still mandatory for DO-178B
> > certification, but I may be wrong on this.  Now the general aviation
> > market is only a part of the markets that Barco addresses, and usually
> > GA aircraft tend to be on the higher end of the spectrum (i.e. at
> > least turboprop, not piston-engined).
>
> DO-178B has never mandated Ada - but has 'highly recommended' it.

Sorry that's Def-Stan-0055/56 that has the recommendations

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

"Bernd Specht" <Bernd.Specht@gmx.com> wrote in message
news:Xns94E678C765E8CBerndSpechtgmxcom@151.189.20.10...
> "Martin Dowie" <martin.dowie@btopenworld.com> wrote in
> news:c7onju$smf$1@titan.btinternet.com:
>
> > There are GNAT ports to AVR microcontrollers (see
> > http://sourceforge.net/projects/avr-ada).
>
> 1. We work on 8051 and HC11, there is _no_ reason to switch to other
> processors (especially the programming language is no reason).
>
> 2. For professional development, it's not enough to have a compiler. You
> need a complete tool-chain (e.g. configurator, debugger ...)
>
> > But compiler vendors aren't going to support targets for which they see
> > no demand - so if you seriously want an Ada compiler for such things
> > ASK! :-)
>
> No, thats wrong. Would you go to GM and tell them: "Please build me a car
> whith five wheels,60 inch wheels, ..."? NO, you go to your local dealer,
see
> what he has and then buy it or leave it.

That's wrong in your (and perhaps even my) view - but that is the reality of
the situation with Ada (or any other non-C/C++ language) compilers.

Think of it more providing marketing information. One or two requests or
enquiries about support for a particular platform won't do anything but if
they get requests from tens of potential customers they then now there is a
market there to exploit.

-- Martin

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

berndtrog@yahoo.com (Bernd Trog) wrote in news:cbdd91ae.0405111134.5b2a75c5
@posting.google.com:

> Bernd.Specht@gmx.com (Bernd Specht) wrote in message 
> 
>> 1. We work on 8051 and HC11, there is _no_ reason to switch to other 
>> processors (especially the programming language is no reason).
> 
> Try http://gel.sourceforge.net/ada_example.php
> It's a cross compiler for 68HC11 and 68HC12.


1. It runs on Linux. Linux is still no acceptable platform for us. Maybe, we 
will give it a chance if there is Windows version.

2. It lacks a few good features of Ada. Exception handling for example.

3. Compare it with available C/C++ environments (e.g. Metroworks for HC11). 
Is there a  quick and easy  installation procedure? Is there a GUI? What 
about a debugger?

> Use an Ada-to-C compiler for 8051.

Why not directly C?

Think about that: we do  not get paid for using a nice programming language, 
we earn our money with writing working applications (regardless what 
language we use).

Re: Ada used in General Aviation (GA) applications?

[Richard Riehle]

> You could ask, but the answer is "No - but we'll let you pay to develop
> it and you can wait a year to get it." You're absolutely right on this
> score. *NOBODY* is going to pay a vendor and wait for the privilege of
> using Ada. They're going to select their board/chip and they're going to
>   use whatever compilers/development environments are available for that
> hardware. If Ada isn't sitting on the shelf waiting to be bought, then
> Ada doesn't play.

Ada compiler publishers have been dwelling in the DoD world for such
a long time that any tendency toward entrepreneurialism has vanished.
I talked to one DoD contractor's management a few years ago and
suggested they had an opportunity to enter a marketplace (not with
Ada, but other skills) and become successful in that market.  Their
response, "But who is going to fund it?"

There is a risk associated with any new venture.   Entrepreneurs tend
to be driven by their vision, their dedication to that vision, and their
unwavering confidence that it is a vision with commercial value.  At
one time, before the wimps took charge, Aonix seemed almost ready
to pursue that kind of vision.

Among those who have grown up in the DoD contracting world, courage
seems in short supply.  Some of the people who build tools for Ada
don't use their own Ada compilers and tools for building their other
products.    I don't need to identify those companies.  They know who
they are.  However, they don't have a sense of how ashamed they
should be of their lack of vision, lack of courage, and lack of
entrepreneurial will.

"We will only build a product if someone asks for it," is not a particularly
good business strategy.   One can keep a company alive, for a while, using
that strategy, but it is not a sustainable posture.   Entrepreneurs are
constantly
trying to find new markets, not simply cling to existing ones.

There have been a few examples of risk takers in the Ada industry.  RR
Software comes to mind.  Meridian comes to mind.   OC Systems.
There are a few others.  The fact that these have not been a resounding
success has acted as a deterrent for others.   One of the companies that
had a chance to make Ada popular, Rational, flubbed that opportunity.
I don't know if it was inept management,  lack of courage, or just being
too busy to put any energy into seeking commercial success for Ada.
Whatever it was, the language product that got them started, Ada,
seems to have vanished from Rational's main marketing thrust.

The only companies that actively and energetically market their Ada
products, at present, seem to be ACT and DDC-I.  Every month,
I get a newsletter from DDC-I that updates me on what they are
doing, customer news, product news, and even a little column by
a guest writer.  No other Ada compiler publisher takes the trouble
to do anything like that.   If you want to be on their distribution, I
suggest you send  them an email.  They are quite easy to deal
with on such matters.

Marin is correct when he suggests that the only way to make Ada
successful is for people to being creating products that use it.  We
can whine about the fact that more people are not choosing it, we
can complain about the stupidity of the LM management on JSF,
we can wring our hands about the downside of abrogating the
mandate.   None of that is worth much.    What is worth a lot
is for those who know and love Ada to build commercial products
using it.  Sell your shrink-wrapped Ada application to the general
marketplace.  Let people know you used Ada for development.
Once we have those kind of successes, Ada can stand on its own
and will be recognized for the value it actually provides.

Richard Riehle

Re: Ada used in General Aviation (GA) applications?

[Steve]

"Bernd Specht" <Bernd.Specht@gmx.com> wrote in message
news:Xns94E678C765E8CBerndSpechtgmxcom@151.189.20.10...
[snip]
>
> No, thats wrong. Would you go to GM and tell them: "Please build me a car
> whith five wheels,60 inch wheels, ..."? NO, you go to your local dealer,
see
> what he has and then buy it or leave it.
>

I strongly disagree with this statement, it is also inconsistant with my
experience.

In my experience successful vendors listen very carefully to their customers
and adapt to their wants and needs.  These companies know to listen very
carefully because very few customers give feedback.

If GM were to start getting a bunch of requests for cars with five 60 inch
wheels, they would be much more likely to look into the market for such a
car than if they had received no requests.

If an Ada vendor were to receive multiple requests for a compiler targeting
a specific processor, it would be silly for them to not at least consider
targeting that processor.

Steve
(The Duck)

RE: Ada used in General Aviation (GA) applications?

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: Bernd.Specht@gmx.com [mailto:Bernd.Specht@gmx.com]
...
| > Use an Ada-to-C compiler for 8051.
| 
| Why not directly C?
Because it cost 60% less to developp and because the result will contain 8
time less bugs with a comparable test effort. Studies showing that are on
the net.

If you develop a small software, that will almost never change, and be
embedded in 10 000 cell phones, then recurrent cost (harware for example)
are much more important than non reccurent cost (software developement for
example). This usually result in choosing the cheaper CPU whitout
considering software tools.

But even in this case, this is something to evaluate. Using Ada will result
in:
1 - cheaper developpement
2 - more reliable software
3 - better evolutivity (important in product line context, like cell phones)
4 - better portability

You should evaluate the cost of settling an Ada compilation environment for
your target, and balance it with those points.

-- 
Lionel Draghi

Re: Ada used in General Aviation (GA) applications?

[Peter Amey]

Bernd Specht wrote:
> Marin David Condic <nobody@noplace.com> wrote in
> news:409F69CB.8020604@noplace.com: 
> 
>  
> 
>>If Ada wants to be a player in General Aviation avionics it needs to do
>>something to either a) target & capture some more "general" market that
>>will provide the $$$ that will build the tools/talent/support
> 
> 
> 
> There is a wide range of embedded applications on 8- and 16-bit processors 
> like 8051 and HC11. But - no (Ada-)compiler vendor offers a compiler for 
> such targets.
> 

There might not be general Ada solutions for these processors but there 
a SPARK one.  Because SPARK is designed to require little or no run-time 
library support it is possible to design and analyse in SPARK and then 
compile to C en route to running on small processors.  The translation 
is simple because SPARK in unambiguous and because issues like array 
bounds violation have all been dealt with and eliminated at the SPARK 
design level.  I am presenting a paper on this at Ada Europe this year.

Peter

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Richard Riehle wrote:
 >
 > "We will only build a product if someone asks for it," is not a
 > particularly good business strategy.   One can keep a company alive,
 > for a while, using that strategy, but it is not a sustainable
 > posture.   Entrepreneurs are constantly trying to find new markets,
 > not simply cling to existing ones.
 >

 From what I can see, most of the Ada vendors are sitting in the
background with Ada and selling what they can sell, but looking to other
ventures as their future. I can sympathize with the attitude of "We're
driven by what our customers want..." but here's the flaw: It keeps your
ever dwindling customer base happy with you while finding ABSOLUTELY NO
*NEW* business. (Unless, of course, you're getting out of the Ada 
business - which then makes you irrelevant to the discussion at hand.)

How many people currently use Ada in real-world development jobs? 1%?
5%? Whatever it is - its *not* a big number. What is Ada going to do to
try to persuade the 95%-99% of the developers out there that they should
use Ada? All the arguments we've been hearing for years about long term
costs and reliability and such just have *not* sold. So the vendors and
the Ada fans can all go home grumbling about how stupid the general
computing public is and watch Ada die, or they can try to get
*innovative* and find a way to attract some *NEW* business.





 > There have been a few examples of risk takers in the Ada industry.
 > RR Software comes to mind.  Meridian comes to mind.   OC Systems.
 > There are a few others.  The fact that these have not been a
 > resounding success has acted as a deterrent for others.

Some of it is an issue of resources. It can take a lot of cash to
promote a technology. But if individual entrepreneurial companies can't
do it on their own, they can consider that a rising tide raises all
ships and band together to take some joint action. Go off and develop
some strategy for getting Ada adopted in a more lucrative area and then
stay focused on that strategy. I don't know why the vendors don't get
together and pool some resources to try to advance Ada technology
because at this point its going down the tubes and going to take them
*all* out of business. Perhaps they just figure they're migrating to
selling other technology and don't really care if Ada has any future.





 >
 > Marin is correct when he suggests that the only way to make Ada
 > successful is for people to being creating products that use it.  We
 > can whine about the fact that more people are not choosing it, we can
 > complain about the stupidity of the LM management on JSF, we can
 > wring our hands about the downside of abrogating the mandate.   None
 > of that is worth much.    What is worth a lot is for those who know
 > and love Ada to build commercial products using it.  Sell your
 > shrink-wrapped Ada application to the general marketplace.  Let
 > people know you used Ada for development. Once we have those kind of
 > successes, Ada can stand on its own and will be recognized for the
 > value it actually provides.
 >


Amen! Amen! and Amen! Every time I hear someone here blaming it all on
the "stupidity of managers" or "DoD Contractor greed" or "The hacker 
mentality that likes debugging" or any number of other excuses, I 
cringe. We could debate the truth or falsity of those excuses forever. 
Let's assume they're all 100% true. How does that get Ada adopted 
anywhere that matters? How many Ada programming jobs does that create? 
If its all true and the conclusion is that its all hopeless, then let's 
just put Ada in a coffin, bury her in the ground and get off this 
newsgroup and go talk about something that has a future.

If nobody does anything entrepreneurial with Ada, then there are no Ada 
jobs, students will consider learning Ada a waste of their time, big 
companies like LockMart will see Ada as a niche language that nobody 
uses and not worth their time/money to adopt, managers will perceive it 
as too big a risk and too high a cost to use on their projects, Ada 
vendors have nobody to sell support contracts to, etc., etc., etc.

The reason Ada is dying is because nobody is off building 
entrepreneurial products with it. All the "free" software in the world 
will not make Ada a hit if there isn't somewhere along the line some 
revenue with which to cut paychecks and purchase Ada vendor support, 
right? So dream up an end product that might have a market and in which 
Ada can play a role in making it a success. When that product is selling 
and generating revenue, Ada gets some life breathed back into it.

MDC


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

It doesn't hurt to ask, but the vendors ought to be doing their own 
market research. If Vendor X thinks that they want to sell embedded 
compilers, they ought to a) find out what the popular processors are in 
the market of interest, b) ask the people who use those processors if 
they would have any real interest in using Ada if it were available, c) 
ask them what else they'd need to see in the way of support tools in 
order to close a deal and d) invest in building what the potential 
customers want to buy.

Waiting for someone to come knocking at your door begging you to please 
sell them a product is a sure fire way to go out of business quickly. I 
hope the vendors aren't counting on that.

BTW: I don't think that Ada has much of a chance of gaining huge success 
(by which I mean revenue) in trying to target small embedded computers. 
It would cost a lot to do and the potential buyers don't buy millions of 
compilers. That, and they are enormously disinterested in Ada. Better to 
get some other market segment that might be easier to address and let it 
spill over into embedded programming.

MDC


Martin Dowie wrote:
> 
> Think of it more providing marketing information. One or two requests or
> enquiries about support for a particular platform won't do anything but if
> they get requests from tens of potential customers they then now there is a
> market there to exploit.
> 
> -- Martin
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Waiting for someone to come to your door and beg you to make them a 
product will put you out of business. GM, et alia, spend huge amounts of 
money building "concept cars" and doing market research to try to figure 
out what customers will want to buy. And not just their *existing* 
customers either. They go ask people who *don't* buy their product to 
figure out how to make something that will get them to hand over some 
cash too.

"We build whatever our existing customer's ask for..." is a recepie for 
how to go out of business over time.

MDC

Steve wrote:
> 
> I strongly disagree with this statement, it is also inconsistant with my
> experience.
> 
> In my experience successful vendors listen very carefully to their customers
> and adapt to their wants and needs.  These companies know to listen very
> carefully because very few customers give feedback.
> 
> If GM were to start getting a bunch of requests for cars with five 60 inch
> wheels, they would be much more likely to look into the market for such a
> car than if they had received no requests.
> 
> If an Ada vendor were to receive multiple requests for a compiler targeting
> a specific processor, it would be silly for them to not at least consider
> targeting that processor.
> 
> Steve
> (The Duck)
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

We could debate the exact numbers forever. I wasn't claiming that this 
was an exact representation but rather illustrative of how an industry 
works. I used to program digital TV boxes and the company sold *lots* of 
units. But the staff of people actually programming the boxes was 
relatively small (and getting smaller - which is why I left! :-) The 
point being that just because there might be a lot of HC11 processors 
sold, it doesn't follow that there will be lots of HC11 Ada compilers sold.

If someone wanted to push Ada in the embedded market, I'd think the best 
way to do it would be to actually be in the board business. Design a 
board for a specific part of the market, target Ada to it, provide a 
whole suite of Ada related development tools to make the job of 
programming it a snap and then sell boards/development kits. "Sure, you 
could use C to program my spiffy little board, but it would take you 
longer because all the critical leverage items are all Ada related..."

MDC

Bernd Specht wrote:
> 
> Well, if there were someone who sells *millions* on  HC11/8051 products and 
> needs only *one* compiler, then I would agree. But - in this range it is 
> often the case, that a small company sells a few hundrets or thousands 
> units, but *every* developer has its own develop environment. So to sell 
> *millions* of units, there would be a lot of companies, who need a lot of 
> development tools (_not only_ compiler, but the _whole_ toolchain). But - 
> toolset must not be expensive (if one sells a few thousand boards, they 
> won't be willing to pay $1000_000 for the tools).
> 
> 

-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

The "Compile Ada to C then compile C to the target..." approach has been 
enormously popular and successful, hasn't it? :-)

Nobody wants to use two compilers, nobody wants to cobble the whole mess 
together and nobody wants to fight all the rest of the development tools 
that will be targeted to C. The whole argument amounts to "Go out and 
spend lots of extra money, take lots of extra time and go through lots 
of extra pain in order to have the *privilege* of programming your 
little board in Ada - all for a job that might only involve a couple of 
guys for a few months." It has not sold well and I doubt it ever will. 
Ada is either right there on the shelf with an embedded development kit 
or the embedded guys go elsewhere.

MDC


Peter Amey wrote:
> 
> There might not be general Ada solutions for these processors but there 
> a SPARK one.  Because SPARK is designed to require little or no run-time 
> library support it is possible to design and analyse in SPARK and then 
> compile to C en route to running on small processors.  The translation 
> is simple because SPARK in unambiguous and because issues like array 
> bounds violation have all been dealt with and eliminated at the SPARK 
> design level.  I am presenting a paper on this at Ada Europe this year.
> 
> Peter
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

That assumes "All Other Things Being Equal". They never are. Especially 
in the embedded world.

MDC

Lionel.DRAGHI@fr.thalesgroup.com wrote:

> Because it cost 60% less to developp and because the result will contain 8
> time less bugs with a comparable test effort. Studies showing that are on
> the net.
> 
> If you develop a small software, that will almost never change, and be
> embedded in 10 000 cell phones, then recurrent cost (harware for example)
> are much more important than non reccurent cost (software developement for
> example). This usually result in choosing the cheaper CPU whitout
> considering software tools.
> 
> But even in this case, this is something to evaluate. Using Ada will result
> in:
> 1 - cheaper developpement
> 2 - more reliable software
> 3 - better evolutivity (important in product line context, like cell phones)
> 4 - better portability
> 
> You should evaluate the cost of settling an Ada compilation environment for
> your target, and balance it with those points.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

RE: Ada used in General Aviation (GA) applications?

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: Marin David Condic [mailto:nobody@noplace.com]
...
| That assumes "All Other Things Being Equal". They never are. 
| Especially in the embedded world.

Take those other things into account in the balance.

I notice that this argument is used more and more, and most of the time by
those underestimating language influence on developement cost.
Just ask yourself what kind of "other things being not equal" is not a
detail compared to cutting down half developpement time or dividing problem
after release by 8 or shrinking fix time per bug by two.

-- 
Lionel Draghi

Re: Ada used in General Aviation (GA) applications?

[Georg Bauhaus]

Marin David Condic <nobody@noplace.com> wrote:
: 
: Nobody wants to use two compilers, nobody wants to cobble the whole mess 
: together and nobody wants to fight all the rest of the development tools 
: that will be targeted to C.

There are quite a few languages using C as a target language,
and then having that compiled automatically. Among them you will
find Eiffel, and, nota bene, C++ (by Comeaucomputing). You won't
have to cobble anything together with these compilers, and you
are asked not to try.


-- Georg

Re: Ada used in General Aviation (GA) applications?

[Robert C. Leif]

The US DoD has one major technology transfer success. They are
extremely capable of dropping a bomb on someone. Unfortunately, this
very limited success does NOT appear to benefit the US economy. In
fact, part of the Clinton boom was the result of the transfer of some
very bright creative individuals from military to civilian
entrepreneurial projects. The Ada paradox is that the Ada marketing
capability is inversely proportional to the Ada technology capability.

For 2004 we will have support from Microsoft of SIGAda, which is
wonderful. Unfortunately, we have minimal support for creation of
real-time code with Microsoft operating systems, particularly Windows
CE and embedded XP.  One of the great strengths of Ada is low level
coding.  At Microsoft's own embedded show in San Diego, their experts
could not tell me how to write code for a board!  A# has reached the
feasibility stage.  Unfortunately, it is an academic exercise and not
a commercial product. It has demonstrated that existing J code
production software can be modified to produce the ECMA intermediate
language.  Languages with much smaller markets than Ada have now been
hosted to produce ECMA code and operate under Microsoft Visual Studio.
 One should compile directly to the Microsoft intermediate language
and then use the Microsoft tools to run on all of the major
microprocessors. If for some reason, a microprocessor does not work
with Microsoft tools, a language independent product that permits
Microsoft tools to be used with that processor should be developed.
This language independent tool should have a better return on capital
than retargeting an Ada compiler.

Two old proverbs are relevant: 1) If you can not beat them join them.
2) Render unto Bill Gates (Caesar) what is Bill Gates'.

As far as work on other languages, I agree that the Ada compiler
vendors have undercut their own products. Gucci does not sell to
Walmart. However, there is one market where they should be competing,
XML.  XML is parallel to Ada and does not compete. It also a larger
market than that for compilers. I use XMLSpy to validate XML schemas. 
The error messages from GNAT are orders of magnitude better.  An Ada
XForms generator would be competitive. An XForms generator that is a
plug-in for Word could be very profitable. Even, an Ada plug-in for
Xforms would be a useful product.

As far as commercialization is concerned, ASIS provides the capability
of creating tools to divide up the royalties on software.  This would
permit the development of for profit, commercial, sources available
software, which is the logical evolutionary step beyond "Free
Software".

Bob Leif
 

"Richard  Riehle" <adaworks@earthlink.net> wrote in message news:<tVdoc.16693$V97.6543@newsread1.news.pas.earthlink.net>...
> Ada compiler publishers have been dwelling in the DoD world for such
> a long time that any tendency toward entrepreneurialism has vanished.
> I talked to one DoD contractor's management a few years ago and
> suggested they had an opportunity to enter a marketplace (not with
> Ada, but other skills) and become successful in that market.  Their
> response, "But who is going to fund it?"
> 
> There is a risk associated with any new venture.   Entrepreneurs tend
> to be driven by their vision, their dedication to that vision, and their
> unwavering confidence that it is a vision with commercial value.  At
> one time, before the wimps took charge, Aonix seemed almost ready
> to pursue that kind of vision.
> 
> Among those who have grown up in the DoD contracting world, courage
> seems in short supply.  Some of the people who build tools for Ada
> don't use their own Ada compilers and tools for building their other
> products.    I don't need to identify those companies.  They know who
> they are.  However, they don't have a sense of how ashamed they
> should be of their lack of vision, lack of courage, and lack of
> entrepreneurial will.
> 
> "We will only build a product if someone asks for it," is not a particularly
> good business strategy.   One can keep a company alive, for a while, using
> that strategy, but it is not a sustainable posture.   Entrepreneurs are
> constantly
> trying to find new markets, not simply cling to existing ones.
> 
> There have been a few examples of risk takers in the Ada industry.  RR
> Software comes to mind.  Meridian comes to mind.   OC Systems.
> There are a few others.  The fact that these have not been a resounding
> success has acted as a deterrent for others.   One of the companies that
> had a chance to make Ada popular, Rational, flubbed that opportunity.
> I don't know if it was inept management,  lack of courage, or just being
> too busy to put any energy into seeking commercial success for Ada.
> Whatever it was, the language product that got them started, Ada,
> seems to have vanished from Rational's main marketing thrust.
> 
> The only companies that actively and energetically market their Ada
> products, at present, seem to be ACT and DDC-I.  Every month,
> I get a newsletter from DDC-I that updates me on what they are
> doing, customer news, product news, and even a little column by
> a guest writer.  No other Ada compiler publisher takes the trouble
> to do anything like that.   If you want to be on their distribution, I
> suggest you send  them an email.  They are quite easy to deal
> with on such matters.
> 
> Marin is correct when he suggests that the only way to make Ada
> successful is for people to being creating products that use it.  We
> can whine about the fact that more people are not choosing it, we
> can complain about the stupidity of the LM management on JSF,
> we can wring our hands about the downside of abrogating the
> mandate.   None of that is worth much.    What is worth a lot
> is for those who know and love Ada to build commercial products
> using it.  Sell your shrink-wrapped Ada application to the general
> marketplace.  Let people know you used Ada for development.
> Once we have those kind of successes, Ada can stand on its own
> and will be recognized for the value it actually provides.
> 
> Richard Riehle

Re: Ada used in General Aviation (GA) applications?

[Mike Silva]

Bernd.Specht@gmx.com (Bernd Specht) wrote in message news:<Xns94E6E7A0A5A51BerndSpechtgmxcom@151.189.20.10>...
 
> > Use an Ada-to-C compiler for 8051.
> 
> Why not directly C?
> 
> Think about that: we do  not get paid for using a nice programming language, 
> we earn our money with writing working applications...
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Just add "for the least amount of time and money expended" and perhaps
you've answered your own question.

Re: Ada used in General Aviation (GA) applications?

[Mike Silva]

Peter Amey <peter.amey@praxis-cs.co.uk> wrote in message news:<2geb04F1pte3U1@uni-berlin.de>...
 
> There might not be general Ada solutions for these processors but there 
> a SPARK one.  Because SPARK is designed to require little or no run-time 
> library support it is possible to design and analyse in SPARK and then 
> compile to C en route to running on small processors.  The translation 
> is simple because SPARK in unambiguous and because issues like array 
> bounds violation have all been dealt with and eliminated at the SPARK 
> design level.  I am presenting a paper on this at Ada Europe this year.
> 
> Peter

I think this is a _very_ exciting approach and I hope it catches on.  Good luck!

Mike

RE: Ada used in General Aviation (GA) applications?

[Bernd Specht]

Lionel.DRAGHI@fr.thalesgroup.com wrote in
news:mailman.114.1084352783.313.comp.lang.ada@ada-france.org: 

> 
> 
>| -----Message d'origine-----
>| De: Bernd.Specht@gmx.com [mailto:Bernd.Specht@gmx.com] ...
>| > Use an Ada-to-C compiler for 8051.
>| 
>| Why not directly C?
> Because it cost 60% less to developp and because the result will
> contain 8 time less bugs with a comparable test effort. Studies showing
> that are on the net.


60% less???

Throw away current develop-tools.
Throwing away the staffs know how.
Start from scratch.
Write Ada to generate C, instead the employees writing C-code.

60% less??? Maybe over the next 100 years. But till then the company is 
dead.

Our business is to make profit. You are an academician?

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

 (Bernd Specht) writes:
> wrote in
> 
> 
> > 
> > 
> >| -----Message d'origine-----
> >| De: ...
> >| > Use an Ada-to-C compiler for 8051.
> >| 
> >| Why not directly C?
> > Because it cost 60% less to developp and because the result will
> > contain 8 time less bugs with a comparable test effort. Studies showing
> > that are on the net.
> 
> 
> 60% less???
> 
> Throw away current develop-tools.

If you use an Ada-to-C compiler, you will probably not have to throw
away many tools.  You can keep your C compiler, configuration
management system, editor, documentation system, etc.

> Throwing away the staffs know how.

This would be a silly mistake.  Most of that know-how is in the
problem space, in this case avionics; not in the particular
implementation language.

> Start from scratch.

Again, no.  You can reuse your tested and proven C code from new Ada
programs.  And all the business-specific knowledge is reusable.

> Write Ada to generate C, instead the employees writing C-code.

Yes, this is a good approach, because C code generated from Ada has
fewer bugs than hand-written C code; just like assembler generated
from C has fewer bugs than hand-written assembler.

> 60% less??? Maybe over the next 100 years. But till then the company
> is dead.

No.  Studies show that the return on investment is quite rapid.  The
time it takes to learn Ada is negligible compared to the time it takes
to learn avionics or any other business.  In fact, the time spent
writing the code is a fraction of the time it takes to debug, test and
validate the code.  This is where Ada pays for itself many times over:
it cuts into the debugging, testing and validation time.

> Our business is to make profit. You are an academician?

If this is the case, then you should be prepared to use
state-of-the-art, not second-choice languages; just like you should
generally use the tools most appropriate for your job.

-- 
Ludovic Brenta.

Re: Ada used in General Aviation (GA) applications?

[Mark Lorenzen]

Ludovic Brenta <ludovic.brenta@insalien.org> writes:

[snip]

>> Throwing away the staffs know how.
>
> This would be a silly mistake.  Most of that know-how is in the
> problem space, in this case avionics; not in the particular
> implementation language.

Good point! I really, really hope for the OP that the staff know-how
is in the problem space. This is a very good counter-argument for the
all-recurring "but all our developer know C/C++".

[snip]

>
> -- 
> Ludovic Brenta.

- Mark Lorenzen

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

snarflemike@yahoo.com (Mike Silva) wrote in
news:20619edc.0405120909.6ba1a793@posting.google.com: 

> Bernd.Specht@gmx.com (Bernd Specht) wrote in message
> news:<Xns94E6E7A0A5A51BerndSpechtgmxcom@151.189.20.10>... 
>  
>> > Use an Ada-to-C compiler for 8051.
>> 
>> Why not directly C?
>> 
>> Think about that: we do  not get paid for using a nice programming
>> language, we earn our money with writing working applications...
>   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> Just add "for the least amount of time 

Whats about the time to train the staff?
Whats about the time to get familiar with the new tools?
Whats about the time to get familiar with the new language?
Is it economic to throw away the whole know how?

> and money expended" 

Whats about the money for the new tools?
Whats about the money for training?


Time is money - the time you lose due to the change is _your_ money you 
lose. Do _YOU_ really think _this_ is a benefit???


> you've answered your own question.

Take you time andf think about. Maybe you find your errors in your 
calculation (hope you are not responsible for your companies health).

Re: Ada used in General Aviation (GA) applications?

[Pascal Obry]

Bernd.Specht@gmx.com (Bernd Specht) writes:

> Whats about the time to train the staff?

That's the whole point. This will almost count for nothing if your staff
has good computer science skills. Learning a new language is not a problem
for skilled people and we know that the gain will be there with Ada.

But if your staff can't understand something more high level that C, then I
agree with you.

Pascal.

-- 

--|------------------------------------------------------
--| Pascal Obry                           Team-Ada Member
--| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE
--|------------------------------------------------------
--|              http://www.obry.org
--| "The best way to travel is by means of imagination"
--|
--| gpg --keyserver wwwkeys.pgp.net --recv-key C1082595

Re: Ada used in General Aviation (GA) applications?

[Richard Riehle]

"Bernd Specht" <Bernd.Specht@gmx.com> wrote in message
news:Xns94E678C765E8CBerndSpechtgmxcom@151.189.20.10...
> "Martin Dowie" <martin.dowie@btopenworld.com> wrote in
> news:c7onju$smf$1@titan.btinternet.com:
>
> > There are GNAT ports to AVR microcontrollers (see
> > http://sourceforge.net/projects/avr-ada).
>
> 1. We work on 8051 and HC11, there is _no_ reason to switch to other
> processors (especially the programming language is no reason).

I have a student at NPS who is enthusiastic about Ada.  He actually
likes it.  His thesis project is related to the HC11.  Guess what
language he is forced to use instead of Ada.   Interesting thing is
that this project involves a large U.S. Navy software system. However,
he will not be able to write his code in his preferred language.

This is happening quite a bit.  I teach Ada, as part of a class called
Programming Paradigms.  Included in this class, I also teach C,
C++, Eiffel, Lisp, and touch on several other languages (e.g., one
lecture on Smalltalk).    Some of my students really like Ada. Most
of the professors who are thesis advisors don't like Ada, so the
refuse to let students use it.  Still, some professors are a little
more open about such things.   When the thesis advisor says,
"OK. Use Ada if you wish," and there is no compiler available,
we lose another opportunity to make our case.

I suppose this is my fault for not taking the time (I have so much
of it) to port Ada to these other platforms myself.   But my time
is spent reading programming assignments, grading exams, and
counseling students (which is a lot of time, it turns out), and there
just isn't time to do the fun things such as port GNAT to the
interesting platforms we might like to see it on.

As Marin has observed, as long as the compiler publishers are
operating on the basis of, "Our customers don't ask for that
capability," these other platforms are unlikely to be supported.
No one can blame them for that attitude.   Most of them are
small, understaffed, and understand that Ada is no longer a
"cash cow."    ACT has gone out on a limb with JGNAT, and
suffered for it - no sales.  Aonix, when it was still an Ada company,
took some risks with ObjectAda.   RR Software has always put
itself at risk, but I suspect its compiler sales are not at an all-time
high.   Rational stopped actively marketing Ada a long time ago
and seems to be in a holding pattern with the language.

Those of us who are Ada advocates have only ourselves to blame.
Had we put the energy into building commercial products instead
of relying on the DoD those many years ago, we would be much
further ahead today.   A few companies did take the risk.  In some
cases, they did well.   Others have vanished, not because of their
choice of Ada, but because of market forces.

At this point, we cannot depend on enlightened management at
DoD contractors.  It is a scarce commodity in that domain (with
some remarkable exceptions within LM and elsewhere).   Most
software developers are moving on to the next fad even as we
debate this issue.  Remember when people were saying, seven
years ago, "The language wars are over and C++ won."  Well,
they are not over.  Java and C# are gaining a large following,
perhaps larger than C++.

At present, Ada lacks a committed consortium (ARA, etc. seems
pretty ineffectual), a corporate sponsor, any government support,
and no coordinate approach to marketing and promotion.  It is
now a loose association of enthusiasts who hang on to the hope
that the resurrection is near, much like a bunch of millenial zealots.
While such faith is admirable, we must keep in mind the admonition,
"Faith without works is dead."

The Ada enthusiast at LM is not going to persuade the new manager
who prefers C++ of the benefits of Ada.  The Ada expert who did
so well on his last project at Raytheon will not turn the tide back
toward Ada when so many there have lost sight of why it was a
good idea for earlier systems.   The only hope is for those experts
to identify an marketing opportunity (even band together) and use
Ada for a new product.    Give up the security of you Boeing
pension, your LM health benefits, and launch a new product
using Ada.   Put you belief in Ada to the test.   That is the only
way Ada will survive over the next five or ten years.

Richard Riehle

Re: Ada used in General Aviation (GA) applications?

[Vinzent 'Gadget' Hoefler]

Bernd Specht wrote:

>Whats about the time to train the staff?

If they're already programmers at worst two weeks to get most of the
concepts.

Less if you only need the concepts they already have in C, but that's
a bad idea. ;)

>Whats about the time to get familiar with the new tools?

A couple of days. Hey, they don't have to write makefiles anymore. :)

>Whats about the time to get familiar with the new language?

Well, that takes longer (just like in C). But it's a parallel process.
You can do that while already writing code.

(And BTW, I still don't get it. Writing code is a very small part of a
whole development process. A much bigger part is testing and
debugging. And that's where Ada cuts a lot.)

>Is it economic to throw away the whole know how?

What? You don't do that. The know-how isn't just syntax. The know-how
is what they do. The system. Not just some *language*.

I still set the same bits in the same register, no matter if I do it
in Assembler, C or Ada. But it took me about half a year to understand
the parts of the whole system. *This* is the know-how. And that
doesn't get *woosh* just because I switch to another language.

>Whats about the money for the new tools?

GNAT is free, if you're not relying on professional support.

>Whats about the money for training?

Some books. They're not *that* expensive. And hey, learning Ada is
*fun*!

(BTW, currently at my company we all here are forced to learn Java,
obviously someone seems to think this is a benefit.)

>Time is money - the time you lose due to the change is _your_ money you 
>lose. Do _YOU_ really think _this_ is a benefit???

Clearly, yes.

For instance, I am currently writing on an embedded system, it's a
complete rewrite of something that already exists in Assembler. If I'd
do that in C it would take *much* longer (and would have less features
just to compensate for that longer development time).

>Take you time andf think about. Maybe you find your errors in your 
>calculation (hope you are not responsible for your companies health).

Well. I am not responsible for the companies health, but WTH, I am
*interested* in it. They pay me for what I am doing. So, does that
count? :)


Vinzent.

-- 
To err is human, to forgive beyond the scope of the operating system.

Re: Ada used in General Aviation (GA) applications?

[Peter Amey]

Marin David Condic wrote:
> The "Compile Ada to C then compile C to the target..." approach has been 
> enormously popular and successful, hasn't it? :-)

Yes, considering we only came up with the SPARK -> C idea last year it 
has been rather popular :-)

> Nobody wants to use two compilers, nobody wants to cobble the whole mess 
> together and nobody wants to fight all the rest of the development tools 
> that will be targeted to C. The whole argument amounts to "Go out and 
> spend lots of extra money, take lots of extra time and go through lots 
> of extra pain in order to have the *privilege* of programming your 
> little board in Ada - all for a job that might only involve a couple of 
> guys for a few months." It has not sold well and I doubt it ever will. 
> Ada is either right there on the shelf with an embedded development kit 
> or the embedded guys go elsewhere.

There is some truth in what you say, the process has to be convenient. 
I think this hinders use of the approach for Ada in general.  The use of 
SPARK changes the gearing somewhat:

1.  You are achieving more than "the privilege of programming your 
little board in Ada", you are gaining the ability to do very high levels 
of validation much earlier in the development process.  Before you even 
get to C you can have eliminated use of unset variables, proved freedom 
from all run-time errors, shown that values remain in 
design-domain-defined ranges (regardless of what C predefined type they 
end up being implemented as), demonstrated separation of critical and 
non-critical information flows and even proved that certain important 
properties of the code hold.

2.  The translation process can be completely transparent.  As far as we 
are concerned on one project, we have what amounts to a SPARK to target 
compiler.  We don't need to process the C manually at all.

3.  If you are constrained to deliver something like MISRA-C then it is 
orders of magnitude easier to arrange for the translator to represent 
the exact SPARK design in something which is unequivocally MISRA-C than 
to demonstrate retrospectively that hand-crafted C conforms to the 
subset rules.

4.  For very critical applications, the C provides an intermediate 
representation between the SPARK and the object code and this makes 
object code verification much easier than it otherwise would be.  For 
example, array assignments at the Ada level are "unwrapped" at the C 
level.  This intermediate representation allows one very hard step: 
comparing Ada to machine code to be replaced by two much easier ones: 
Ada to (simple) C and C to machine code.

5.  You don't need, and therefore don't have to validate, a run-time 
library.

So far from having to "spend lots of extra money, take lots of extra 
time and go through lots of extra pain" we find the combination allows 
us to produce higher quality at lower cost than writing directly in C.

regards


Peter

Re: Ada used in General Aviation (GA) applications?

[Dmitry A. Kazakov]

On Wed, 12 May 2004 12:30:44 GMT, Marin David Condic
<nobody@noplace.com> wrote:

>BTW: I don't think that Ada has much of a chance of gaining huge success 
>(by which I mean revenue) in trying to target small embedded computers. 
>It would cost a lot to do and the potential buyers don't buy millions of 
>compilers.

I believe you are wrong here. In the area we are working, each large
vendor of an embedded application is surrounded by a swarm of smaller
firms to which it outsource software development and testing. Each
satellite firm has the complete development tool chain. These tools
are very expensive. It is a very lucrative market. [Some tool chain
vendors just take GCC, build loader, add a pair windows and sell that
for many k$.]

>That, and they are enormously disinterested in Ada.

True. To get there one has to offer better tools, a better compiler
would be not enough. Present tools are very bad, but the problem is
that the marked is saturated and closed. People there are conservative
and arrogant. It is difficult to get there, even if you have a better
product.

>Better to 
>get some other market segment that might be easier to address and let it 
>spill over into embedded programming.

No. It would take too long. I am afraid that Ada is missing the
embedded programming market which in the foreseeable future will
become no less important than the conventional one.

--
Regards,
Dmitry Kazakov
www.dmitry-kazakov.de

RE: Ada used in General Aviation (GA) applications?

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: Bernd.Specht@gmx.com [mailto:Bernd.Specht@gmx.com]
...

| 60% less???
Take a look a the studies on the net.
 
...
| Our business is to make profit. You are an academician?
This is also my business, i am not an academician. 
Instead of casting doubts on my own competences, just take a look at the
figures.

-- 
Lionel Draghi

Re: Ada used in General Aviation (GA) applications?

[Dmitry A. Kazakov]

On Wed, 12 May 2004 12:42:44 GMT, Marin David Condic
<nobody@noplace.com> wrote:

>If someone wanted to push Ada in the embedded market, I'd think the best 
>way to do it would be to actually be in the board business. Design a 
>board for a specific part of the market, target Ada to it, provide a 
>whole suite of Ada related development tools to make the job of 
>programming it a snap and then sell boards/development kits. "Sure, you 
>could use C to program my spiffy little board, but it would take you 
>longer because all the critical leverage items are all Ada related..."

Very true

--
Regards,
Dmitry Kazakov
www.dmitry-kazakov.de

Re: Ada used in General Aviation (GA) applications?

[Bernd Specht]

Vinzent 'Gadget' Hoefler <nntp-2004-05@t-domaingrabbing.de> wrote in 
news:bu76a05mpq39k8ouqo56ojgnct6g35ugug@jellix.jlfencey.com:

> Bernd Specht wrote:
> 
>>Whats about the time to get familiar with the new language?
> 
> Well, that takes longer (just like in C). But it's a parallel process.
> You can do that while already writing code.

Don't forget, the employees are real C-guys. You can't brainwash them at one 
afternoon.

> (And BTW, I still don't get it. Writing code is a very small part of a
> whole development process. A much bigger part is testing and
> debugging. And that's where Ada cuts a lot.)
> 
>>Is it economic to throw away the whole know how?
> 
> I still set the same bits in the same register, no matter if I do it
> in Assembler, C or Ada. But it took me about half a year to understand

Ok, but they write code for accessing a controller register in a few 
minutes, to write it Ada with all those rep specs they would  need much 
longer.
In our business the first prio is "time to market".

>>Whats about the money for the new tools?
> 
> GNAT is free, if you're not relying on professional support.

There is not GNAT for 8051/HC11.

>>Whats about the money for training?
> 
> Some books. They're not *that* expensive. And hey, learning Ada is
> *fun*!

If you give C-experts a book to learn Ada, then you will get programmers 
writing C-code in Ada syntax. A real training that they understand _and_ 
accept that they do not have to use pointers and pointzer arithmetic will 
take much more than  a book. 

> (BTW, currently at my company we all here are forced to learn Java,
> obviously someone seems to think this is a benefit.)

Well, it's not bad to know. 

>>Time is money 

A big company may live with a few developers on training for weeks (month). 
A small company has not the financial power to be able to close for some 
week and have no income (but costs) in that time.

End of "discussion" (was Re: Ada used in General Aviation (GA) applications?)

[Bernd Specht]

"Richard  Riehle" <adaworks@earthlink.net> wrote in news:buEoc.18396
$V97.11476@newsread1.news.pas.earthlink.net:

> 
> I suppose this is my fault for not taking the time (I have so much
> of it) to port Ada to these other platforms myself. 

Hey, wouldn't this be an good assignment for a student group? ;-)

> Those of us who are Ada advocates have only ourselves to blame.

There is a bigg problem: Most Ada-advocats are too conceived. Look at this 
thread, I said I would give Ada a chance if there were a version for our 
processors. What was the reply? "You are stupid if you don't switch over", 
"You don't do your job well if you work with C" ... (bla bla).
This really doesn't help.


Best is to leave this group and let Ada silent die.

RE: Ada used in General Aviation (GA) applications?

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: Bernd.Specht@gmx.com [mailto:Bernd.Specht@gmx.com]
...
| Whats about the time to train the staff?
| Whats about the time to get familiar with the new tools?
| Whats about the time to get familiar with the new language?
...
| Whats about the money for the new tools?
| Whats about the money for training?

Put this in the balance, the conclusion is not obvious.

Just some point here :
- Ada learning curve is generally over estimated, because people expect a
high level language to be mastered in about the same time than for C++. In
Ada, your are able to fix alone problem in existing code and do simple
addition very quickly.
 
- Ada benefits are generally underestimated, because even experienced people
don't trust the figures. They just expect some marginal benefits.

- in our experience, language and related tools training is negligible
compared to the the time to master the domain.

-- 
Lionel Draghi

Re: Ada used in General Aviation (GA) applications?

[Vinzent 'Gadget' Hoefler]

Bernd Specht wrote:

>Vinzent 'Gadget' Hoefler <nntp-2004-05@t-domaingrabbing.de> wrote in 
>news:bu76a05mpq39k8ouqo56ojgnct6g35ugug@jellix.jlfencey.com:
>
>> Bernd Specht wrote:
>> 
>>>Whats about the time to get familiar with the new language?
>> 
>> Well, that takes longer (just like in C). But it's a parallel process.
>> You can do that while already writing code.
>
>Don't forget, the employees are real C-guys. You can't brainwash them at one 
>afternoon.

True. But that's the two weeks I talked about before. :) Anything else
is learning by doing, damn it's just programming in another language.
(And if the guys are smart and want to deliver working code instead of
doing overtime while spending hours for debugging, this is not a
problem.)

>>>Is it economic to throw away the whole know how?
>> 
>> I still set the same bits in the same register, no matter if I do it
>> in Assembler, C or Ada. But it took me about half a year to understand
>
>Ok, but they write code for accessing a controller register in a few 
>minutes, to write it Ada with all those rep specs they would  need much 
>longer.

Would they? In general? The problem aren't the rep specs. Usually the
programming problem is much higher at some logic level. And Ada helps
there.

And I still don't see where the know how gets lost.

>In our business the first prio is "time to market".

I know, I know. It's all about "time to market". Sometimes it may help
the company to even deliver crap, when they only deliver it early.

But IMO especially in the embedded market this is a bad idea.
Unfortunately I am not running a company of my own, so I can't tell if
you will keep your customers when they get buggy systems but two weeks
earlier. ;-)

>>>Whats about the money for the new tools?
>> 
>> GNAT is free, if you're not relying on professional support.
>
>There is not GNAT for 8051/HC11.

That's where the problem lies. If you can't find an affordable
Ada-compiler for the architecture you are using, you're out of luck.
Sad but true. I can imagine, that you won't convince some hard-core
C-guys to use Ada if that Ada is then converted to C before it
compiles... (I can already hear them crying "But then we can write it
in C ourselves!").

>>>Whats about the money for training?
>> 
>> Some books. They're not *that* expensive. And hey, learning Ada is
>> *fun*!
>
>If you give C-experts a book to learn Ada, then you will get programmers 
>writing C-code in Ada syntax.

IBTD. IMO it mostly depends if they get a grasp on the basic concept
of typing instead of declaring everything as int.

>A real training that they understand _and_ 
>accept that they do not have to use pointers and pointzer arithmetic will 
>take much more than  a book. 

Oh, they can use pointers. But why should they? Parameter modes aren't
that hard to understand and this is a lot of where pointers are needed
in C.

>> (BTW, currently at my company we all here are forced to learn Java,
>> obviously someone seems to think this is a benefit.)
>
>Well, it's not bad to know.

Know thy enemy, yessssir. ;-)

>A big company may live with a few developers on training for weeks (month). 
>A small company has not the financial power to be able to close for some 
>week and have no income (but costs) in that time.

Well, of course. I am always forgetting that I am an Ada enthusiast
and so almost everything I know about Ada I learned in my spare time.
If you have guys who do "it's just my job from nine to five", you
won't get that...

So for me it was easier to decide, I just had to convince my boss that
I will use Ada for that system. And this part was relatively easy, he
doesn't like C either and Java was not a real option.

But *you* would have to convince a couple of developers who probably
never heard anything good about Ada...


Vinzent.

-- 
Wenn alle Stricke reissen, kann man sich immer noch erschiessen.

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

Bernd Specht wrote:
> A big company may live with a few developers on training for weeks
> (month).  A small company has not the financial power to be able to
> close for some week and have no income (but costs) in that time.

There are ways to spread the investment (not cost) of training over
time and still have some revenue during training.  As a manager, you
certainly know how you can have one developer at a time invest time
learning new tools between phases of a project.

Perhaps you can even convince your customers to share this investment
with you.  For example, one customer may agree to allow a project to
slip by a couple of weeks without penalty, with the understanding that
ROI will be on the next project in the form of several weeks gained on
the schedule, and fewer bugs.  In particular, the DO-178B
certification process is greatly facilitated by use of Ada; surely
customers are interested in expediting this process.

Also, I am a bit worried by your statement.  Does this imply that your
company does no proactive research?  In this case, how can you be
innovative?  And if you are not innovative, how can you survive in
avionics?

-- 
Ludovic Brenta.


-- 
Use our news server 'news.foorum.com' from anywhere.
More details at: http://nnrpinfo.go.foorum.com/

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Lots of people here seem to lack much understanding of or sympathy for 
"Time To Market" - perhaps because they come from industries where this 
is not a big pressure. But you're right - in lots of business sectors 
(the ones that generate money for compiler vendors) time to market is 
*everything*. All the maintenance costs and long term bug problems in 
the world pale in comparison to "Time To Market".

So a company that is, for example, in the business of making some kind 
of controller board and programming it would look at a new project this 
way: We've got a new board similar to what we've been using all along. 
We have software that does some significant percentage of the job 
already written for existing products. We got a C compiler and debugger 
and other support tools sitting right here next to the new board ready 
to go. Our guys can get started programming this thing *today* and have 
it ready (including excess debugging time) in 3 or 4 months.

What? You'd like me to use Ada? But first I've got to go find an Ada 
compiler targeted to my little controller board, right? And then maybe 
I'm going to have to cobble it together with some C compiler? Along the 
way I get to worry about efficiency and if all this 
Ada-to-C-to-Machine-Code is going to generate good enough code to do the 
job. And then all my existing tools don't work right because they're all 
hooked to C in some manner, right? So I've got to fix them or live 
without them. With whatever existing code I still want to bring along, 
I've got to stop and write some sort of Ada binding to it to make it 
usable. Then I've got to rebuild experience in my staff with the 
compiler and all the tools so they don't get wrapped around the axle 
just trying to get it to do what they need it to do. Plus I've got to 
send my staff off to Ada school (against their will) and get them 
trained up to truly be effective with Ada. (Won't happen in two weeks, 
BTW) All this is going to take how many months before I can get started 
generating the code I need? And what is it I get out of it? Will my 
competitor sit still and do nothing while I'm off trying to convert to Ada?

This is an incredibly tough nut to crack and if Ada is serious about 
playing in that market, it ABSOLUTELY needs to be sitting on the shelf 
right next to the development board with 100% of everything you'd get 
with the corresponding C compiler and 100% of the quality of its output 
result. It needs that just in order to play the game. It won't win 
unless it *ALSO* brings with it enough developmental leverage to EXCEED 
the Time To Market the developer has at present with C. He's got to 
overcome the lag time of training up his resistant staff and rebuilding 
whatever essential tools he may have in place already.

The only way out of it is to either start a business that sells an 
incredibly wonderful embedded board for use in some market and supply 
with its development kit an Ada compiler with all the goodies. Or 
alternately, some home-hobbyist-entrepreneur cobbles the parts together 
and starts building an end product that happens to be programmed in Ada 
and that end product starts becomming successful. It won't happen by 
trying to persuade project managers (like me - who happens to be a fan 
of Ada) to switch over to Ada without a clear path to success sitting 
right there in front of them.

My advice would be to ignore the embedded market for the time being 
because it has by and large ignored Ada and is just too expensive for 
which to produce a sufficient product. Get in the door on a bigger 
market that might have some related connection to the embedded world and 
then see if it spills over into that realm.

MDC



Bernd Specht wrote:
> 
> Ok, but they write code for accessing a controller register in a few 
> minutes, to write it Ada with all those rep specs they would  need much 
> longer.
> In our business the first prio is "time to market".
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Dmitry A. Kazakov wrote:
 >
 > I believe you are wrong here. In the area we are working, each large
 > vendor of an embedded application is surrounded by a swarm of smaller
 >  firms to which it outsource software development and testing. Each
 > satellite firm has the complete development tool chain. These tools
 > are very expensive. It is a very lucrative market. [Some tool chain
 > vendors just take GCC, build loader, add a pair windows and sell that
 >  for many k$.]
 >
It tends to not work that way here. Embedded systems software is seldom
farmed out to subcontractors. I'd agree that there is some market for
compilers here, and it isn't as grim as I characterized it. But
basically you find hyundreds of companies out there selling some little
embedded board or chip set and they have a development kit with a C
compiler & all the goodies sitting right there and they're practically
giving it away in order to sell boards. How is some alternative vendor
going to come in there with an Ada compiler and convince the developer
he needs to use theirs (without all the supporting goodies) and it will 
only cost him $20,000 a year in annual support contracts?

It won't fly unless the compiler is right there with the rest of the
development kit and doesn't cost a fortune to get.



 >
 > No. It would take too long. I am afraid that Ada is missing the
 > embedded programming market which in the foreseeable future will
 > become no less important than the conventional one.
 >
It wouldn't take too long if Ada focused in on a branch of software
systems to which its existing strengths play and whos practitioners are 
not already Ada-haters. The embedded market already hates Ada with the 
white hot intensity of a thousand suns for a variety of historical 
reasons. In language surveys "Ada" ranks behind "Other" in languages 
used to program embedded systems. I'd *love* to see Ada get some nice 
share of the embedded market, but I think it is lost for the time being 
because it is too expensive in which to play and has too much historic 
resistance to overcome.

If Ada went for something related - like communications or math or 
related fields - and tried to target that audience, they might be more 
receptive. That and they don't demand as much support gagetry just in 
order to be a player. So you can get in there easier and start building 
a market and because it might have a connection to the embedded world, 
it lets you gradually start to try to get into that arena.

MDC



-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Exactly. If you want a job programming in Ada, you've got to build it 
for yourself. I've always been willing to entertain ideas about 
commercial products that might be built as a garage-venture - but the 
product itself has to have a perceived market and become a successful 
revenue generator. That won't happen just because it was programmed in 
Ada. But if the enthusiasts put their minds to it and dream up a useful 
application where the strengths of Ada might provides some competitive 
advantage, that will start making the "Ada Ressurection" come about.

It *must* generate revenue or its just a hobby. Hobbies don't tend to 
keep language vendors and languages alive. Or is that all Ada aspires to?

MDC


Richard Riehle wrote:
> 
> The Ada enthusiast at LM is not going to persuade the new manager
> who prefers C++ of the benefits of Ada.  The Ada expert who did
> so well on his last project at Raytheon will not turn the tide back
> toward Ada when so many there have lost sight of why it was a
> good idea for earlier systems.   The only hope is for those experts
> to identify an marketing opportunity (even band together) and use
> Ada for a new product.    Give up the security of you Boeing
> pension, your LM health benefits, and launch a new product
> using Ada.   Put you belief in Ada to the test.   That is the only
> way Ada will survive over the next five or ten years.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

When you get 20% of the embedded systems market, I'll congradulate you 
on your enormous success. And I truly hope you do that. Really.

Just remember that others have tried to sell the 
"Ada-to-C-to-Machine-Code" answer in the past and you don't see a whole 
lot of those little controller boards getting programmed in Ada. People 
use the development kit that came with the board for the most part and 
don't want to add extrra layers of stuff on top of it. Especially if it 
involves any work on their part to do so. They want to take it out of 
the box and start programming it. Give them that kind of answer and 
perhaps you'll sell a lot of them.

MDC


Peter Amey wrote:
> 
> 
> Marin David Condic wrote:
> 
>> The "Compile Ada to C then compile C to the target..." approach has 
>> been enormously popular and successful, hasn't it? :-)
> 
> 
> Yes, considering we only came up with the SPARK -> C idea last year it 
> has been rather popular :-)
> 

-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Dmitry A. Kazakov]

On Thu, 13 May 2004 12:01:02 GMT, Marin David Condic
<nobody@noplace.com> wrote:

>Dmitry A. Kazakov wrote:
> >
> > I believe you are wrong here. In the area we are working, each large
> > vendor of an embedded application is surrounded by a swarm of smaller
> >  firms to which it outsource software development and testing. Each
> > satellite firm has the complete development tool chain. These tools
> > are very expensive. It is a very lucrative market. [Some tool chain
> > vendors just take GCC, build loader, add a pair windows and sell that
> >  for many k$.]

>It tends to not work that way here. Embedded systems software is seldom
>farmed out to subcontractors. I'd agree that there is some market for
>compilers here, and it isn't as grim as I characterized it. But
>basically you find hyundreds of companies out there selling some little
>embedded board or chip set and they have a development kit with a C
>compiler & all the goodies sitting right there and they're practically
>giving it away in order to sell boards.

Right, but I am not talking about naked boards. Little could be done
there, here I agree with you. I am talking about

1) the next layer. Many our customers need a board + multitasking +
TCP/IP stack + USB + field bus. They also want it portable [that
becomes a real issue for many customers.] Isn't it what Ada is for? 

2) the whole tool chain. You don't get it with a board.

>How is some alternative vendor
>going to come in there with an Ada compiler and convince the developer
>he needs to use theirs (without all the supporting goodies) and it will 
>only cost him $20,000 a year in annual support contracts?

This is approximately what a C based tool chain costs.

>It won't fly unless the compiler is right there with the rest of the
>development kit and doesn't cost a fortune to get.

Right

> > No. It would take too long. I am afraid that Ada is missing the
> > embedded programming market which in the foreseeable future will
> > become no less important than the conventional one.

>It wouldn't take too long if Ada focused in on a branch of software
>systems to which its existing strengths play and whos practitioners are 
>not already Ada-haters. The embedded market already hates Ada with the 
>white hot intensity of a thousand suns for a variety of historical 
>reasons. In language surveys "Ada" ranks behind "Other" in languages 
>used to program embedded systems.

I am afraid, that's in the past. Nowadays many are just unaware that
Ada exists.

>I'd *love* to see Ada get some nice 
>share of the embedded market, but I think it is lost for the time being 
>because it is too expensive in which to play and has too much historic 
>resistance to overcome.

But the point is that embedded rapidly grows. I do not propose to go
after the present customers. No, we should go after the new ones.
Those who know nothing about Ada. They would readily switch to any
language other than C. And they will, unfortunately, to C++, C#, Java,
but not to Ada!

>If Ada went for something related - like communications or math or 
>related fields - and tried to target that audience, they might be more 
>receptive. That and they don't demand as much support gagetry just in 
>order to be a player. So you can get in there easier and start building 
>a market and because it might have a connection to the embedded world, 
>it lets you gradually start to try to get into that arena.

--
Regards,
Dmitry Kazakov
www.dmitry-kazakov.de

Re: Ada used in General Aviation (GA) applications?

[Mike Silva]

Bernd.Specht@gmx.com (Bernd Specht) wrote in message news:<Xns94E7CB3F1C485BerndSpechtgmxcom@151.189.20.10>...

> Throw away current develop-tools.
> Throwing away the staffs know how.
> Start from scratch.

Yeah, I can see your point.

> Write in C, instead the employees writing assembly-code.

Nope, I'd never risk it!
 
> 60% less??? Maybe over the next 100 years. But till then the company is 
> dead.

Switching from assembly to C certainly did kill a lot of companies. 
It was a crazy idea.  Don't do it!

Mike

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

"Richard Riehle" <adaworks@earthlink.net> wrote in message
news:buEoc.18396$V97.11476@newsread1.news.pas.earthlink.net...
> I suppose this is my fault for not taking the time (I have so much
> of it) to port Ada to these other platforms myself.   But my time
> is spent reading programming assignments, grading exams, and
> counseling students (which is a lot of time, it turns out), and there
> just isn't time to do the fun things such as port GNAT to the
> interesting platforms we might like to see it on.

A "Dummies Guide to porting GNAT" covering both processor
and OS would be a huge help. I'd love to have Ada available for
SymbianOS (or it's predecessor EPOC32).

-- Martin

Re: End of "discussion" (was Re: Ada used in General Aviation (GA) applications?)

[Robert I. Eachus]

Bernd Specht wrote:

> There is a bigg problem: Most Ada-advocats are too conceived. Look at this 
> thread, I said I would give Ada a chance if there were a version for our 
> processors. What was the reply? "You are stupid if you don't switch over", 
> "You don't do your job well if you work with C" ... (bla bla).
> This really doesn't help.

Hmmm.  Maybe I saw the answers differently.  Apparently there is an Ada 
compiler for the HC11, and you can get a SPARK toolchain for any 
embedded controller.

If I were given the choice to code in SPARK and doing my debugging 
during compilation, or spending lots of time with in-circut emulators I 
wouldn't have to think about the choice at all.  Sorry you see the 
picture differently.

-- 

                                           Robert I. Eachus

"The terrorist enemy holds no territory, defends no population, is 
unconstrained by rules of warfare, and respects no law of morality. Such 
an enemy cannot be deterred, contained, appeased or negotiated with. It 
can only be destroyed--and that, ladies and gentlemen, is the business 
at hand."  -- Dick Cheney

Re: Ada used in General Aviation (GA) applications?

[Mike Silva]

Vinzent 'Gadget' Hoefler <nntp-2004-05@t-domaingrabbing.de> wrote in message news:<bu76a05mpq39k8ouqo56ojgnct6g35ugug@jellix.jlfencey.com>...
> Bernd Specht wrote:
> 
> >Whats about the time to train the staff?
> 
> If they're already programmers at worst two weeks to get most of the
> concepts.

And really, how much of Ada are they likely to need for 8051 and HC11
projects?  (And I keep coming back to the SPARK->C approach as
sounding like a good fit for little processors)

Mike

Re: Ada used in General Aviation (GA) applications?

[Pascal Obry]

Bernd.Specht@gmx.com (Bernd Specht) writes:

> Vinzent 'Gadget' Hoefler <nntp-2004-05@t-domaingrabbing.de> wrote in 
> news:bu76a05mpq39k8ouqo56ojgnct6g35ugug@jellix.jlfencey.com:
> 
> > Bernd Specht wrote:
> > 
> >>Whats about the time to get familiar with the new language?
> > 
> > Well, that takes longer (just like in C). But it's a parallel process.
> > You can do that while already writing code.
> 
> Don't forget, the employees are real C-guys. You can't brainwash them at one 
> afternoon.

And don't forget that this is just a wrong argument! When it comes to Ada you
always hear "Well it takes time and money to leard a new langauge". Now let me
ask some questions :

   - What was the cost of learning C++ ?
   - What was the cost of learning Java/JSP/EJB/.../.../... ?
   - What was (will be) the cost of learning C# and .NET ?
   - What will be the cost of learning the next-super-mega-over-cool-language ?

I've never heard these arguments for other language that Ada, does it
really cost nothing ?

Pascal.

-- 

--|------------------------------------------------------
--| Pascal Obry                           Team-Ada Member
--| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE
--|------------------------------------------------------
--|              http://www.obry.org
--| "The best way to travel is by means of imagination"
--|
--| gpg --keyserver wwwkeys.pgp.net --recv-key C1082595

Re: Ada used in General Aviation (GA) applications?

[Lutz Donnerhacke]

* Pascal Obry wrote:
> And don't forget that this is just a wrong argument! When it comes to Ada
> you always hear "Well it takes time and money to leard a new langauge".
> Now let me ask some questions :
>
>    - What was the cost of learning C++ ?

Nothing. You can program for our customers in C while learning C++.
C++ is nothing more than C with Classes and the Template Library.

>    - What was the cost of learning Java/JSP/EJB/.../.../... ?

Nothing. You can program for our customers in C while learning Java.
Java is nothing more than C++ only using a sligthly different syntax.
The libs are easy to use: Just tune an example from the next website.

>    - What was (will be) the cost of learning C# and .NET ?

Nothing. You can program for our customers in C while learning C#.
C# is nothing more than C++ only using a sligthly different syntax.
.Net is easy to use: Just tune an example from the MSDN.

>    - What will be the cost of learning the next-super-mega-over-cool-language ?

Nothing. You can program for our customers in C while learning it.
It will be nothing more than C or C++ with a sligthly different syntax.
Examples are always available, because everbody uses it.

> I've never heard these arguments for other language that Ada, does it
> really cost nothing ?

It look other than C. You can't be productive while learning.
There are no examples on other websites how to implement the current task.

Of course Perl is more different, but Perl has CPAN.
Of course CPAN Modules are work in progress, if you are going deep, but
there are Modules for every fucking task. And all important modules are
written in C anyway ... So you can still programm in C while writing Perl.

There is a very simple, but important fact against Ada: The whole Network
and OS interfaces are written in C. Only a small part of the interfaces can
be imported into Ada, because the major list of system specific constants
are only available as preprocessor macros. It get worse if you need
structures.

So the main part of developing ordinary applications in Ada is writing a
seperate inferface to the system or the network. Nobody will pay for a list
of already defined constants. Of course, you can write a C wrapper in order
to get a interface, but this is silly. You will be expected to simply
include the system header files. So your language has to be C compatible as
much as possible. Interfacing with Ada is too expensive.

Not kidding :-(
\f
Reimplemented syslog(3) because this part of the libc is not thread safe:

...
   type Socket_Family is (PF_UNSPEC, PF_LOCAL);
   for Socket_Family use (
     PF_UNSPEC       => 0,
     PF_LOCAL        => 1
   );
   function PF_UNIX return Socket_Family renames PF_LOCAL;
   function PF_FILE return Socket_Family renames PF_LOCAL;

   type Socket_Addr(family : Socket_Family) is record
      case family is
	 when PF_LOCAL =>
	    path : String(1 .. 108);
	 when PF_UNSPEC =>
	    null;
      end case;
   end record;
   
   for Socket_Addr use record
      family at 0 range 0 .. 15;
      path   at 2 range 0 .. 108*8 - 1;
   end record;
   type Socket_Addr_Ptr is access constant Socket_Addr;
   
   log_file : aliased constant Socket_Addr := (
     family => PF_LOCAL,
     path   => "/dev/log" & (9 .. 108 => ASCII.NUL)
   );
   
   type Socket_Type is (SOCK_STREAM, SOCK_DGRAM, SOCK_RAW, SOCK_RDM,
     SOCK_SEQPACKET, SOCK_PACKET);
   for Socket_Type use (
     SOCK_STREAM => 1,
     SOCK_DGRAM => 2,
     SOCK_RAW => 3,
     SOCK_RDM => 4,
     SOCK_SEQPACKET => 5,
     SOCK_PACKET => 10
   );
   
   type File_Descriptor is new Integer;
   
   function c_socket(family : Socket_Family; mode : Socket_Type; protocol : Integer) return File_Descriptor;
   pragma Import(C, c_socket, "socket");
   
   procedure c_close(desc : File_Descriptor);
   pragma Import(C, c_close, "close");
   
   procedure c_sendto(socket : File_Descriptor;
     msg : String; msg_len : Natural;
     flag : Integer;
     addr : Socket_Addr_Ptr; addr_len : Natural
   );
   pragma Import(C, c_sendto, "sendto");
   
   procedure Send_To(Socket : File_Descriptor; Msg : String; To : Socket_Addr_Ptr) is
      pragma Inline(Send_To);
   begin
      c_sendto(Socket, Msg, Msg'Length, 0, To, To.all'Size / 8);
   end Send_To;
   
...

Re: Ada used in General Aviation (GA) applications?

[Randy Brukardt]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40A35FF6.9050703@noplace.com...
> Lots of people here seem to lack much understanding of or sympathy for
> "Time To Market" - perhaps because they come from industries where this
> is not a big pressure. But you're right - in lots of business sectors
> (the ones that generate money for compiler vendors) time to market is
> *everything*. All the maintenance costs and long term bug problems in
> the world pale in comparison to "Time To Market".

In which case, Ada is doomed (and we all probably are, because buggy
programs will kill us). Ada is about building programs the right way, with
few bugs and ability to fix/enhance in the future. None of that matters in
the least when time to market is everything. Ship any old crap, and fix it
next time. Or never bother.

If the world does not start insisting on programs that actually work (and
thus reduce the time to market pressure), then there is no hope.

                  Randy.

Re: Ada used in General Aviation (GA) applications?

[Randy Brukardt]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40A36368.7050104@noplace.com...
...
> It wouldn't take too long if Ada focused in on a branch of software
> systems to which its existing strengths play and whos practitioners are
> not already Ada-haters. The embedded market already hates Ada with the
> white hot intensity of a thousand suns for a variety of historical
> reasons. In language surveys "Ada" ranks behind "Other" in languages
> used to program embedded systems. I'd *love* to see Ada get some nice
> share of the embedded market, but I think it is lost for the time being
> because it is too expensive in which to play and has too much historic
> resistance to overcome.

Ada's share of non-embedded markets is zero to any reasonable precision that
you want to use. I don't forsee that changing in any significant way (the
$25 I can put into marketing Ada and Claw isn't going to make any
difference). The only place it gets used is in larger embedded systems (like
avionics, ATC, etc.) But perhaps you are using some odd-ball definition of
"embedded"??

                   Randy.

Re: Ada used in General Aviation (GA) applications?

[Bernd Trog]

Bernd.Specht@gmx.com (Bernd Specht) wrote in message 

> > GNAT is free, if you're not relying on professional support.
> 
> There is not GNAT for 8051/HC11.

Well, there is the gcc(gnat) for hc11.
If you like, you can build an Ada cross compiler for the hc11 with this line:

../configure --target=m68hc11 --enable-languages=ada,c; make

This should even work on windoze, if you have the MinGW installed.

http://gel.sourceforge.net/ada_example.php

Symbian OS (was: Re: Ada used in General Aviation (GA) applications?)

[Alexander E. Kopilovich]

Martin Dowie wrote:

>A "Dummies Guide to porting GNAT" covering both processor
>and OS would be a huge help. I'd love to have Ada available for
>SymbianOS (or it's predecessor EPOC32).

Being more or less familiar with Symbian OS (mostly SonyEricsson's flavour -
Symbian 7.0, much less Nokia's Series 60 flavour - Symbian 6.1) I think that
any guide of porting GNAT will not bring you near to that aim.

Actually, current Symbian C++ development toolset consist of C++ compiler
(usually Metrowerks CodeWarrior is used for Symbian 7.0 and MSVC 6.0 for
Symbian 6.1, although there are SDKs for other compilers, for example,
Borland's C++ Builder), resource compiler, linker (with addition for
resources) - all that produces an executable for emulator only - and emulator
itself. Production version of the program (which will be loaded into 
smartphone) has to be compiled and built with another toolset, which is
GCC-based.

So, in principle one may try to port GNAT for Symbian, but there will be
severe haedaches with GCC versions, and you'll not be able to use the emulator
for development.

Then, even if you adapt GNAT for Windows for use with the emulator, you still
be quite restricted, because the emulator has problems with support of tasking
- in Symbian terminology it supports "threads", but not "processes". Well,
C++ development on emulator has the same limitation, so I just warn you that
you should not expect too much from Ada tasking on emulator (if you aren't
going to develop better emulator).

But all that is less than half of work, I think - because there are mountains
of APIs for Symbian (and many of them differ significantly between Symbian's
flavours). And it will be quite serious work to provide Ada bindinds even for
most needed APIs. (As you can expect, those APIs/classes/etc. are only
partially documented, and the source code is not available after EPOC32 R5).

Actually I thought that RR Software's Janus/Ada might be more proper thing
(than GNAT) for targetting at Symbian, and several times I tempted to tell
Randy about this opportunity... but there is a lot of work, and perhaps small
chances that that work will be compensated by sales. If there were interest
and some funding from one of smartphone vendors that use or plan to use
Symbian (Nokia, SonyEricsson, Siemens, Samsung, etc. ... I'm not sure about
Motorola) - it will be another matter, but they still are trying to push Java
(without much success, though - as far as I can see).




Alexander Kopilovich                      aek@vib.usr.pu.ru
Saint-Petersburg
Russia

Re: Ada used in General Aviation (GA) applications?

[Bartłomiej Świercz]

In article <slrnca7amb.18s.lutz@taranis.iks-jena.de>, Lutz Donnerhacke wrote:
> Nothing. You can program for our customers in C while learning C++.
> C++ is nothing more than C with Classes and the Template Library.

It's not truth, don't forget about type checking and save behaviour like
'static_cast'.

> There is a very simple, but important fact against Ada: The whole Network
> and OS interfaces are written in C. Only a small part of the interfaces can
> be imported into Ada, because the major list of system specific constants
> are only available as preprocessor macros. It get worse if you need
> structures.

I agree with you in 100%. And I have other point of view: if someone
ports a library to Ada it's usually too old version of library for me :(

> So the main part of developing ordinary applications in Ada is writing a
> seperate inferface to the system or the network. Nobody will pay for a list
> of already defined constants. Of course, you can write a C wrapper in order
> to get a interface, but this is silly. You will be expected to simply
> include the system header files. So your language has to be C compatible as
> much as possible. Interfacing with Ada is too expensive.

In my opinion Ada should be move to embedded market (for new boards and
processors) because PC market has too old C tradition ...

Best regards,
-- 
Bart�omiej �wiercz
http://sknauk.wpk.p.lodz.pl
PGP Public Key: http://wilk.wpk.p.lodz.pl/swierczu/pgp/

Re: Ada used in General Aviation (GA) applications?

[tmoran]

>the least when time to market is everything. Ship any old crap, and fix it
>next time. Or never bother.
   In the somewhat longer run, of course, crap will lose repeat customers,
thus making opportunities for new vendors.  If Windows had been really
good, would Linux exist?

Re: Ada used in General Aviation (GA) applications?

[Pascal Obry]

Lutz Donnerhacke <lutz@iks-jena.de> writes:

> * Pascal Obry wrote:
> > And don't forget that this is just a wrong argument! When it comes to Ada
> > you always hear "Well it takes time and money to leard a new langauge".
> > Now let me ask some questions :
> >
> >    - What was the cost of learning C++ ?
> 
> Nothing. You can program for our customers in C while learning C++.
> C++ is nothing more than C with Classes and the Template Library.
> 
> >    - What was the cost of learning Java/JSP/EJB/.../.../... ?
> 
> Nothing. You can program for our customers in C while learning Java.
> Java is nothing more than C++ only using a sligthly different syntax.
> The libs are easy to use: Just tune an example from the next website.
> 
> >    - What was (will be) the cost of learning C# and .NET ?
> 
> Nothing. You can program for our customers in C while learning C#.
> C# is nothing more than C++ only using a sligthly different syntax.
> .Net is easy to use: Just tune an example from the MSDN.
> 
> >    - What will be the cost of learning the
> >    next-super-mega-over-cool-language ?
> 
> Nothing. You can program for our customers in C while learning it.
> It will be nothing more than C or C++ with a sligthly different syntax.
> Examples are always available, because everbody uses it.

Good, now I'm waiting for your theory about the fact that there is lot of
courses on C/C++/Java/EJB/JSP/C#/.NET/.../.../... in many institutes... and
there is lot of attendes ! Cost nothing, are you saying that these courses
are just given for free by those guys !

Pascal.

-- 

--|------------------------------------------------------
--| Pascal Obry                           Team-Ada Member
--| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE
--|------------------------------------------------------
--|              http://www.obry.org
--| "The best way to travel is by means of imagination"
--|
--| gpg --keyserver wwwkeys.pgp.net --recv-key C1082595

Re: Ada used in General Aviation (GA) applications?

[Alexander E. Kopilovich]

Lutz Donnerhacke wrote:

> >    - What was the cost of learning C++ ?
>
> Nothing. You can program for our customers in C while learning C++.

Yes, at least I always managed to do that. What is significant here, is that
you can program in C inside C++ classes and their methods.

> C++ is nothing more than C with Classes and the Template Library.

This is appropriate viewpoint for dealing with C++ programs using C.

> >    - What was the cost of learning Java/JSP/EJB/.../.../... ?
>
> Nothing. You can program for our customers in C while learning Java.

Almost yes, at least I always managed to do that; but I knew in advance about
threads and inheritance - without that it would be practically impossible
to write many sorts of Java programs.

> Java is nothing more than C++ only using a sligthly different syntax.

Plus threads and interfaces (you can avoid interfaces in you own code, but
you still must understand them in a code produced by your co-worker or
predecessor).

And minus enumerations.

> The libs are easy to use: Just tune an example from the next website.

Perhaps "just try an example from the next website" would be more proper
expression -;)

>    - What was (will be) the cost of learning C# and .NET ?
>
> Nothing. You can program for our customers in C while learning C#.

I think this is mostly wrong statement - because C# contains too many small
things that weren't in C/C++ (for one little example - boxing), and those
things will appear too often in code produced by your co-workers and
predecessors. If you don't know them you'll be unable to understand code,
that was written by your neighbour (or subcontractor). You'll guess in those
place and sometimes you'll miss.

So you may not be proficient in C#, but you still must be somehow acquitanted
with it.

> C# is nothing more than C++ only using a sligthly different syntax.

It depends on which C++ you have used before. I think that if you used
Borland's C++ Builder, it will be nearer to truth (although even in this case
not quite true) than if you used MSVC or CodeWarrior or GCC.

> .Net is easy to use: Just tune an example from the MSDN.

This may be true as long as you have no uncommon needs.

> > I've never heard these arguments for other language that Ada, does it
> > really cost nothing ?
>
> It look other than C. You can't be productive while learning.

If you have Pascal (well, Object Pascal/Delphi) background in addition to
your C background then probably you can do the same thing - you can consider
Ada as Object Pascal with slightly different syntax.

> There are no examples on other websites how to implement the current task.

And this is indeed great difference.

>...
>There is a very simple, but important fact against Ada: The whole Network
>and OS interfaces are written in C.

Actually, the absence of their Ada counterparts is correlated with the
scarcity of Ada examples.




Alexander Kopilovich                      aek@vib.usr.pu.ru
Saint-Petersburg
Russia

Re: Ada used in General Aviation (GA) applications?

[Alexander E. Kopilovich]

Randy Brukardt wrote:

> > in lots of business sectors
> > (the ones that generate money for compiler vendors) time to market is
> > *everything*. All the maintenance costs and long term bug problems in
> > the world pale in comparison to "Time To Market".
>
> In which case, Ada is doomed

So far Ada is well alive (presence of several maintained compilers is
evident indication of that). SPARK gradually gains popularity. So the meaning
of "doomed" here is quite unclear. 

>  (and we all probably are, because buggy programs will kill us).

Not at all, those buggy programs will kill themselves before they can reach
most of us. And this is good that those programs are buggy enough, because
otherwise they will be able to reach too many of us.

> Ada is about building programs the right way, with
> few bugs and ability to fix/enhance in the future.

This claim is too universal and therefore inevitably wrong. There can't be
a universal right way to build programs for all application domains and in
all circumstances and for all reasonable requirements (which are in turn,
dictated by circumstances).

> None of that matters in
> the least when time to market is everything. Ship any old crap, and fix it
> next time. Or never bother.

Moreover, the market (and the vendors) really needs no-so-good products for
two purposes: first. too good products need not to be replaced with better
products often enough; second, not-so-good products are needed for justifying
higher prices for better products of the same kind offered at the same time.
(Perhaps someone will call this cynical, but it isn't - it is just a careful
observation.)

> If the world does not start insisting on programs that actually work (and
> thus reduce the time to market pressure), then there is no hope.

The world will never insist on that because the world does not know (and can't
know) what does it mean "actually work" for computer programs. The world
evaluates whole things, but not parts of them - the parts belong to
specialists.




Alexander Kopilovich                      aek@vib.usr.pu.ru
Saint-Petersburg
Russia

Re: Ada used in General Aviation (GA) applications?

[Alexander E. Kopilovich]

Pascal Obry wrote:

> Good, now I'm waiting for your theory about the fact that there is lot of
> courses on C/C++/Java/EJB/JSP/C#/.NET/.../.../... in many institutes... and
> there is lot of attendes ! Cost nothing, are you saying that these courses
> are just given for free by those guys !

I offer 5 theories for that:

1) those attendees (or their employers) seek certificates;

2) they seek either socialisation or social adaptation for next stage of 
their careers;

3) they can't learn without teacher's lead and supervision;

4) they need really fast-track learning;

5) they need a collection of implementation patterns;

I think that vast majority of attendees of those courses fall to first 3
categories, and that most of the courses actually can't give more than that
needed for those 3 categories.

But when Lutz said "costs nothing" he surely assumed an experienced programmer,
who is already proficient in C, and certainly most attendees of first 3
categories do not qualify for that criteria.

And regarding courses that are given for free - oh, yes, there are plenty
of them - for example, I recently received notification from ACM about
availability of 150 (or even 300? I don't remember) online courses for
those languages/environments, which are (or most of them are) free for ACM
members (and ACM membership isn't very expensive).




Alexander Kopilovich                      aek@vib.usr.pu.ru
Saint-Petersburg
Russia

Re: Ada used in General Aviation (GA) applications?

[Ole-Hjalmar Kristensen]

Lutz Donnerhacke <lutz@iks-jena.de> writes:

> * Pascal Obry wrote:
> > And don't forget that this is just a wrong argument! When it comes to Ada
> > you always hear "Well it takes time and money to leard a new langauge".
> > Now let me ask some questions :
> >
> >    - What was the cost of learning C++ ?
> 
> Nothing. You can program for our customers in C while learning C++.
> C++ is nothing more than C with Classes and the Template Library.
> 

Pretty much agree here, although you will probably have some false
steps in the beginning if you have not done any object-oriented
programming before.

<snip>

> 
> > I've never heard these arguments for other language that Ada, does it
> > really cost nothing ?
> 
> It look other than C. You can't be productive while learning.
> There are no examples on other websites how to implement the current task.
> 

You are right, there are few examples around. However, I would say
that you can be productive even while learning Ada. At the most basic
level, C, Pascal, and Ada are sufficiently similar that you can design
the system exactly the same way and implement it in any of these
languages without difficulty. Been there, done that, for the very
first prototype of the NATO ATCCIS system. So I would say it depends
on what you are building. If you are building a system from scratch,
you will probably finish just as fast in Ada as in C even if you do
not know Ada already. If the activity is mainly calling existing
libraries written in C or C++, then C or C++ will certainly be faster.

> Of course Perl is more different, but Perl has CPAN.
> Of course CPAN Modules are work in progress, if you are going deep, but
> there are Modules for every fucking task. And all important modules are
> written in C anyway ... So you can still programm in C while writing Perl.
> 
> There is a very simple, but important fact against Ada: The whole Network
> and OS interfaces are written in C. Only a small part of the interfaces can
> be imported into Ada, because the major list of system specific constants
> are only available as preprocessor macros. It get worse if you need
> structures.
> 
> So the main part of developing ordinary applications in Ada is writing a
> seperate inferface to the system or the network. Nobody will pay for a list
> of already defined constants. Of course, you can write a C wrapper in order
> to get a interface, but this is silly. You will be expected to simply
> include the system header files. So your language has to be C compatible as
> much as possible. Interfacing with Ada is too expensive.
> 
Too expensive if most of your code is calling OS and network primitives, yes.

Actually, writing interface functions to the OS/network/database in C
may be the path of least resistance if you are writing portable
software. You define a set of interface functions which abstract away
the details of the system around you. This set of functions is tuned
to the specific needs of your system, not a set of general wrappers
for OS calls. Of course this will work only if the environments are
reasonably similar, but at least Windows, VMS, and the different
versions of UN*X are similar enough that it works. Then, implement
these functions in whatever language seems most appropriate for a
given platform. This way, you can simply include the system header
files as you say (if you choose to do it in C), and there is really no
unnecessary work, as you need such an abstraction layer to minimize
the cost of porting anyway.

<example snipped>

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

Re: Ada used in General Aviation (GA) applications?

[Anders Wirzenius]

"Randy Brukardt" <randy@rrsoftware.com> writes:

> "Marin David Condic" <nobody@noplace.com> wrote in message
> news:40A35FF6.9050703@noplace.com...
> > Lots of people here seem to lack much understanding of or sympathy for
> > "Time To Market" - perhaps because they come from industries where this
> > is not a big pressure. But you're right - in lots of business sectors
> > (the ones that generate money for compiler vendors) time to market is
> > *everything*. All the maintenance costs and long term bug problems in
> > the world pale in comparison to "Time To Market".
> 
> In which case, Ada is doomed (and we all probably are, because buggy
> programs will kill us). Ada is about building programs the right way, with
> few bugs and ability to fix/enhance in the future. None of that matters in
> the least when time to market is everything. Ship any old crap, and fix it
> next time. Or never bother.
> 
> If the world does not start insisting on programs that actually work (and
> thus reduce the time to market pressure), then there is no hope.
> 
>                   Randy.

A free (no cost) hint to all vendors ;):
1. Start a delivery project and use the programming_language_for_fast_delivery.
2. Start a parallel Ada project for the same product.
3. Deliver the product from point 1 to the customer.
4. While he installs and sets up his test environment, continue with the point 2 project.
5. Receive bug reports from the customer.
6. Correct the bugs from point 5 and deliver a new version.
7. Meanwhile, continue with point 2 project.
8. Repeat points 5..7 as many times as you see fit.
9. Finally, replace the delivery with the product from point 2.

Anders

Re: Ada used in General Aviation (GA) applications?

[Kai Glaesner]

Randy,

> Ada's share of non-embedded markets is zero to any reasonable precision
that
> you want to use. I don't forsee that changing in any significant way (the
> $25 I can put into marketing Ada and Claw isn't going to make any
> difference). The only place it gets used is in larger embedded systems
(like
> avionics, ATC, etc.)But perhaps you are using some odd-ball definition of
> "embedded"??

So the question is: what's embedded?

When I was working for a small avionics company several years ago producing
"Flight Inspection" and navigation avionics they developed their software in
Modula-2 using VMEbus Motorola boards running OS9. Is that embedded?

I think today's avionics tend to more and more based on COTS hardware and
rely heavily on concurrency, distribution, being physical connected via
ethernet and using kind of realtime-ORB's as an communication
infrastructure.

Do you catch the drift? These are all domains Ada is good in.

I dont think the problem are Ada compilers being targeted to the wrong
processor, but software managers in avionic business not knowing about Ada
capabilities in the above mentioned architectures. This leeds to e.g UPSAT
(now Garmin) deploying a PC104-based(certified) Moving Map Display (the
MX20), dvelopped in C/C++ running under Windoze NT(!).

Regards

Kai

RE: Ada used in General Aviation (GA) applications?

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: Lutz Donnerhacke [mailto:lutz@iks-jena.de]
...
| It look other than C. You can't be productive while learning.

That's absolutly *not* our experience here in Thales Communications. 
Beginners on my (rather complex) project understand quickly the code,
because Ada is crystal clear. 
They are able to fix bug quickly because the language and the compiler catch
most stupid syntax error.

We have several time succesfully integrated beginners (first job and no Ada
knowledge). It was easy, provided that peer reviews are done and that
experienced programmers are available for sharp questions, and that they
start with simple coding task.

Those beginners will know 50% of Ada long before knowing 10% of our domain,
and that's a much bigger cost.

-- 
Lionel Draghi 

PS : to be franck, I won't hire someone with programming experience that
can't master Ada basis in a week.

Re: Ada used in General Aviation (GA) applications?

[Andersen Jacob Sparre]

Anders Wirzenius wrote:

> A free (no cost) hint to all vendors ;):
> 1. Start a delivery project and use the
>    programming_language_for_fast_delivery.
> 2. Start a parallel Ada project for the same product.
> 3. Deliver the product from point 1 to the customer.

But what is point 2 is done before point 1?  (rumour say that it
happens)

Well maybe you can sell 2 for a higher price than 1.

Jacob (who works in a field, where there is no prize for being number two)
-- 
"Any, sufficiently advanced, technology is indistinguishable from magic."

Re: Ada used in General Aviation (GA) applications?

[Martin Dowie]

<Lionel.DRAGHI@fr.thalesgroup.com> wrote in message
news:mailman.134.1084535057.313.comp.lang.ada@ada-france.org...

> That's absolutly *not* our experience here in Thales Communications.
> Beginners on my (rather complex) project understand quickly the code,
> because Ada is crystal clear.
> They are able to fix bug quickly because the language and the compiler
catch
> most stupid syntax error.

A company I'm familiar with (cough!) take non-computer degree
qualified people and train them to be Software Engineers in high-intensity
courses. The language used - Ada. Why? See the article @

http://www.computerweekly.com/Article106744.htm

Try that with C/C++...

With Eiffel/Java you might stand a hint of a chance.

-- Martin

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

Anders Wirzenius writes:
> A free (no cost) hint to all vendors ;):
> 1. Start a delivery project and use the
>    programming_language_for_fast_delivery.
> 2. Start a parallel Ada project for the same product.
> 3. Deliver the product from point 1 to the customer.
> 4. While he installs and sets up his test environment, continue with
>    the point 2 project.
> 5. Receive bug reports from the customer.
> 6. Correct the bugs from point 5 and deliver a new version.
> 7. Meanwhile, continue with point 2 project.
> 8. Repeat points 5..7 as many times as you see fit.
> 9. Finally, replace the delivery with the product from point 2.

This is not really a no-cost strategy; it basically acknowledges that
the non-Ada code is to be thrown away.  Engineers working on that code
will rightly feel they are not being recognised, so the quality of the
first project would be abysmal.  This plus the cost of developing the
thing twice is not negligible.

I would prefer an approach where a partial delivery occurs early to
please the customer(s), but this partial delivery would be 100% Ada
and would be partial in functionality, not in quality.

Of course, the vendor ofsuch software needs to educate the customers,
saying "see, this does only part of what you want, but it is so solid
you can entrust your life to it.  The other parts of what you want can
now be implemented quickly and with the same quality level".

-- 
Ludovic Brenta.

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

"Kai Glaesner" writes:
> So the question is: what's embedded?
> 
> When I was working for a small avionics company several years ago
> producing "Flight Inspection" and navigation avionics they developed
> their software in Modula-2 using VMEbus Motorola boards running
> OS9. Is that embedded?

Definitely.

> I think today's avionics tend to more and more based on COTS
> hardware and rely heavily on concurrency, distribution, being
> physical connected via ethernet and using kind of realtime-ORB's as
> an communication infrastructure.

At Barco we design the displays down to the layout of the boards.
Most of the ASICs and processors are COTS, but not much more.

> Do you catch the drift? These are all domains Ada is good in.

Yes.  We use Ada for almost everything.

> I dont think the problem are Ada compilers being targeted to the
> wrong processor, but software managers in avionic business not
> knowing about Ada capabilities in the above mentioned
> architectures. This leeds to e.g UPSAT (now Garmin) deploying a
> PC104-based(certified) Moving Map Display (the MX20), dvelopped in
> C/C++ running under Windoze NT(!).

Certified to what level?  When we do DO-178B level A or B we certainly
do not use Windows XP.  If it's level E then it is not life-critical,
and the only potential problem is customers complaining they've paid
millions in development contracts just to get the blue screen of
death :)

-- 
Ludovic Brenta.

Re: Ada used in General Aviation (GA) applications?

[Jacob Sparre Andersen]

Ludovic Brenta wrote:

> I would prefer an approach where a partial delivery occurs early to
> please the customer(s), but this partial delivery would be 100% Ada
> and would be partial in functionality, not in quality.

Why is it that everybody seem to say that it is slower to get programs
written in Ada out of the door, than those written in other languages?

Are there any data indicating that this is the case?

Or some sensible arguments for it?

Jacob
-- 
�For there are only two reasons why war is made against a
 republic: The one, to become lord over her: the other, the 
 fear of being occupied by her.�       -- Nicolo Machiavelli

Re: Ada used in General Aviation (GA) applications?

[Ludovic Brenta]

Jacob Sparre Andersen writes:
> Ludovic Brenta wrote:
>
>> I would prefer an approach where a partial delivery occurs early to
>> please the customer(s), but this partial delivery would be 100% Ada
>> and would be partial in functionality, not in quality.
>
> Why is it that everybody seem to say that it is slower to get programs
> written in Ada out of the door, than those written in other languages?
>
> Are there any data indicating that this is the case?
>
> Or some sensible arguments for it?

I also think that developing in Ada is faster than in C.  However,
even then, the pressure for time to market may make it impractical to
do the Right Thing.  Under such pressures, the common temptation is to
deliver lower quality, because the most time is spent testing and
debugging the program.  My temptation is to deliver less but with high
quality.  In the case of avionics, testing is mandatory for
certification anyway, so skipping that hase is not an option.

Another aspect of Ada is that it encourages people to do the Right
Thing, and think before they act.  Many will feel that thinking before
acting will lead to slower development.  My personal feeling is that,
whatever the size of the project, it always pays off.

-- 
Ludovic Brenta.

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Jacob Sparre Andersen wrote:

> Why is it that everybody seem to say that it is slower to get programs
> written in Ada out of the door, than those written in other languages?
> 
> Are there any data indicating that this is the case?
> 
> Or some sensible arguments for it?

I'm aware of 2 real-world studies, in 2 different domains, of real 
projects, done both in C and Ada, with good metrics, that show that Ada 
reaches deployment in half the time/cost of C. As a result, I see no 
evidence to support claims that C-like languages have better 
time-to-market characteristics than Ada.

These same studies show Ada has a factor of 4 fewer post-deployment 
errors, and a factor of 10 less time/cost to correct an error.

Note to MDC: All other things are not equal, but that was true of these 
projects as well, and Ada still came out with a factor of 2 advantage on 
time to deployment.

-- 
Jeff Carter
"Brave Sir Robin ran away."
Monty Python and the Holy Grail
59

Re: Ada used in General Aviation (GA) applications?

[Robert I. Eachus]

Martin Dowie wrote:

> A company I'm familiar with (cough!) take non-computer degree
> qualified people and train them to be Software Engineers in high-intensity
> courses. The language used - Ada. Why? See the article @
> 
> http://www.computerweekly.com/Article106744.htm

The most telling quote in the article is: "Although the candidates learn 
the Ada programming language, which is widely used in the defence 
industry, the aim is to teach software engineering principles rather 
than Ada programming."

Teaching software engineering to some programmers is HARD.  Teaching 
software engineering to hot-shot programmers is, in my experience, 
impossible. (There are many converts, but they never come from teaching, 
only from experience.)
-- 

                                           Robert I. Eachus

"The terrorist enemy holds no territory, defends no population, is 
unconstrained by rules of warfare, and respects no law of morality. Such 
an enemy cannot be deterred, contained, appeased or negotiated with. It 
can only be destroyed--and that, ladies and gentlemen, is the business 
at hand."  -- Dick Cheney

Re: Ada used in General Aviation (GA) applications?

[Anders Wirzenius]

Andersen Jacob Sparre <sparre@jacob.crs4.it> writes:

> Anders Wirzenius wrote:
> 
> > A free (no cost) hint to all vendors ;):
> > 1. Start a delivery project and use the
> >    programming_language_for_fast_delivery.
> > 2. Start a parallel Ada project for the same product.
> > 3. Deliver the product from point 1 to the customer.
> 
> But what is point 2 is done before point 1?  (rumour say that it
> happens)

Don't Violate The Rules! I repeat: Don't Violate The Rules! ;)

Anders

Re: Ada used in General Aviation (GA) applications?

[Anders Wirzenius]

Ludovic Brenta <ludovic.brenta@insalien.org> writes:

> Anders Wirzenius writes:
> > A free (no cost) hint to all vendors ;):
> > 1. Start a delivery project and use the
> >    programming_language_for_fast_delivery.
> > 2. Start a parallel Ada project for the same product.
> > 3. Deliver the product from point 1 to the customer.
> > 4. While he installs and sets up his test environment, continue with
> >    the point 2 project.
> > 5. Receive bug reports from the customer.
> > 6. Correct the bugs from point 5 and deliver a new version.
> > 7. Meanwhile, continue with point 2 project.
> > 8. Repeat points 5..7 as many times as you see fit.
> > 9. Finally, replace the delivery with the product from point 2.
> 
> This is not really a no-cost strategy; it basically acknowledges that
> the non-Ada code is to be thrown away.  Engineers working on that code
> will rightly feel they are not being recognised, so the quality of the
> first project would be abysmal.  This plus the cost of developing the
> thing twice is not negligible.

I hope you noticed my smiley.
The "no cost" was aimed to point to the hint, not to the project.
In stead of throwing it away, you may add a step nr 10:
Sit down with your programmers and review both the Ada code and the programming_language_for_fast_delivery code. That could be a convenient way to do an inhouse training session in programming. After that you may try to reach some consensus about which code to be the finally delivery product which is maintained. 
Allowing both parts ("Ada_people" and "fast_programming_people") to continue to use their preferred coding language would take care of the not_being_recognised feeling and could finally motivate some "fast_programming_people" to study more in detail how things are done by the "Ada_people". And then ... who knows ... :)

> 
> I would prefer an approach where a partial delivery occurs early to
> please the customer(s), but this partial delivery would be 100% Ada
> and would be partial in functionality, not in quality.
>
> Of course, the vendor ofsuch software needs to educate the customers,
> saying "see, this does only part of what you want, but it is so solid
> you can entrust your life to it.  The other parts of what you want can
> now be implemented quickly and with the same quality level".
> 
> -- 
> Ludovic Brenta.

This is a form of the prototyping approach with early reviewing.

Here is another way of doing early reviewing:
In early 80's I worked for a Fortran community. My co-programmer and I started to use Ada as a pseudo-code language. We made a code review of the Ada code and did not start the detailed Fortran coding until the review had approved the pseudo-code. The Ada code was left in the programs as comments.

Anders

Re: Ada used in General Aviation (GA) applications?

[Martin Krischik]

Anders Wirzenius wrote:

> "Randy Brukardt" <randy@rrsoftware.com> writes:

> A free (no cost) hint to all vendors ;):
> 1. Start a delivery project and use the
> programming_language_for_fast_delivery.
> 2. Start a parallel Ada project 
> ....
> 9. Finally, replace the delivery with the product from point 2.

But would say that:

programming_language_for_fast_delivery := Ada;

Which make point 2 and 9 obsolete.

With Regards

Martin

-- 
mailto://krischik@users.sourceforge.net
http://www.ada.krischik.com

Re: Time to market, was: Ada used in General Aviation (GA) applications?

[Anders Wirzenius]

Jeffrey Carter <spam@spam.com> writes:

> 
> These same studies show Ada has a factor of 4 fewer post-deployment
> errors, and a factor of 10 less time/cost to correct an error.
> 
> Note to MDC: All other things are not equal, but that was true of
> these projects as well, and Ada still came out with a factor of 2
> advantage on time to deployment.

Ed. Berard once said about rushing too fast into coding: "If your boss wants to see code, take whatever code you have, give it to your boss, and return to your specification of the real product."

You write about post-deployment, my posting that started this thread was a rant (with a smiley) around the theme: "If your customer want it in the afternoon, take whatever you have, deliver it, and start to do debugging".

Yes, it is in the long run all about (final) deployment of the product. Either you deliver many times, or you deliver a few times (or even only once, free of bugs). It is the long run quality that keeps the company alive and in the business. No doubt the persons behind the company may stay in business also by establish new short life companies when the present company fails to get new orders. You may only fool once, but you may always start fresh with a new company. ;)

Anders

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Why does that have to exclude Time To Market? T.T.M. does *not* mean 
that a system has to be buggy or unsafe. If Ada was providing tools, 
utilities, features, etc., to a given market that meant significant 
leverage to those who develop for that market, wouldn't they get some 
added quality as well just by using Ada and having it perform its checks 
on the code prior to delivery?

Industries have various levels of required quality. I expect more from 
flight certified software than I do from the average open source freebie 
desktop app I downloaded from the internet. Assume Ada has that elusive 
All Other Things Being Equal factor when compared against Language-X in 
a given problem domain. If so, then your time to market is equal, except 
you get fewer bugs as a bonus - and probably some percentage of faster 
development time too because you don't debug as much. Assume Ada tosses 
on top of that some spiffy features, utilities, libraries, etc., that 
reduce that time to market by 50%. Ada suddenly has a *HUGE* competitive 
advantage over Language-X. Enough so that it might persuade someone to 
bite the bullet and take a chance using Ada in the next deployment of a 
product.

So A) Define the market you want, B) Fix it so that all other things are 
nearly equal and C) Provide a truckload of additional leverage so that 
the argument to use Ada is extremely compelling to the guys who have to 
keep their business alive by whupping-up on their competitors who stick 
with Language-X. (Note: It's not compelling enough just to say you'll 
get fewer bugs. Ada has been making that case for years and it hasn't 
persuaded many commercial users.)

Or we could just get cynical about it and abandon hope of seeing Ada 
successful somewhere in the commercial computer world. ;-)

MDC


Randy Brukardt wrote:

> programs will kill us). Ada is about building programs the right way, with
> few bugs and ability to fix/enhance in the future.


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

But then you go broke because you've got double the development costs as 
compared to your competitor. ;-)

MDC

Andersen Jacob Sparre wrote:
> 
> But what is point 2 is done before point 1?  (rumour say that it
> happens)
> 
> Well maybe you can sell 2 for a higher price than 1.

-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Why must it take longer to get to market with Ada? I don't think it 
does. Not when you get the "All Other Things Being Equal" magic ticket 
in hand. The problem is the "All Other Things Being Equal" part. Fix 
that first and you'll instantly be competitive because Ada development 
can be more productive. Add additional leverage for the problem domain 
and it can't be ignored. If you *really* had a way to get to market 
twice as fast as the leading competitor, and the industry folks refused 
to believe you, go open a competing business and blow their doors off 
with your spiffy, extra-added-value, secret weapon: Ada.

MDC



Ludovic Brenta wrote:
> 
> 
> This is not really a no-cost strategy; it basically acknowledges that
> the non-Ada code is to be thrown away.  Engineers working on that code
> will rightly feel they are not being recognised, so the quality of the
> first project would be abysmal.  This plus the cost of developing the
> thing twice is not negligible.
> 
> I would prefer an approach where a partial delivery occurs early to
> please the customer(s), but this partial delivery would be 100% Ada
> and would be partial in functionality, not in quality.
> 
> Of course, the vendor ofsuch software needs to educate the customers,
> saying "see, this does only part of what you want, but it is so solid
> you can entrust your life to it.  The other parts of what you want can
> now be implemented quickly and with the same quality level".
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

I don't think that it does - in principle. But we don't build things in 
the abstract. We build things in the Real World(tm) where a competing 
language may offer lots of leverage that raw Ada alone can't compete 
against. So if we want to see Ada succeed somewhere rather than just 
grouse about how the people in that market sector are bone-heads for not 
picking Ada, then we ought to look at what they do, how they do it, what 
help they need, what would make their lives better, etc., and make sure 
Ada provides it.

Or having learned all that stuff and invented a better answer, go off 
and compete against them and win. Then *they* can go off grumbling about 
how Ada is a stupid idea and we control the industry just because we 
somehow cheated or got lucky. :-)

MDC



Jacob Sparre Andersen wrote:
> 
> Why is it that everybody seem to say that it is slower to get programs
> written in Ada out of the door, than those written in other languages?
> 
> Are there any data indicating that this is the case?
> 
> Or some sensible arguments for it?
> 
> Jacob


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Keep in mind that I have done a study that showed a 50% reduction in 
development time and a factor of 4 improvement in defects. You may in 
fact be thinking of that very study here. I've made the statement here 
before.

However, a) it is difficult-to-impossible to claim (from my study) that 
Ada was the only factor and b) enough of it was an "all other things 
being equal" environment to make the claim believable/useful. It was an 
embedded system environment with virtually all of the code home-grown. 
It was *not* an environment in which the OS, GUI, Database, Support 
Libraries, Etc., were all written in C and all the development tools 
were designed with C programming in mind.

You want to take over - for example - Motorola cell phone applets? 
That's an interesting market, but you don't even have an Ada compiler 
supported for that environment - so all other things aren't equal. Then 
there's all the development tools & libraries around it and no similar 
suite of stuff for Ada. There is no "Development Kit" - so Ada can't 
play on even an equal footing - or play at all. All the abstract studies 
in the world citing double-your-productivity-and-reduce-your-bugs are 
useless unless and until you've got that development kit sitting right 
there available for use.

My argument is not to deny the productivity and defect gains, but to 
insist that Ada needs to target a lucrative market and give it 
everything it needs to take advantage of the potential productivity gains.

MDC


Jeffrey Carter wrote:
> 
> 
> I'm aware of 2 real-world studies, in 2 different domains, of real 
> projects, done both in C and Ada, with good metrics, that show that Ada 
> reaches deployment in half the time/cost of C. As a result, I see no 
> evidence to support claims that C-like languages have better 
> time-to-market characteristics than Ada.
> 
> These same studies show Ada has a factor of 4 fewer post-deployment 
> errors, and a factor of 10 less time/cost to correct an error.
> 
> Note to MDC: All other things are not equal, but that was true of these 
> projects as well, and Ada still came out with a factor of 2 advantage on 
> time to deployment.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Try searching the web for "Single Board Computers" and see what pops up. 
Look at what the vendors of those boards provide as development kits. 
Make a list of *everything* they give you when you buy a development 
kit. (Often an RTOS and all the things you cite below) If a developer 
shooting to use that board doesn't see an Ada compiler in that toolset, 
then Ada won't get used.

MDC



Dmitry A. Kazakov wrote:
> 
> Right, but I am not talking about naked boards. Little could be done
> there, here I agree with you. I am talking about
> 
> 1) the next layer. Many our customers need a board + multitasking +
> TCP/IP stack + USB + field bus. They also want it portable [that
> becomes a real issue for many customers.] Isn't it what Ada is for? 
> 
> 2) the whole tool chain. You don't get it with a board.
> 
> 
>>How is some alternative vendor
>>going to come in there with an Ada compiler and convince the developer
>>he needs to use theirs (without all the supporting goodies) and it will 
>>only cost him $20,000 a year in annual support contracts?
> 
> 
> This is approximately what a C based tool chain costs.
> 
> 
>>It won't fly unless the compiler is right there with the rest of the
>>development kit and doesn't cost a fortune to get.
> 
> 
> Right
> 
> 
>>>No. It would take too long. I am afraid that Ada is missing the
>>>embedded programming market which in the foreseeable future will
>>>become no less important than the conventional one.
>>
> 
>>It wouldn't take too long if Ada focused in on a branch of software
>>systems to which its existing strengths play and whos practitioners are 
>>not already Ada-haters. The embedded market already hates Ada with the 
>>white hot intensity of a thousand suns for a variety of historical 
>>reasons. In language surveys "Ada" ranks behind "Other" in languages 
>>used to program embedded systems.
> 
> 
> I am afraid, that's in the past. Nowadays many are just unaware that
> Ada exists.
> 
> 
>>I'd *love* to see Ada get some nice 
>>share of the embedded market, but I think it is lost for the time being 
>>because it is too expensive in which to play and has too much historic 
>>resistance to overcome.
> 
> 
> But the point is that embedded rapidly grows. I do not propose to go
> after the present customers. No, we should go after the new ones.
> Those who know nothing about Ada. They would readily switch to any
> language other than C. And they will, unfortunately, to C++, C#, Java,
> but not to Ada!
> 
> 
>>If Ada went for something related - like communications or math or 
>>related fields - and tried to target that audience, they might be more 
>>receptive. That and they don't demand as much support gagetry just in 
>>order to be a player. So you can get in there easier and start building 
>>a market and because it might have a connection to the embedded world, 
>>it lets you gradually start to try to get into that arena.
> 
> 
> --
> Regards,
> Dmitry Kazakov
> www.dmitry-kazakov.de


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

And we disagree on this.... how?

MDC

Randy Brukardt wrote:
> 
> Ada's share of non-embedded markets is zero to any reasonable precision that
> you want to use. I don't forsee that changing in any significant way (the
> $25 I can put into marketing Ada and Claw isn't going to make any
> difference). The only place it gets used is in larger embedded systems (like
> avionics, ATC, etc.) But perhaps you are using some odd-ball definition of
> "embedded"??
> 
>                    Randy.
> 
> 
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Dmitry A. Kazakov]

On Mon, 17 May 2004 12:25:05 GMT, Marin David Condic
<nobody@noplace.com> wrote:

>Try searching the web for "Single Board Computers" and see what pops up. 
>Look at what the vendors of those boards provide as development kits. 
>Make a list of *everything* they give you when you buy a development 
>kit. (Often an RTOS and all the things you cite below) If a developer 
>shooting to use that board doesn't see an Ada compiler in that toolset, 
>then Ada won't get used.

First of all, our customers do not want to be dependent on some
definite RTOS or board. It is a chance for Ada, provided that things
like networking will be standardized.

Secondly they want the whole tool chain, which is bigger than just
RTOS. For example, quite often they require interfacing to
Matlab/Simulink etc.

--
Regards,
Dmitry Kazakov
www.dmitry-kazakov.de

Re: Ada used in General Aviation (GA) applications?

[Randy Brukardt]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40A8AF7C.5000608@noplace.com...
> And we disagree on this.... how?

You said: "The embedded market already hates Ada with the white hot
intensity of a thousand suns for a variety of historical reasons." And I
pointed out that the only significant use of Ada is in embedded systems.
That means either your statement is wrong, or Ada is doomed - in fact it is
already dead, like a tree with Dutch Elm Disease. The leaves haven't found
out yet.

Both RRS and Aonix spent a lot of effort trying to push Ada into more
mainstream programming. Neither company was successful, and both companies
have pretty much given up. Besides, non-embedded programming is on its way
to extinction anyway - those projects and jobs are rapidly moving to very
low-cost countries. Obviously, quality isn't even on the radar.

Ada's only chance for further expansion is in various kinds of embedded
systems (bigger ones; Ada isn't really suited for single-board computers).

                        Randy.

> MDC
>
> Randy Brukardt wrote:
> >
> > Ada's share of non-embedded markets is zero to any reasonable precision
that
> > you want to use. I don't forsee that changing in any significant way
(the
> > $25 I can put into marketing Ada and Claw isn't going to make any
> > difference). The only place it gets used is in larger embedded systems
(like
> > avionics, ATC, etc.) But perhaps you are using some odd-ball definition
of
> > "embedded"??
> >
> >                    Randy.
> >
> >
> >
> >
>
>
> --
> ======================================================================
> Marin David Condic
> I work for: http://www.belcan.com/
> My project is: http://www.jsf.mil/NSFrames.htm
>
> Send Replies To: m   o   d   c @ a   m   o   g
>                     c   n   i       c   .   r
>
>      "Face it ladies, its not the dress that makes you look fat.
>      Its the FAT that makes you look fat."
>
>          --  Al Bundy
>
> ======================================================================
>

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Marin David Condic wrote:

> Keep in mind that I have done a study that showed a 50% reduction in 
> development time and a factor of 4 improvement in defects. You may in 
> fact be thinking of that very study here. I've made the statement here 
> before.

One was a study in the domain of jet-engine control, which I think was 
presented by some guy named Condic. Is there a link to this study? That 
would be a nice thing to have. The other is the Rational/Verdix study in 
the domain of Ada compilers, which is available on AdaIC.org. There are 
other studies also available on the AdaIC site, such as John 
McCormick's. They don't present their results in a comparable way, but 
the general message is the same for all of them.

-- 
Jeff Carter
"I blow my nose on you."
Monty Python & the Holy Grail
03

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Randy Brukardt wrote:

> Ada's only chance for further expansion is in various kinds of embedded
> systems (bigger ones; Ada isn't really suited for single-board computers).

I've long thought there should be an embedded board with something like 
32 processors, about as good as an 80387 w/FPU, each with its own hunk 
of memory, and an Ada compiler that targets the board that distributes 
tasks across processors. Inexpensive, true parallelism should be a big 
advantage for embedded systems.

Someday, when I win the lottery ...

-- 
Jeff Carter
"I blow my nose on you."
Monty Python & the Holy Grail
03

Re: Ada used in General Aviation (GA) applications?

[Simon Wright]

Anders Wirzenius <anders@no.email.thanks.invalid> writes:

> Here is another way of doing early reviewing:
> In early 80's I worked for a Fortran community. My co-programmer and
> I started to use Ada as a pseudo-code language. We made a code
> review of the Ada code and did not start the detailed Fortran coding
> until the review had approved the pseudo-code. The Ada code was left
> in the programs as comments.

I did the same, '87 or so, with assembler (for a Mascot kernel for a
Ferranti F1600-series machine); the Ada code had to compile too,
though there were two points where 'magic occurs here'.

Led to some entertaining rumours about an Ada compiler for the
machine.

-- 
Simon Wright                               100% Ada, no bugs.

Re: Ada used in General Aviation (GA) applications?

[Simon Wright]

"Randy Brukardt" <randy@rrsoftware.com> writes:

> Ada's only chance for further expansion is in various kinds of
> embedded systems (bigger ones; Ada isn't really suited for
> single-board computers).

I guess that depends on the size of the single board! for example,
http://www.radstone.co.uk/products/dsp_product.aspx?id=79&fid=5

You're right, of course, in that it's not for use in set-top boxes or
toasters.

-- 
Simon Wright                               100% Ada, no bugs.

Re: Ada used in General Aviation (GA) applications?

[Peter Amey]

Jeffrey Carter wrote:
> Marin David Condic wrote:
> 
>> Keep in mind that I have done a study that showed a 50% reduction in 
>> development time and a factor of 4 improvement in defects. You may in 
>> fact be thinking of that very study here. I've made the statement here 
>> before.
> 
> 
> One was a study in the domain of jet-engine control, which I think was 
> presented by some guy named Condic. Is there a link to this study? That 
> would be a nice thing to have. The other is the Rational/Verdix study in 
> the domain of Ada compilers, which is available on AdaIC.org. There are 
> other studies also available on the AdaIC site, such as John 
> McCormick's. They don't present their results in a comparable way, but 
> the general message is the same for all of them.
> 

On the subject of defect rates, I think data from the Lockheed C130K 
(Hercules II) project is relevant.  Independent V&V was carried out by 
the UK MoD (the lead customer).  The V&V covered a variety of systems 
produced by a variety of sub-contractors in a variety of languages.  All 
the software inspected had already been cleared to DO178B level A or B.

The results showed that:

1.  Significant errors remained despite the prior FAA clearance.
2.  Code written in C had, on average, 10 times as many errors as that 
written in Ada (and 100 times more than that written in SPARK).
3.  No statistically significant difference in error rate between level 
A versus level B systems could be found.

See: "Software Static Code Analysis Lessons Learned" by Andy German, 
QinetiQ Boscombe Down. DoD CrossTalk Journal, November 2003. 
http://www.stsc.hill.af.mil/crosstalk/2003/11/index.html

and

"Correctness by Construction: Better Can Also Be Cheaper"  (PDF 312kb) 
Peter Amey, Praxis Critical Systems Limited. CrossTalk Magazine, March 
2002.  http://www.praxis-cs.co.uk/pdfs/c_by_c_better_cheaper.pdf

Peter

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

I've seen Ada used by defense contractors in more workstation-ish 
applications doing network communication, etc. That's not really 
"Embedded". In what *is* "embedded", it is typically restricted to the 
larger processors in defense related applications. In most commercial 
stuff and most of the smaller processor arenas (making up most of the 
"embedded" market) Ada gets ranked behind "other" in language used to 
implement. C is typically the language of choice - mostly because that's 
what comes with the development kit. However you slice it, Ada is near 
zero in utilization in the embedded community and is basically hated and 
derided - or simply ignored. My guess is that vendors are not going to 
make much money off of that market and that trying to cater to it is too 
expensive.


I don't know who ACT's customers are, but the last time I checked they 
were not selling a compiler that is aimed at "embedded" systems. They 
market a compiler that mostly wants to run on top of Unix and isn't 
targeted to popular embedded processors. Perhaps it *can* be utilized in 
some embedded/realtime applications, but I don't think they go out of 
their way to try to make a kit to do so. (Perhaps they'd like to bid on 
the next engine control I'll be working on?) So I'm guessing that there 
must be *some* usage of Ada in areas that are not what most people would 
call "embedded".

I think we don't disagree on the fact that Ada is not highly utilized in 
anything outside of the Defense/Realtime market and that this market is 
dwindling. We might even agree as to some of the reasons why Ada is 
losing out. We probably even agree that Ada is like that tree with the 
Dutch Elm Disease - dying but not entirely gone yet.

Where we disagree is in what to do about it. You're convinced it is 
doomed and hopeless - why don't you just give up and spend your time on 
something you *do* think has a future? (I'm not trying to be mean here - 
if you really believe it is hopeless, you would likely be better off 
focusing in on something you *do* believe has a future.) I'm convinced 
that it could be turned around and perhaps Ada could find a market in 
which it could thrive - but it needs a new strategy to do so. So far, I 
don't see the vested interests in Ada calling for a meeting to discuss 
what market they'd like to shoot for and how best to address it with a 
new strategy. (I could probably host it if they asked real nice. :-)

If Ada really *is* better and *does* produce better products at a lesser 
cost, then perhaps it could be a competitive advantage in some market. 
That means that the believers in Ada should try to utilize Ada in some 
manner to build some end-product and sell it. We've got compilers. We've 
got some tools & libraries. (They could obviously be better integrated 
and more standardized) Why not dream up some useful product and go build 
it using Ada? Start *selling* something that uses Ada as a component and 
you generate compiler business to keep the likes of ACT, RR, Aonix, et 
alia, in the Ada-making business.

MDC



Randy Brukardt wrote:
> 
> You said: "The embedded market already hates Ada with the white hot
> intensity of a thousand suns for a variety of historical reasons." And I
> pointed out that the only significant use of Ada is in embedded systems.
> That means either your statement is wrong, or Ada is doomed - in fact it is
> already dead, like a tree with Dutch Elm Disease. The leaves haven't found
> out yet.
> 
> Both RRS and Aonix spent a lot of effort trying to push Ada into more
> mainstream programming. Neither company was successful, and both companies
> have pretty much given up. Besides, non-embedded programming is on its way
> to extinction anyway - those projects and jobs are rapidly moving to very
> low-cost countries. Obviously, quality isn't even on the radar.
> 
> Ada's only chance for further expansion is in various kinds of embedded
> systems (bigger ones; Ada isn't really suited for single-board computers).
> 

-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Simon Wright]

Marin David Condic <nobody@noplace.com> writes:

> Set top boxes actually have pretty large processors & memories. They
> really are not that far off from being a PC in most respects. The last
> one I was programming was using the GCC compiler - so it would not
> have been that hard to target Gnat to it and potentially started a
> whole new market area for Ada. But the intellectual resistance there
> was high and all the related infrastructure was C, C++ or Java
> oriented so it was a tough sell. Had there been a set-top Ada OS, it
> might have been a different story.
> 
> Now toasters? That I can't speak to. :-)

I was thinking more of the price point! a few thousand UKP per card is
more than most of us would want to pay ..

I see from a UK PC magazine that the TiVo (capitalisation?) is
basically a Linux box.

-- 
Simon Wright                               100% Ada, no bugs.

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Marin David Condic wrote:
> Well, check my .sig - I'm "Condic". 

I know. I thought I didn't need to put a smiley on that.

-- 
Jeff Carter
"Death awaits you all, with nasty, big, pointy teeth!"
Monty Python & the Holy Grail
20

Re: Ada used in General Aviation (GA) applications?

[Jeffrey Carter]

Marin David Condic wrote:

> Keep in mind that while we software guys would like to have all sorts of 
> spiffy hardware to make our jobs easier, that the end product is going 
> to be some board whos price is driven by the parts count & complexity. 
> The cost of reproducing a million copies of the software in it are near 
> zero. The cost of 32 processors * 1m boards is pretty non-trivial. 
> Sooner or later someone starts saying "Well do we really *need* that 
> many processors or could the software guys somehow shoehorn it all into 
> one little inexpensive 8 bit processor?" ;-)

Using current-but-one chip technology, it should be possible to put 32 
80387+FPU-equivalent processors with memory on one chip pretty cheaply. 
Ideally the board with these processors would sell for the same price as 
the board with the 8-bit processor.

-- 
Jeff Carter
"Death awaits you all, with nasty, big, pointy teeth!"
Monty Python & the Holy Grail
20

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Well, if you're smart enough about the hardware to have a low-cost 
option, it could be worth following up on as a possible development 
idea. 32 processors for the price of one and you get to program it in 
Ada? Might be an attractive idea to someone....

MDC


Jeffrey Carter wrote:
> 
> Using current-but-one chip technology, it should be possible to put 32 
> 80387+FPU-equivalent processors with memory on one chip pretty cheaply. 
> Ideally the board with these processors would sell for the same price as 
> the board with the 8-bit processor.
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Randy Brukardt]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40A9FBD4.40707@noplace.com...
> You're convinced it is
> doomed and hopeless - why don't you just give up and spend your time on
> something you *do* think has a future? (I'm not trying to be mean here -
> if you really believe it is hopeless, you would likely be better off
> focusing in on something you *do* believe has a future.)

I'm afraid I can't think of anything that does have a future. At least
nothing worth doing (serving hamburgers at McDonalds isn't very exciting).
The software business in general is a waste of time; if you create a useful
product, someone will clone it and give it away as open source (or you could
do that yourself, guaranteeing no return unless you charge robber baron
prices for support). Thus I've stuck with Ada...

                      Randy.

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Well, I hope you find some luck with that. I've always thought that 
accounting software was one of the areas where Open Source could work. 
The basic accounting stuff goes out free, but you charge for the extras 
- like whenever the tax laws change, etc., you sell new modules to the 
supported customers. Get them adicted to using the base package and 
you've got a good user base who may be willing to spend some $$$. 
Certainly Linux is crying out for something like that.

Like anything else, you sometimes need to look at the assets you've got 
on hand and decide how you could best deploy them in some spin-off or 
entirely new area because the existing market is dwindling. The world is 
full of examples where companies and industries had to reinvent 
themselves. Look at cruise ships for one obvious one. Ada is in that 
boat too I think (no pun intended). Its a good language and with some 
new focus on a new market, it might find a way to reinvent itself.

MDC

Randy Brukardt wrote:
> 
> I'm afraid I can't think of anything that does have a future. At least
> nothing worth doing (serving hamburgers at McDonalds isn't very exciting).
> The software business in general is a waste of time; if you create a useful
> product, someone will clone it and give it away as open source (or you could
> do that yourself, guaranteeing no return unless you charge robber baron
> prices for support). Thus I've stuck with Ada...
> 
>                       Randy.
> 
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Jeff C,]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40A9FBD4.40707@noplace.com...

> I don't know who ACT's customers are, but the last time I checked they
> were not selling a compiler that is aimed at "embedded" systems. They
> market a compiler that mostly wants to run on top of Unix and isn't
> targeted to popular embedded processors. Perhaps it *can* be utilized in
> some embedded/realtime applications, but I don't think they go out of
> their way to try to make a kit to do so. (Perhaps they'd like to bid on
> the next engine control I'll be working on?) So I'm guessing that there
> must be *some* usage of Ada in areas that are not what most people would
> call "embedded".
>


I dont disagree with your initial points (deleted) but the stuff above is
not entirely correct.

ACT certainly does "make a kit" to be used for embedded systems.
They are the "preferred" vendor for vxWorks (Real-time embedded OS)
they have or had a product for LynxOS (Real-time embdded though admittedly
unix like OS)
They will support (for a fee) proprietary kernels and no executive type
installs.

They target PowerPC which is a popular embedded computer...Perhaps not in
your world...

I have heard rumors of ARM support at various times (although again you are
correct they dont market it for ARM)

As for the "Perhaps they'd like to bid on  the next engine control I'll be
working on?"

I would say...perhaps your right..

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Actually, I'm looking at a new control project that would use PowerPC 
and VxWorks. Current leaning is towards Aonix for it. If ACT had a 
soup-to-nuts offering in that area, I'd want to know about it - but I'm 
not the ultimate customer. (I can recommend and maybe specify in the 
proposal, but the customer has to bob their head up and down. :-)

It isn't just a matter of having a compiler that targets an embedded 
processor/OS and cross-compiles though. For one thing, most of the 
systems I'm dealing with don't have an OS, so you need a compiler that 
can go right down to the bare silicon & comes with its own suitable RTK. 
  For another thing, it can't just be a compiler that happens to 
generate the right code for the machine and getting all the pieces 
married up is left as an exercise for the student. The schedule I'm 
looking at just plain won't allow for any messing around with cobbling 
together pieces. I need something where I can take it out of the box, 
plug it in and get started putting code in my machine.

I don't know if ACT offers something that well integrated. I just know 
that in a prior life when I had been asking ACT about providing an 
embedded compiler for a PowerPC project (no OS) they politely sent me 
off to talk to someone else. I rather took that to mean they weren't 
interested in the embedded market much.

Keep in mind also that (in terms of numbers of units sold) most embedded 
machines are pretty much smaller than your garden variety PowerPC. 
People mostly concede that they would have a hard time getting Ada to 
run on things like the HC11, etc. Now Spark? That might work well there.

MDC



Jeff C, wrote:
> 
> 
> 
> I dont disagree with your initial points (deleted) but the stuff above is
> not entirely correct.
> 
> ACT certainly does "make a kit" to be used for embedded systems.
> They are the "preferred" vendor for vxWorks (Real-time embedded OS)
> they have or had a product for LynxOS (Real-time embdded though admittedly
> unix like OS)
> They will support (for a fee) proprietary kernels and no executive type
> installs.
> 
> They target PowerPC which is a popular embedded computer...Perhaps not in
> your world...
> 
> I have heard rumors of ARM support at various times (although again you are
> correct they dont market it for ARM)
> 
> As for the "Perhaps they'd like to bid on  the next engine control I'll be
> working on?"
> 
> I would say...perhaps your right..
> 
> 
> 
> 


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[Georg Bauhaus]

Randy Brukardt <randy@rrsoftware.com> wrote:
 
: The software business in general is a waste of time; if you create a useful
: product, someone will clone it and give it away as open source

Do you mean "at no cost"? (which is close to what Microsoft is doing with
compilers and libraries for the wuold-be non-commercial users,
and certainly what they are doing with mediaplayer (lamp for free,
you can buy oil)

: (or you could
: do that yourself, guaranteeing no return unless you charge robber baron
: prices for support).

Is there noone willing to pay smaller sums for bug fixing and enhancements?

Re: Ada used in General Aviation (GA) applications?

[Jeff C,]

"Marin David Condic" <nobody@noplace.com> wrote in message
news:40AC983B.10208@noplace.com...
> Actually, I'm looking at a new control project that would use PowerPC
> and VxWorks. Current leaning is towards Aonix for it. If ACT had a
> soup-to-nuts offering in that area, I'd want to know about it - but I'm
> not the ultimate customer. (I can recommend and maybe specify in the
> proposal, but the customer has to bob their head up and down. :-)
>

ACT does have an integrated product (including complete tornado integration)
for vxWorks.
As for an OS'less type approach...I would guess that you are correct that
there are better choices.

Re: Ada used in General Aviation (GA) applications?

[Simon Wright]

Marin David Condic <nobody@noplace.com> writes:

> Actually, I'm looking at a new control project that would use
> PowerPC and VxWorks. Current leaning is towards Aonix for it. If ACT
> had a soup-to-nuts offering in that area, I'd want to know about it
> - but I'm not the ultimate customer. (I can recommend and maybe
> specify in the proposal, but the customer has to bob their head up
> and down. :-)

We are using GNAT with VxWorks. Of course if you're using VxWorks
you're going to be a Wind River customer, so a lot of the
environmental questions come from them ...

ACT offer "integration with Tornado" but I have to say we aren't great
users of it. Most of our code is developed & tested under Windows, not
my choice but we move on, and choosing the target to build for is
pretty painless within either IDE (a little bit more hands-on with the
Emacs-based GLIDE).

On the command line it's

  gnatmake -Pmy_project

vs

  powerpc-wrs-vxworks-gnatmake -XTARGET=target -Pmy_project

(of course we had to spend some time getting those project files to
recognise the build target etc!)

To me it seems pretty painless. Of course people accessing memory
after it's been freed is always going to be an issue, most especially
when the problem only shows up on the target.

-- 
Simon Wright                               100% Ada, no bugs.

Re: Ada used in General Aviation (GA) applications?

[Richard Riehle]

"Jeff C," <jcreem@yahoo.com> wrote in message
news:1cdrc.14916$zw.12231@attbi_s01...
>
> "Marin David Condic" <nobody@noplace.com> wrote in message
> news:40AC983B.10208@noplace.com...
> > Actually, I'm looking at a new control project that would use PowerPC
> > and VxWorks. Current leaning is towards Aonix for it. If ACT had a
> > soup-to-nuts offering in that area, I'd want to know about it - but I'm
> > not the ultimate customer. (I can recommend and maybe specify in the
> > proposal, but the customer has to bob their head up and down. :-)
> >
>
> ACT does have an integrated product (including complete tornado
integration)
> for vxWorks.
> As for an OS'less type approach...I would guess that you are correct that
> there are better choices.
>
Including, I suspect, DDC-I and Green Hills.  Don't know about ICC.  There
are small, struggling compiler publishers out there that specialize in
embedded
systems.   We need to give them an opportunity to compete on these projects
too.   In the case of the three I just named, we know that they all have a
commitment to Ada.  Some of the other companies, that used to be focused
on Ada, have not seemed to demonstrate much more than token commitment in
recent times.    They apparently have "other fish to fry."

Richard Riehle

Re: Ada used in General Aviation (GA) applications?

[Marin David Condic]

Georg Bauhaus wrote:

> 
> Do you mean "at no cost"? (which is close to what Microsoft is doing with
> compilers and libraries for the wuold-be non-commercial users,
> and certainly what they are doing with mediaplayer (lamp for free,
> you can buy oil)
> 
Clearly, they almost give away compilers because they want people to 
develop apps for their OS and thus sell more OS's to the general public. 
Compilers are a lot of work and don't have (proportionally) many 
customers. OS's are something every computer needs.

So the lesson ought to be "Develop the compiler in order to develop 
something of more general usefulness" Perhaps not down to the consumer 
level, but at least some level where the end market is bigger than the 
people who develop software.

A really good example: Develop the compiler to target some embedded 
processor board/kit. Sell the embedded computers for profit and view the 
compiler as a necessary expenditure. One of the general complaints about 
Ada in the embedded world has always been "Even if it is terrific, I 
can't get it with the embedded machine I'm using..." Well, sell them the 
embedded machine! ;-)

That's not the only way to go, but it is an example of the compiler 
being a value-added product that helps you make enough money to keep the 
compiler work alive.


> 
> Is there noone willing to pay smaller sums for bug fixing and enhancements?
> 
Probably, but is it enough to keep the vendor in business? People are 
willing to pay Microsoft for MSVC++ and for quarterly updates, but I'll 
bet it still is a "loss-leader" that is subsidized by Microsoft in order 
to keep the OS sales going.

The trick here is to say "I've got to find a larger market for my goods 
and services in order to keep the business afloat..." The generalized 
market of "software developers" may be willing to spend some money, but 
are there enough of them to keep a dozen different languages and 
hundreds of different vendors in business? The prices are relatively low 
because the supply exceeds the demand and Ada is fighting for a slice of 
that pie against larger competitors. Can Ada continue to be a niche 
player in a not-all-that-profitable market? It either needs a larger 
slice of that market or the vendors have to find a way to branch out.

Let's cogitate here a second.... Microsoft sells compilers in order to 
sell operating systems... Ada has a compiler..... What if Ada had an 
operating system..... Hmmmmm....... Monkey see, Monkey do? :-)

MDC



-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m   o   d   c @ a   m   o   g
                    c   n   i       c   .   r

     "Face it ladies, its not the dress that makes you look fat.
     Its the FAT that makes you look fat."

         --  Al Bundy

======================================================================

Re: Ada used in General Aviation (GA) applications?

[I R T]

"Kai Glaesner" <this.is@void.de> writes:

> With companies like Garmin bringing state-of-the-art avionics down to
> affordable levels it wont take long until a "glass-cockpit" becomes standard
> in GA.
>
> Now I wonder about what role Ada plays and will play in this business
> segment. I hope it's a big one but anyone out there who knows (or can make
> serious guess)?

No, there seem to be two approaches:

1. C / C++
2. Proprietory languages like LUSTRE

LUSTRE seems popular in Europe.

"Verimag is mainly involved in the development of the synchronous language 
LUSTRE, which is the core language of the industrial environment SCADE, 
developped by Esterel-Technologies, and used, in particular by: 

Airbus, for the onboard software of Airbus A340/600 and A380.
Schneider-Electric, for nuclear power plant control software. "

"Choice of Airbus, Audi, Dassault Aviation, Eurocopter, General Motors, 
Hispano-Suiza, Pratt & Whitney, Rockwell Collins, STMicroelectronics, and 
Texas Instruments. 

"Correct-by-construction design is the only practical solution to the problems 
strangling the productivity of embedded software and electronic systems 
developers
."

http://www.esterel-technologies.com/v3/?id=38513

Re: Ada used in General Aviation (GA) applications?

[I R T]

Jeffrey Carter <spam@spam.com> writes:

> I've long thought there should be an embedded board with something
> like 32 processors, about as good as an 80387 w/FPU, each with its own
> hunk of memory, and an Ada compiler that targets the board that
> distributes tasks across processors. 

And in a price , part count sensitive market like embedded systems
who is going to buy this in quantity ?

Re: Ada used in General Aviation (GA) applications?

[I R T]

Simon Wright <simon@pushface.org> writes:

> I was thinking more of the price point! a few thousand UKP per card is
> more than most of us would want to pay ..
>
> I see from a UK PC magazine that the TiVo (capitalisation?) is
> basically a Linux box.

And my netgear router is basically a linux box.

Re: Ada used in General Aviation (GA) applications?

[I R T]

http://www.ghs.com/products/rtos/AdaMULTI_bare_board.html