can one build commercial applications with latest gnat and other licenses related questions...

can one build commercial applications with latest gnat and other licenses related questions...

[klobert]

I am looking for some clear information related with ada compiler licensing.
Is it possible to build commercial applications without releasing the
source code using latest "libre" version of gnat ?
Or is it necessary to buy a $15K commercial license from adacore in 
order to build commercial applications ?
What's the difference in license between gcc ada and gnat ?
What's the difference in functionality between these:
(gnat, gccada3.44, gccada4.1.0 ? )
What happened to the Eclat compiler ?
Is thee any ada compiler under BSD license ?
Thanks for your help.
Klobert.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

klobert wrote:
> I am looking for some clear information related with ada compiler 
> licensing.
> Is it possible to build commercial applications without releasing the
> source code using latest "libre" version of gnat ?

Yes.
If by commercial you mean selling just the binaries to third parties, no.

> Or is it necessary to buy a $15K commercial license from adacore in 
> order to build commercial applications ?

No.
You have a number of compiler vendors offering tools.
However, if you want to use GNAT, and sell just binaries to third parties,
then likely yes.

> What's the difference in license between gcc ada and gnat ?

Many differences.
FSF GNAT is based on a more recent GCC.
Chances are that AdaCore's GCC is based on GCC 3.4.6 in many cases,
and more intensely maintained.


> What's the difference in functionality between these:
> (gnat, gccada3.44, gccada4.1.0 ? )

Bug fixes, ASIS (yes, no, some), varying levels of support for the 
new Ada 2005 features, and more.

> What happened to the Eclat compiler ?

Don't know.

> Is thee any ada compiler under BSD license ?

Not aware of one.

HTH, Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

There is more information on these pages:

http://en.wikibooks.org/wiki/Ada_Programming/Installing
http://www.ada-france.org/debian/debian-ada-policy.html#The-variants-of-GNAT

As Georg hinted, do not confuse "commercial" and "proprietary". These
are orthogonal concepts; you can make your software:

- commercial (costs money) and proprietary (e.g. Oracle)
- non-commercial (i.e. no cost) and proprietary (e.g. Java, Adobe
Reader)
- commercial (costs money) and free as in freedom (e.g. GNAT Pro,
MySQL, Qt, Sourcery G++)
- non-commercial (no-cost) and free as in freedom (e.g. GCC, Apache).

If your software is free as in freedom (the latter two cases), you can
use GNAT GPL Edition. Otherwise, you must use either GNAT Pro or GCC
from the Free Software Foundation. MySQL and Qt have similar
arrangements.

You can also make your software proprietary but distribute it in
source-only form, perhaps under non-disclosure agreement. You would
then require your customers to compile your sources for themselves,
using whichever compiler thay have at hand. In that case, you can use
GNAT GPL Edition to develop and test your program: since you are not
distributing binaries, the licensing terms of the GNAT run-time library
do not affect your software.

You can also make commercial, proprietary software that you do not
distribute at all to your customers; instead, you make it available as
a web service, and require your customers to pay a subscription fee for
the use of the service. In this model, you can use GNAT GPL Edition,
since you are not distributing your software at all.

There may be other models that have not yet been discussed on this
list. I think this subject matter is worth a FAQ or a Wiki page
somewhere, if such does not already exist.

HTH

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Stephen Leake]

klobert <klobert@sbcglobal.net> writes:

> I am looking for some clear information related with ada compiler
> licensing.

Then you are probably in the wrong place. The only place to get
reliable information on legal issues is from a lawyer, or from the
compiler vendors. 

But that doesn't stop the rest of us from chiming in :).

> Is it possible to build commercial applications without releasing
> the source code using latest "libre" version of gnat ? 

You are probably not using "commercial" to mean what I think it means.

"commercial" means "making money". So yes, you can make money without
releasing source code that is under the GPL license; you sell the
service of running your code, rather than selling the code or binaries
produced from it.

But you probably meant:

"can I sell binaries without source code, if compiled with GNAT
GPL-2006"

The answer to that is "no", because the binaries will contain the
compiler runtime, which is licensed under the Gnu GPL, which requires
source code distribution.

> Or is it necessary to buy a $15K commercial license from adacore in
> order to build commercial applications ? 

That is one choice. It is highly recommended. Note that you are
actually buying a support contract, and their support is excellent.

> What's the difference in license between gcc ada and gnat ? 

I'll assume by 'gcc ada' you mean the official Free Software
Foundation distribution of the GNAT compiler in the GCC (Gnu Compiler
Collection). And by 'gnat' you mean the public release GNAT GPL-2006.

The gcc ada runtime has the GMGPL license, which means it does _not_
require you to distribute source code with binaries. 

The GNAT GPL-2006 runtime has the GPL license, which does require you
to distribute source code with binaries.

> What's the difference in functionality between these: (gnat,
> gccada3.44, gccada4.1.0 ? ) 

They nominally contain the _same_ compiler, but they are slightly
different because of release cycles.

A major difference is that some of the tools that are in GNAT GPL-2006
are not in gcc ada; notably ASIS, and the debugger may not be
Ada-aware (depending on exactly how your distribution is configured).

Also, the GNAT public releases are more reliable (more thoroughly
tested). That's because the gcc release cycle does _not_ wait for Ada
compiler issues, while the GNAT public release cycle does. That may
become less of an issue over time.

I believe gcc ada 4.1.0 has the same Ada 2005 features as GNAT
GPL-2005, but I'm not sure. gcc ada 3.4.4 has no Ada 2005 features.

> What happened to the Eclat compiler ? 

Never heard of it. There is a SourceForge project for it, but no
downloads. Writing a useful Ada compiler is several man-years of work,
so I doubt that will ever get anywhere.

> Is thee any ada compiler under BSD license ? 

The best list of compilers I know of is at
http://www.adaic.com/compilers/comp-tool.html 

That list would probably not include volunteer efforts.

> Thanks for your help. 

You're welcome. Hope this helps.

-- 
-- Stephe

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> klobert wrote:
> > I am looking for some clear information related with ada compiler
> > licensing.
> > Is it possible to build commercial applications without releasing the
> > source code using latest "libre" version of gnat ?
> 
> Yes.
> If by commercial you mean selling just the binaries to third parties, no.
> 

> > What's the difference in functionality between these:
> > (gnat, gccada3.44, gccada4.1.0 ? )
> 
> Bug fixes, ASIS (yes, no, some), varying levels of support for the new
> Ada 2005 features, and more.

I'd like to add, that AFAIS nothing forbids you to do development with
gnat-2006 (using the full ASIS support for debugging and perhaps
generating source code and then compile the project with gccada and
ship these binaries closed source style to the customer.


> > Is thee any ada compiler under BSD license ?
> 
> Not aware of one.

And its actually not necessary: The Toolchain might be under GPL as it
likes (it's OK in my book that nobody can create a closed source tool
from the compiler), but what is annoying is if the GPL infects any
tool compiled _with_ the compiler. This is less freedom not more.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

Stephen Leake wrote :
> klobert <klobert@sbcglobal.net> writes:
>> What's the difference in functionality between these: (gnat,
>> gccada3.44, gccada4.1.0 ? )
>
> They nominally contain the _same_ compiler, but they are slightly
> different because of release cycles.
>
> A major difference is that some of the tools that are in GNAT GPL-2006
> are not in gcc ada; notably ASIS, and the debugger may not be
> Ada-aware (depending on exactly how your distribution is configured).
>
> Also, the GNAT public releases are more reliable (more thoroughly
> tested). That's because the gcc release cycle does _not_ wait for Ada
> compiler issues, while the GNAT public release cycle does. That may
> become less of an issue over time.
>
> I believe gcc ada 4.1.0 has the same Ada 2005 features as GNAT
> GPL-2005, but I'm not sure. gcc ada 3.4.4 has no Ada 2005 features.

What I gathered by looking at the diffs of the Ada front-end and
library between releases:

GCC 3.4 < GCC 4.0 < GNAT GPL 2005 < GCC 4.1 < GCC 4.2 < GNAT GPL 2006

IIRC, most of the changes between GNAT GPL 2005 and GCC 4.1 are in the
Ada.Containers library. Support for Ada 2005 gradually improves with
each release, and seems to be complete starting with GNAT GPL 2006; at
least that's what AdaCore's marketing leads me to believe. I don't know
where each GNAT Pro release fits in that sequence.

Contrast this with the versions of the back-ends, which affect the
quality of the optimiser:

GCC 3.4.0 < GNAT GPL 2005 (3.4.4 + patches) < GNAT GPL 2006 (3.4.6) <
GCC 4.0 < GCC 4.1 < GCC 4.2

GCC 4.2 is not released yet; it is in stage 3: only bug fixes and no
more new features.

The reason why GNAT GPL is still based on GCC 3.4.x is because GCC 4.0
introduced Tree-SSA, a major change in the internal architecture of the
compiler, which has brought some instability (read: bug boxes). GCC 4.1
corrects most if not all of that instability.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

M E Leypold writes :
> what is annoying is if the GPL infects any
> tool compiled _with_ the compiler. This is less freedom not more.

Less freedom for you, but more for your users. The GPL protects users
from programmers who would restrict their freedom.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> Less freedom for you, but more for your users. The GPL protects users
> from programmers who would restrict their freedom.

But the application programmer *is* the user of the compiler.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

Michael Bode wrotes :
> Ludovic Brenta writes:
>
> > Less freedom for you, but more for your users. The GPL protects users
> > from programmers who would restrict their freedom.
>
> But the application programmer *is* the user of the compiler.

The question was not about the compiler, it was about the run-time
library.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

>> But the application programmer *is* the user of the compiler.
>
> The question was not about the compiler, it was about the run-time
> library.

I thought the problem arises if one instantiates generics. This is
usage of an Ada compiler. I also thought the run-time could be linked
dynamically and thus be distributed separate. For now I'm fine with
gnat 3.15p, but maybe that also doesn't work for me licence-wise as...

...in an related thread there is a discussion of the license for
GtkAda. I just looked at some source files of the CVS Version
(https://libre2.adacore.com/cvsweb/GtkAda/) and they still read:

...
-- As a special exception, if other files instantiate generics from  --
...

If this means you need a written statement signed by Robert Dewar to
be sure you won't be sued by Adacore if you use it in a CSS project,
then probably I better read that Java book on my bookshelf. (I'm not
the one who decides what licence my code has, I'm only the one who
decides what language it is written in.)

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey Creem]

Michael Bode wrote:
> "Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:
> 
> 
>>>But the application programmer *is* the user of the compiler.
>>
>>The question was not about the compiler, it was about the run-time
>>library.
> 
> 
> I thought the problem arises if one instantiates generics. This is
> usage of an Ada compiler. I also thought the run-time could be linked
> dynamically and thus be distributed separate. For now I'm fine with
> gnat 3.15p, but maybe that also doesn't work for me licence-wise as...
> 


This is not correct. If the runtime is GPL. Dynamically linking to it 
really does not help under the strictest interpretations of the GPL. You 
are thinking that some portion of the runtime is LGPL.


It is not. It is GPL with a linking/generics exception (in the case of 
old GNAT <=3.15 releases and FSF derived GCC releases) which is actually 
slightly more permissive than the LGPL.

The runtime is "GPL" in the case of the Libre/adacore GPL releases.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Jeffrey Creem <jeff@thecreems.com> writes:

> This is not correct. If the runtime is GPL. Dynamically linking to it
> really does not help under the strictest interpretations of the
> GPL. 

If dynamic linking constitutes a 'derived work' this would mean that
Microsoft's EULA is in fact much more 'Free' than GPL. You can easily
create a 'derived work' of Microsoft Windows(TM) by writing any
program that uses any Windows DLL. And you don't have to stick
Microsoft's EULA to it, you can even licence it under GPL.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Michael Bode <m.g.bode@web.de> writes:

> ...in an related thread there is a discussion of the license for
> GtkAda. I just looked at some source files of the CVS Version
> (https://libre2.adacore.com/cvsweb/GtkAda/) and they still read:
> 
> ...
> -- As a special exception, if other files instantiate generics from  --
> ...
> 
> If this means you need a written statement signed by Robert Dewar to
> be sure you won't be sued by Adacore if you use it in a CSS project,
> then probably I better read that Java book on my bookshelf. (I'm not
> the one who decides what licence my code has, I'm only the one who
> decides what language it is written in.)

That exactly was my problem here after hearing the quote in
question. 

My _original_ problem was, that ACT doesn't state the copyright
situation on their libre-site, and that the available sources have the
GMGPL header in the ada files but not mentioning of a linking
exception in COPYING or README (compare that with florist (3.15p or
prior), which states the linking exception very clearly in the README,
which would be the traditional place to do). So the actual licensing
is somewhat unclear especially in the recent historical context of ACT
having acquired the habit to change licenses to GPL.

I'll not try to assess this situation yet, until I have done more
research. 

But if you want to have an opportunity to get some ugly ideas on your
own, just diff the "GPL" florist sources distributed from the ACT site
against the the 3.15p florist sources. Don't forget to throw away
Makefile*, INSTALL*, README* and configure* first since they would
have no impact on the license and configure is automatically generated
and accounst for around half the size of the diff.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Jeffrey Creem <jeff@thecreems.com> writes:

> Michael Bode wrote:
> > "Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:
> >
> >>>But the application programmer *is* the user of the compiler.
> >>
> >>The question was not about the compiler, it was about the run-time
> >>library.
> > I thought the problem arises if one instantiates generics. This is
> > usage of an Ada compiler. I also thought the run-time could be linked
> > dynamically and thus be distributed separate. For now I'm fine with
> > gnat 3.15p, but maybe that also doesn't work for me licence-wise as...
> >
> 
> 
> This is not correct. If the runtime is GPL. Dynamically linking to it
> really does not help under the strictest interpretations of the
> GPL. You are thinking that some portion of the runtime is LGPL.

> It is not. It is GPL with a linking/generics exception (in the case of
> old GNAT <=3.15 releases and FSF derived GCC releases) which is
> actually slightly more permissive than the LGPL.



Only that relicensing from LGPL -> GPL is almost impossible (w/o
agreement of _all_ copyright holders, which also would include
contributors of significant patches, so just read that as: impossible)
whereas the substamce of the discussion last hear in this very
newsgroup seems to be that there seems nothing to keep a distributor
to just strip the linking exceaption in a derived work.

I'd be happy to hear otherwise.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Michael Bode <m.g.bode@web.de> writes:

> Jeffrey Creem <jeff@thecreems.com> writes:
> 
> > This is not correct. If the runtime is GPL. Dynamically linking to it
> > really does not help under the strictest interpretations of the
> > GPL. 
> 
> If dynamic linking constitutes a 'derived work' this would mean that
> Microsoft's EULA is in fact much more 'Free' than GPL. You can easily

You're completely right in that respect. That is exactly why a number
of people advocate the following:

 - Use of GPL for tools and applications (that protects the community
   from someone borgifying (embrace and extend) these _programs_ and
   putting out incompatible "value added" versions (which would result
   in a vendor lockin).

 - Use of LGPL, GMGPL or BSD license for libraries. 

The last point is often contested, but one might notice that history
shows that environements (speak libraries) which absolutely preclude
the development of closed source software for certain purposes (think
about an NDA on methods and manufacturing processes your customer
wants you to sign), have a problem to build a non-hobbyist
community. Many of this projects/environments have changed the
license to a more liberal one. Some examples:

 - KDE (Qt)
 - Ocaml
 - Minix :-)

On the other side GPL has become a weapon of marketing in some hands:
I.E. MySQL not only put the client runtime unter GPL after, I think,
version 3) but (rumor has it) also has a most peculiar interpretation
of the GPL, namely that using MySQL specific syntax in one of your
(scripted) programs already constitutes linking.

I leave it to the reader to judge how much of this applies to the
recent license changes surrounding Gnat and the libraries (mostly the
libraries, since there is FSF-Gnat), since I'm not done with my
reasearch yet and also have other things to do.

> create a 'derived work' of Microsoft Windows(TM) by writing any
> program that uses any Windows DLL. And you don't have to stick
> Microsoft's EULA to it, you can even licence it under GPL.


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey Creem]

Michael Bode wrote:
> Jeffrey Creem <jeff@thecreems.com> writes:
> 
> 
>>This is not correct. If the runtime is GPL. Dynamically linking to it
>>really does not help under the strictest interpretations of the
>>GPL. 
> 
> 
> If dynamic linking constitutes a 'derived work' this would mean that
> Microsoft's EULA is in fact much more 'Free' than GPL. You can easily
> create a 'derived work' of Microsoft Windows(TM) by writing any
> program that uses any Windows DLL. And you don't have to stick
> Microsoft's EULA to it, you can even licence it under GPL.
> 

Which is why I indicated that this was under the strictest 
interpretations of the GPL. I am not indicating what would "stand up in 
court" but I am telling you what I believe is the FSF position.

If you take something like the GNU scientific library and dynamically 
link to it would be seen as trying to work around the GPL and not seen 
as trying to meet the intent of the license since if that is what the 
authors wanted, they would have gone with the LGPL for that library.

The question of whether or not it is more or less free than a microsoft 
EULA is irrelevant because the "free" here is the FSF version of free 
which has almost nothing to do with what everyone seems to want it to mean.

FSF free is intended to use copyright law to whenever possible force 
authors to release their software as "free" whenever they release their 
software.

Anyone that thinks that the intention of the GPL is anything other than 
that is totally missing the point of the FSF.

Note this is specifically why the LGPL was changed from library GPL to 
lesser GPL because people were too often just assuming that all 
libraries should use the LGPL and thus there were not enough cases where 
people were forced to release their code under "free" terms.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

M E Leypold wrote:
> Michael Bode <m.g.bode@web.de> writes:


>> If dynamic linking constitutes a 'derived work' this would mean that
>> Microsoft's EULA is in fact much more 'Free' than GPL. You can easily
> 
> You're completely right in that respect.

State that you want the freedom of deriving closed source programs
from free software libraries and everyone knows what you are really
talking about, I think.

-

Do the respective MS EULAs permit modifying the DLL source code and
distributing works based on these modified DLLs using any license of
your choosing?

-

I find it a bit dangerous to say the word "right" in a discussion
of legal issues. That aside, the word "freedom" always has points
of reference. Freedom for whom to do what to what, provided that, etc.
As Ludovic has said, it is important to keep in mind that Free Software
uses some specific points of references when it says what free means
in the GPL.

Playing with the word "free" saying less free or more free without
stating the other parts of the notion of "free" is in fact a well
known omission trick in persuasive rhetoric, used when the speakers
want to fool others - and sometimes themselves.
Ahmadinedjad has been trying this game, too, referring to some
deliberately incomplete definition of freedom as in freedom of press,
when he pointed out that he thinks that western press is in fact not
free.

We are not free to stone someone to death, and we are not free to
distribute closed source software that is derived from a Free Software
library. Fair enough, I'd say.

Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Jeffrey Creem <jeff@thecreems.com> writes:

> Michael Bode wrote:
> > Jeffrey Creem <jeff@thecreems.com> writes:
> >
> >>This is not correct. If the runtime is GPL. Dynamically linking to it
> >>really does not help under the strictest interpretations of the
> >> GPL.
> > If dynamic linking constitutes a 'derived work' this would mean that
> > Microsoft's EULA is in fact much more 'Free' than GPL. You can easily
> > create a 'derived work' of Microsoft Windows(TM) by writing any
> > program that uses any Windows DLL. And you don't have to stick
> > Microsoft's EULA to it, you can even licence it under GPL.
> >

> The question of whether or not it is more or less free than a
> microsoft EULA is irrelevant because the "free" here is the FSF
> version of free which has almost nothing to do with what everyone
> seems to want it to mean.

Not quite. The FSF is offering also the LGPL (and they are using it
for Glibc and the Gcc runtime, BTW. Meditate about that).

Since it's not FSF offering the libraries in question, it's not the FSF
concept of "free" that is at issue here, but the concept of people
offering libraries under GPL.

There is no pointing to the FSF here. After all one (as a library
author) can freely decide between GPL, LGPL and GMGPL. Its their right
to offer libraries under GPL. OK. But they must also stand the
statement that (in a library) that implies less freedom and that it is
their doing not the FSF's. And if one goes further they must also
stand the question why they are doing that.

But the FSF, after all, has nothing todo with the GPL-releases of Gnat.

> Anyone that thinks that the intention of the GPL is anything other
> than that is totally missing the point of the FSF.

No. It would be missing the intentions of folks which release
libraries under GPL.

> Note this is specifically why the LGPL was changed from library GPL to
> lesser GPL because people were too often just assuming that all
> libraries should use the LGPL and thus there were not enough cases
> where people were forced to release their code under "free" terms.

"forced to release". I think I can rest my case here. 

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Marius Amado-Alves]

Remember that if you distribute your work in source code form you are  
not bound by the GPL.

That is, suppose you write program P that depends on GPLed unit G.  
You deliver P in source code form only and without redistributing G  
(in any form) alongside. This distribution is not required to be  
under GPL.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[gshapovalov]

I am sorry for the repetition, but I feel I have to make it clear
again..

Michael Bode wrote:
> ...in an related thread there is a discussion of the license for
> GtkAda. I just looked at some source files of the CVS Version
> (https://libre2.adacore.com/cvsweb/GtkAda/) and they still read:
>
> ...
> -- As a special exception, if other files instantiate generics from  --
> ...

Exactly because of this clause still in the specs and no COPYING
provided and somewhere in some file (not normally expected place) there
was a statement that is is *GPL* I contacted GtkAda developers to
clarify this (when I was repackaging GtkAda for Gentoo). This is the
responce I got:

--------------------------------------------------------------------
From: Arnaud Charlet <charlet@adacore.com>
To: George Shapovalov <george@gentoo.org>
Date: 19.5.06 10.20

George,

All the packages available in libre.adacore.com are available under
the pure GPL license, this includes among others GtkAda and XML/Ada.

Regards,

Arno
------------------------------------------------------------------------

I don't think he could have been any more explicit in his responce. So,
it seems GtkAga and everything coming from AdaCore is pure GPL or
commercial nowadays..
(closely matching the Trolltech situation at the moment AFAICS).

If anybody has any further questions I strongly suggest contacting
AdaCore directly and making another case to clean-up their sources..

George

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Jeffrey Creem <jeff@thecreems.com> writes:

> If you take something like the GNU scientific library and dynamically
> link to it would be seen as trying to work around the GPL and not seen
> as trying to meet the intent of the license since if that is what the
> authors wanted, they would have gone with the LGPL for that library.

Then I would like to hear what the FSF thinks 'using a library' is as
opposed to 'derive work from a library'.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

gshapovalov@gmail.com writes:

>> ...
>> -- As a special exception, if other files instantiate generics from  --
>> ...
>
> Exactly because of this clause still in the specs and no COPYING
> provided and somewhere in some file (not normally expected place) there
> was a statement that is is *GPL* I contacted GtkAda developers to
> clarify this (when I was repackaging GtkAda for Gentoo). This is the
> responce I got:
>
> --------------------------------------------------------------------
> From: Arnaud Charlet <charlet@adacore.com>
> To: George Shapovalov <george@gentoo.org>
> Date: 19.5.06 10.20
>
> George,
>
> All the packages available in libre.adacore.com are available under
> the pure GPL license, this includes among others GtkAda and XML/Ada.
>
> Regards,
>
> Arno
> ------------------------------------------------------------------------
>
> I don't think he could have been any more explicit in his responce. So,
> it seems GtkAga and everything coming from AdaCore is pure GPL or
> commercial nowadays..
> (closely matching the Trolltech situation at the moment AFAICS).

IMO this means they would be actively deceiving users of GtkAda about
the licence in probably any (haven't checked that) source file.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

M E Leypold wrote:
> 
> Jeffrey Creem <jeff@thecreems.com> writes:


>> The question of whether or not it is more or less free than a
>> microsoft EULA is irrelevant because the "free" here is the FSF
>> version of free which has almost nothing to do with what everyone
>> seems to want it to mean.
> 
> Not quite. The FSF is offering also the LGPL (and they are using it
> for Glibc and the Gcc runtime, BTW. Meditate about that).

The FSF released the runtime with its license after much
deliberation, and they emphasize that they prefer plain
GPL, even for libraries.


> There is no pointing to the FSF here. After all one (as a library
> author) can freely decide between GPL, LGPL and GMGPL. Its their right
> to offer libraries under GPL. OK. But they must also stand the
> statement that (in a library) that implies less freedom

From the perspective of the author of a GPLed library, the GPL implies
*more* freedom, *not* less, because library derived works will be freely
available.
Anything else will indeed be against the intent of the GPL which is
stated in the Preamble, "to make sure the software is free for all
its users" (by way of forcing derived works to be GPL, too).

From the FAQ at libre.adacore.com:

                 | ... | GNAT GPL Edition
| Intended users | ... | Free Software developers


-- Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> M E Leypold wrote:
> > Michael Bode <m.g.bode@web.de> writes:
> 
> 
> >> If dynamic linking constitutes a 'derived work' this would mean that
> >> Microsoft's EULA is in fact much more 'Free' than GPL. You can easily
> > You're completely right in that respect.
> 
> State that you want the freedom of deriving closed source programs
> from free software libraries and everyone knows what you are really
> talking about, I think.

Yes, and why not?

 (a) Read that not as closed as in "locking my customer in", but as
     closed as in "my customers want me to sign an NDA (of his
     concepts embodied in the source) and wants to be sure its
     binding".

 (b) Other communites (Ocaml, Qt) don't have a problem with that.


> Do the respective MS EULAs permit modifying the DLL source code and
> distributing works based on these modified DLLs using any license of
> your choosing?

No, but neither would the LGPL and the GMGPL. So why GPL instead of
the other two? There must be an answer to that ... :-).

And that "any license of your choosing" is slightly ironic if I
consider that ACT has just completely stripped the linking exceaptions
from an only sligtly changed florist library.

> I find it a bit dangerous to say the word "right" in a discussion

Talking about "legal issues" is always dangerous. It's in society like
"nuclear" in physics. Don't touch it, for experts only. 

But don't fear. The counterspell IANAL, IANAL effectively protects you
from the danger the discussion of your rights inevitably incurs. Or so
I'm told :-).

(Please add more smileys here where needed. YMMV. Ahem.)

> of legal issues. That aside, the word "freedom" always has points
> of reference. Freedom for whom to do what to what, provided that, etc.

Dependency. If I rely on situation X now, who has the power to change
it and what's the cost for me to deal with the change.

> As Ludovic has said, it is important to keep in mind that Free Software
> uses some specific points of references when it says what free means
> in the GPL.

Hello ... There is also LGPL. Why is that so? No point of reference?


> Playing with the word "free" saying less free or more free without
> stating the other parts of the notion of "free" is in fact a well
> known omission trick in persuasive rhetoric, used when the speakers
> want to fool others - and sometimes themselves.

Yes.

> Ahmadinedjad has been trying this game, too, referring to some
> deliberately incomplete definition of freedom as in freedom of press,
> when he pointed out that he thinks that western press is in fact not
> free.

> We are not free to stone someone to death, and we are not free to

Strange comparison. Rethorical, eh?

> distribute closed source software that is derived from a Free Software
> library. Fair enough, I'd say.

LGPL. LGPL. LGPL.

I think you're missing some points. But never mind. _This thread was
about GPL Gnat and I dare say that the right of reverting to GPL was
part of the agreement between ACT and the FSF when GNat was integrated
into GCC. Since there is another GMGPL Ada compiler now the discussion
about GPL Gnat is indeed mood. I don't care.

But about the libraries I completely disagree, but it's hard to count
"moral arguments" of the rather all inclusive type like your last one
(free, fairness, freedom free). I dare say if one follows an argument
based on a definition that freedom is absence of lock-ins and
dependencies (on the future well meaning of vendors / copyright
holders / whatever) and ask where and when dependencies arise, one
would come to a different result.

After all, as I said, the world is not black and white and the
question is not GPL or not, but there are also GMGPL and LGPL.

If, of course you think, that "free" in any sense only pairs well with
"free" -- like your above arguments seems to indicate -- I'd suggest
that futureapps.de (Your Employer / company?) would stop using free
Linux and free Apache to server their business web pages thus unfairly
and unfreely commercially profiting (boo, boo) from the free software.

I think that rather indicates that tool use must be distinguished from
modifying software (a derived work). Libraries are somewhat in between
and it is open to discussion inhowfar _using_ a library make a derived
work from your program. Is i.e. my source compiled at AIX a derived
work of the AIX libc? Hardly. 

It even becomes more absurd if we're talking about the compiler
runtime: Compiling with GPL Gnat make your excutable a derive work of
the GPL Gnat. Sounds like a legal trick to me, nothing else. Not the
spirit of the GPL, rather the letter. But as I said: Discussions about
GPL Gnat are useless, there are alternatives. 

(Discussions about GPL vs. LGPL vs. GMGPL as library license on the
other other side are rather necessary, I think).

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

Michael Bode wrote:

> I would like to hear what the FSF thinks 'using a library' is as
> opposed to 'derive work from a library'.

"derivative work" and "use" are legal terms, best explained by
legal professionals.

The discussions of whether or not which constitutes what have been
lengthy, and IIRC usually suffered from mixing everyday notions
of "use", "derive", etc., with their relevant meaning in licensing
contracts.



-- Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

gshapovalov@gmail.com writes:

> I am sorry for the repetition, but I feel I have to make it clear
> again..
> 
> Michael Bode wrote:
> > ...in an related thread there is a discussion of the license for
> > GtkAda. I just looked at some source files of the CVS Version
> > (https://libre2.adacore.com/cvsweb/GtkAda/) and they still read:
> >
> > ...
> > -- As a special exception, if other files instantiate generics from  --
> > ...
> 
> Exactly because of this clause still in the specs and no COPYING
> provided and somewhere in some file (not normally expected place) there
> was a statement that is is *GPL* I contacted GtkAda developers to
> clarify this (when I was repackaging GtkAda for Gentoo). This is the
> responce I got:

Ah. thanks. That's the first response I see to my question which might shed
some light om the issue and is not based on generic legal IANAL-advice
or rumours.



> 
> --------------------------------------------------------------------
> From: Arnaud Charlet <charlet@adacore.com>
> To: George Shapovalov <george@gentoo.org>
> Date: 19.5.06 10.20
> 
> George,
> 
> All the packages available in libre.adacore.com are available under
> the pure GPL license, this includes among others GtkAda and XML/Ada.
> 
> Regards,
> 
> Arno


I don't think that's specific enough. He ist just talking about
"package". The distributables in question are (for me):

 - GtkAda 2.4.0, Source as from libre (no license on site).
 - GtkAda 2.4.0, Win32 executables from libre (no license on site).

 - GtkAda 2.2.1, as advertised on freshmeat (as GMGPL!) and still served 

 - GtkAda Sources, various versions, as distributed via anon CVS from libre.

 - Finally: The single files in all versions above which carry a GMGPL
   copyright header. If one where to strip all build mechanisms from
   the original GtkAda and just use the *.ads and *.adb files to build
   a new binding: Would that be GMGPL?

The question still is: Was GtkAda ever under GMGPL? In which sense?
What was the last GMGPL version (ACT doesn't document that very clearly).


> I don't think he could have been any more explicit in his responce. So,

He could have been more specific.

> it seems GtkAga and everything coming from AdaCore is pure GPL or
> commercial nowadays..

> (closely matching the Trolltech situation at the moment AFAICS).

Only that Trolltech AFAIR has a linking exception for QT? 

> If anybody has any further questions I strongly suggest contacting
> AdaCore directly and making another case to clean-up their sources..

I might just do that. I'll keep all informed.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

gshapovalov@gmail.com wrote:

>> ...
>> -- As a special exception, if other files instantiate generics from  --
>> ...
> 
> Exactly because of this clause still in the specs and no COPYING
> provided and somewhere in some file (not normally expected place) there
> was a statement that is is *GPL* I contacted GtkAda developers to
> clarify this (when I was repackaging GtkAda for Gentoo). This is the
> responce I got:
> 
> --------------------------------------------------------------------
> From: Arnaud Charlet <charlet@adacore.com>
> To: George Shapovalov <george@gentoo.org>
> Date: 19.5.06 10.20
> 
> George,
> 
> All the packages available in libre.adacore.com are available under
> the pure GPL license, this includes among others GtkAda and XML/Ada.
> 
> Regards,
> 
> Arno
> ------------------------------------------------------------------------
> 
> I don't think he could have been any more explicit in his responce. So,
> it seems GtkAga and everything coming from AdaCore is pure GPL or
> commercial nowadays..

I don't think his words are suitable for legal determinism, for a
number of reasons.
Among these are: what is "package", does "*availabble* under the pure
GPL" imply "available under nothing else",  etc. Does relevant law
say something about good faith and the exception clause etc. etc.. 
I'm not saying that he has not meant that all AdaCore software is
available to non-customers at libre only under the terms of the GPL,
without exception.

These words should be spoken and dealt with by lawyers, if there
is really a need. Maybe AdaCore employees are really nice people
and won't hunt you if you do something wrong legally, and sort
things out in a satisfactory manner.
The readme.txt file in the GPL edition of AWS still has the sentence
"AWS is distributed under the GMGPL (GNAT Modified GPL) license."
I don't think it means a thing, but it might be a documentation bug.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:

> The question still is: Was GtkAda ever under GMGPL? In which sense?
> What was the last GMGPL version (ACT doesn't document that very clearly).

And how can that be proofed, when Adacore denies what they write in
the source files?

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> M E Leypold wrote:

> Anything else will indeed be against the intent of the GPL which is
> stated in the Preamble, "to make sure the software is free for all
> its users" (by way of forcing derived works to be GPL, too).

Well. We're still talking about a compiler here? Than -- Gnat will
stay free for all it's users even if one would write closed source
software with it.

That one can _buy_ the freedom to write unfree software seems to
specially obnoxious if we stay with the moral argument that it is good
to write free software and that one must be forced to write only free
software with a free tool. So my users freedom has a simple price tag:
15K/year. Cheap that, actually.

Let's put it plain: ACT is using GPL as business weapon. If the ACT
compiler would be the only available one at the market I'd call their
Acdemic offer "bait", since you could only exercise the skill set by
generating revenue for them. But since there is another compiler, I'll
now cease arguing about GPL Gnat. That was never my issue, they are
free to do with their compiler what they want. I'm free to use FSF
Gnat or go to another language.

The library situation (not the Gnat GPL runtime, whose GPL licensing
is only a legal trick, but I'm talking about Florist and GtkAda,
perhaps other libs as well) is another issue and a repeat: It's
hursting the community. All Ada advocacy is in vain, if there are no
(commercially usable) OS bindings or bindings to a common and portable
graphical toolkit. People (small businesses, either software
developers or small engineering companies for their internal tools)
will use what is available without strings attached. No use pining in
c.l.a about the abundance of badly written, unsecure C and C++
software and how much better Ada could do, then.

Some other communites also have the problem of "being forced" to write
free software. Others don't. I wonder how the last kind can cope, if
it is necessary to lock tool use and library use down to "only with
free and open software".

And please note, that I'm an advocate of free software. But I can
hardly justify contributing back to the community in my working time
if the license situation excludes me from using (i.e.) libraries in a
project. I even don't complain about that, if a library has been GPL
from the beginning. About what I really complain, is that libraries
seem to change there license and underhandedly at that.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Michael Bode <m.g.bode@web.de> writes:

> Jeffrey Creem <jeff@thecreems.com> writes:
> 
> > If you take something like the GNU scientific library and dynamically
> > link to it would be seen as trying to work around the GPL and not seen
> > as trying to meet the intent of the license since if that is what the
> > authors wanted, they would have gone with the LGPL for that library.
> 
> Then I would like to hear what the FSF thinks 'using a library' is as
> opposed to 'derive work from a library'.

Good question. The FSF (whose point of view actually has zero impact
on the discussion about GPL software from ACT, since it's not the FSF
forcing ACT in any way to use GPL ...) has, with regard to the GPL,
probably the opinion that linking means "derive a work". 

I, personally think, that has historic reasons. When the GPL was
conceived, dynamic linking hardly existed. The LGPL has been
formulated as a response to situations like those that arouse with
bison: Essentially we have program generator, but since it introduces
pieces of own code (read runtime) into the generated code, that leads
to contamination of everything produced with bison.

The FSF has IMHO not addressed the issue of merely using an interface,
which is already present at the target system (like the libc) not even
in GPL 3. I think, that should constitute a case where the executable
doesn't fall under GPL automatically. Saying that using a machine
readable interface description (like C header files) already makes a
derived work is obviously absurd.  Some rudiments of that kind of
thinking can actually be found in the exception to distribution clause
(don't have do distribute stuff already present at the target system).

Some other examples if anybody actually wants to argue against:

  - I taught a number of people Software Engineering. My texts were
    available under the FDL (Free documentation license). Should I
    have made GPL from that? Everything those people produce in future
    is a derived work of my knowledge.

  - When compiling on a linux system, The libc headers indirectly
    include (at compile time) headers from //usr/include/linux which
    AFAIS come from the Linux source tree, which is under GPL. Does
    that mean, every excutable compiled on a Linux system is under
    GPL?

I could continue her, but (wether there is provision for that in the
GPL or not), I'd suggest that for sanities sake there should be
limits: "Use" vs. "Creating a derived Work". For everybody who has
studied computer since, the difference should be clear: Interface
vs. Implementation. That is not the GPL philosophy, but that is why
the author of any library or tool has the option to put runtime or
library under LGPL or add a linking exception. 

Those who don't want that, are free (!) to not do that. They are even
free not to release anything BTW. But if they choose GPL over LGPL or
GMGPL or whatever I don't want to hear all that stuff about "more
freedom". It's not, not even in the sense of the GNU manifesto
(reducing dependencies ...!). It's an attempt (completely legal) to
lock a library away from commercial use.

I, personally, doubt the value of forced contributions of free
software to the community. Have a look to the BSD communities for a
comparison.

And still I'm for free software and for the GPL (in the right
context). Mind that.


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> Michael Bode wrote:
> 
> > I would like to hear what the FSF thinks 'using a library' is as
> > opposed to 'derive work from a library'.
> 
> "derivative work" and "use" are legal terms, best explained by
> legal professionals.
> 
> The discussions of whether or not which constitutes what have been
> lengthy, and IIRC usually suffered from mixing everyday notions
> of "use", "derive", etc., with their relevant meaning in licensing
> contracts.

IANAL, IANAL, IANAL.

I'd also add, that "license" is AFAIK (IANAL) no legal notion in some
countries. I have been informed that it is something that exists in US
law but as thus and under this name not in German law (i.e.). There is
still copyright and the GPL (for instance) has binding force, because
of copyright law, but the word license is not found in the BGB. My
impression (IANAL) was that whatever corresponds to licensing during
distribution in Germany doesn't fall under contract law, but under
copyright law. But IANAL.

Regards .. Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Michael Bode <m.g.bode@web.de> writes:

> M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes
:
> 
> > The question still is: Was GtkAda ever under GMGPL? In which sense?
> > What was the last GMGPL version (ACT doesn't document that very clearly).
> 
> And how can that be proofed, when Adacore denies what they write in
> the source files?

:-] 

That actually was my acute fear when I read that Rober Dewar quote. It
was not my intention to trigger such a lengthy and heated discussion
on licenses (of all things), but only to get some preliminary
information (and pointers to perhaps documents which reflect license
changes) here before writing to ACT.

Obviously though, the situation is actually murky and my initial
confusion is not my fault. That's comforting. The actual situation
becomes less and less comforting, but never mind that for now.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

M E Leypold wrote:
> Georg Bauhaus <bauhaus@futureapps.de> writes:


>> State that you want the freedom of deriving closed source programs
>> from free software libraries and everyone knows what you are really
>> talking about, I think.
> 
> Yes, and why not?

The GPL Preamble states the intent that software derived from GPL
sources be GPLed. I think, yes, this could be called a lock.
So probably not a good choice for you and your customers, if for some
reason some third party would have the right to obtain your sources,
and you don't want that to happen.
What has AdaCore's business to do with that? If you are not an AdaCore
customer...

It could mean that the wealth of libraries that AdaCore has embraced
will no longer be available to the community of small budget, closed
source businesses. I'm not happy about this, but this is just business
(business tends to have its preferences frequently narrowed down
to what is necessary to get through the next few years. That's
unlike Ada...)


>  (b) Other communites (Ocaml, Qt) don't have a problem with that.

Neither does AdaCore have a problem with closed source applications,
nor does the FSF GNAT, and some libraries. OCaml is publicly funded.
Qt uses a dual licensing model, too, AFAICS.

Besides, the Ada community is not named AdaCore, they have their own
community. Which adds weight to your argument that Ada libraries
might become somewhat dependent on GNAT GPL. They had already been
dependent on GNAT's language extension 'Img.


>> Do the respective MS EULAs permit modifying the DLL source code and
>> distributing works based on these modified DLLs using any license of
>> your choosing?
> 
> No, but neither would the LGPL and the GMGPL. So why GPL instead of
> the other two? There must be an answer to that ... :-).

The answer is: GNAT GPL is intended for Free Software programmers.
The GAP is good for AdaCore and their customers because they can,
for example, find future employees in universities. (You call it
"bait" in another reply.)

>> As Ludovic has said, it is important to keep in mind that Free Software
>> uses some specific points of references when it says what free means
>> in the GPL.
> 
> Hello ... There is also LGPL. Why is that so? No point of reference?

As I said elsewhere, this was discussed and the FSF preference is still
this: avoid the LGPL, because it does not help spreading Free Software.
The main reference is freedom for others, as Ludovic has said. 
The decision to add an exception to FSF GCC libraries enumerates its points
of reference. You will find an explanation why LGPL is not preferred.
GNAT GPL Edition agrees with the preference of the FSF.


What other producers of non-GNU libraries do is their business.
For example, the EFL offers a different approach.

 
>> distribute closed source software that is derived from a Free Software
>> library. Fair enough, I'd say.
> 
> LGPL. LGPL. LGPL.

EFL, BSD, etc. So what? If it is AdaCore's choice to use GPL,
who could stop them? I think that your business's need for a GMGPL
library won't, because your business is not their business.


> I think you're missing some points.

I don't think I'm missing GPL points here.

> I dare say if one follows an argument
> based on a definition that freedom is absence of lock-ins and
> dependencies (...) and ask where and when dependencies arise, one
> would come to a different result.

The GPL is a deliberate choice that deliberately ignores arguments
that (have to) prefer independence of the GPL terms.



> If, of course you think, that "free" in any sense only pairs well with
> "free" -- like your above arguments seems to indicate -- I'd suggest
> that futureapps.de (Your Employer / company?) would stop using free
> Linux and free Apache to server their business web pages thus unfairly
> and unfreely commercially profiting (boo, boo) from the free software.

In fact, we did and do contribute to free software, incl. some money.
Both because it's fair, and because it helps our business. It also
tends to follow open standards. Obviously, we haven't contributed
in a big way. We also use GPL software where we can. 
We can't always, you describe the situation as known to you. But we
don't complain that makers of Free Software disallow use of their GPL
tools and components for closed source software. If you need that,
pay them, if you can.


> Compiling with GPL Gnat make your excutable a derive work of
> the GPL Gnat. Sounds like a legal trick to me, nothing else.

This is just how a license works, and is in accord with the
intent or spirit of the GPL: promote Free Software. The GCC docs
explain why they placed exceptions on the FSF GCC libraries, and
also why this is an exceptional exception.

Unfortunately, the great majority of business entities have been
educated to take advantage of anyone and anything.
This can certainly be in the way of seeing the Free Software model 
as economically viable in many cases, and from many perspectives.

-- Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[george]

(Sorry if this comes as a dupe - the session died while posting. Does
anybody know of a public news server that allows to post?)

M E Leypold wrote:
> I don't think that's specific enough. He ist just talking about
Then just write to him and pose all these questions ;). I am not the
right person to ask as I am technically "just spreading rumors" :).

> "package". The distributables in question are (for me):
>
>  - GtkAda 2.4.0, Source as from libre (no license on site).
This is the one I was referring to in my email, so you may take the
meaning of that responce as concerning this version.

>  - GtkAda 2.4.0, Win32 executables from libre (no license on site).
The GPL talks about distribution of sources and binaries, so whatever
is produced from above sources should be covered (it is a derivative
work technically)..


>  - GtkAda 2.2.1, as advertised on freshmeat (as GMGPL!) and still served
>
>  - GtkAda Sources, various versions, as distributed via anon CVS from libre.
These you should ask about.

>  - Finally: The single files in all versions above which carry a GMGPL
>    copyright header. If one where to strip all build mechanisms from
>    the original GtkAda and just use the *.ads and *.adb files to build
>    a new binding: Would that be GMGPL?
No, you cannot simply strip all other pointers to legal information or
otherwise. The exception clauses were left there simply because they
did not bother so far to strip them. This is the fragment from his
response to my next email talking about some of this:

----------------------------------------------------------------------------------------
> However I can see the individual source files still have that exception
> clause. (Source was obtained from
> https://libre2.adacore.com/GtkAda/
> couple of days ago). Perhaps it should be removed then

Right, that's a desirable thing to do, which we will do at some point.

> I could sed these lines out of the installed files, but
> 1. I do not think I am legally in position to do this,
> 2. I would rather not touch the sources unnecessarily..

Right, I do not think this is worth doing. The mention inside source
files have
no real legal value anyway.
------------------------------------------------------------------------------------------


> The question still is: Was GtkAda ever under GMGPL? In which sense?
> What was the last GMGPL version (ACT doesn't document that very clearly).
It most definitely was, the change to GPL is recent. However I only
started cleaning up Ada in Gentoo recently, so I cannot comment on
particulars.

> > (closely matching the Trolltech situation at the moment AFAICS).
>
> Only that Trolltech AFAIR has a linking exception for QT?
No, it does not, the situation is very much the same now. In fact I
suspect this transition to GPL may have been influenced by Trolltech's
example (they also had somewhat obscure licensing scheme in the
beginning, but transitioned to GPL/commercial as of some years ago).

> > If anybody has any further questions I strongly suggest contacting
> > AdaCore directly and making another case to clean-up their sources..
>
> I might just do that. I'll keep all informed.
Right, I hope they finally clean the sources up and we don't have these
questions coming up every month, like t is now :).

George

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

George Shapovalov writes:
> M E Leypold wrote:
>> > If anybody has any further questions I strongly suggest contacting
>> > AdaCore directly and making another case to clean-up their sources..
>>
>> I might just do that. I'll keep all informed.
> 
> Right, I hope they finally clean the sources up and we don't have these
> questions coming up every month, like t is now :).

I have sent them an email and I'm awaiting a response, which I will
use for the future Debian packages.

Please feel free to ask for yourself, too.  Receiving many emails
asking the same question should prompt them to clarify the situation
and clean up the confusion they've created.  And at the same time,
that'll show them that there are people out there who really care.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"george" <gshapovalov@gmail.com> writes:

> (Sorry if this comes as a dupe - the session died while posting. Does
> anybody know of a public news server that allows to post?)
> 
> M E Leypold wrote:
> > I don't think that's specific enough. He ist just talking about
> Then just write to him and pose all these questions ;). I am not the
> right person to ask as I am technically "just spreading rumors" :).

I will. Nonetheless, your answer was useful. As I explained in another
post, I just wanted to ask c.l.a wether there are any
sources/documents on the licensing question I have overlooked (like
"That articleof Jan 2005 from Robert Dewar which clarified all this
questions once and for all" :-)). It seems that not.

> 
> > "package". The distributables in question are (for me):
> >
> >  - GtkAda 2.4.0, Source as from libre (no license on site).
> This is the one I was referring to in my email, so you may take the
> meaning of that responce as concerning this version.
> 
> >  - GtkAda 2.4.0, Win32 executables from libre (no license on site).

> The GPL talks about distribution of sources and binaries, so whatever
> is produced from above sources should be covered (it is a derivative
> work technically).

Covered by what? That's the question. I mean: There is a strong
indication that GtkAda at least _was_ once GMGPL and exactly that
source is available from libre via CVS. So "all packages", whatever is
meant by "package" seems to be a bit too inclusive.

Also the 2.4.0 source still have the same COPYING and file headers as
the older GTkAda sources, so why they are suddenly GPL?


> >  - GtkAda 2.2.1, as advertised on freshmeat (as GMGPL!) and still served
> >
> >  - GtkAda Sources, various versions, as distributed via anon CVS from libre.
> These you should ask about.
> 
> >  - Finally: The single files in all versions above which carry a GMGPL
> >    copyright header. If one where to strip all build mechanisms from
> >    the original GtkAda and just use the *.ads and *.adb files to build
> >    a new binding: Would that be GMGPL?

> No, you cannot simply strip all other pointers to legal information or
> otherwise. The exception clauses were left there simply because they
> did not bother so far to strip them. This is the fragment from his
> response to my next email talking about some of this:

I did not talk about "stripping pointers to legal information". I
talked about picking single files, which IMHO and as such (but IANAL)
fall under the GMGPL from the supposedly GPLed GtkAda package.


> ----------------------------------------------------------------------------------------
> > However I can see the individual source files still have that exception
> > clause. (Source was obtained from
> > https://libre2.adacore.com/GtkAda/
> > couple of days ago). Perhaps it should be removed then
> 
> Right, that's a desirable thing to do, which we will do at some point.
> 
> > I could sed these lines out of the installed files, but
> > 1. I do not think I am legally in position to do this,
> > 2. I would rather not touch the sources unnecessarily..
> 
> Right, I do not think this is worth doing. The mention inside source
> files have
> no real legal value anyway.

Again that fiction. We will see.


> > The question still is: Was GtkAda ever under GMGPL? In which sense?
> > What was the last GMGPL version (ACT doesn't document that very clearly).

> It most definitely was, the change to GPL is recent. However I only
> started cleaning up Ada in Gentoo recently, so I cannot comment on
> particulars.

Well, well. (Shaking my head).


> 
> > > (closely matching the Trolltech situation at the moment AFAICS).
> >
> > Only that Trolltech AFAIR has a linking exception for QT?

> No, it does not, the situation is very much the same now. In fact I
> suspect this transition to GPL may have been influenced by Trolltech's
> example (they also had somewhat obscure licensing scheme in the
> beginning, but transitioned to GPL/commercial as of some years ago).

> > > If anybody has any further questions I strongly suggest contacting
> > > AdaCore directly and making another case to clean-up their sources.


> > I might just do that. I'll keep all informed.


> Right, I hope they finally clean the sources up and we don't have these
> questions coming up every month, like t is now :).

Dis we have the GtkAda wuestion coming up already? I don't had the
impression. I'm reading c.l.a. now for quite some time and must have
missed that. Especially since Debian distributes GtkAda 2.4.0 as GMGPL
and 2.2.1 is still advertised at freshmeat as GMGPL I can't imagine
that the licensing of GtkAda has been discussed excessively yet.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> > If, of course you think, that "free" in any sense only pairs well with
> > "free" -- like your above arguments seems to indicate -- I'd suggest
> > that futureapps.de (Your Employer / company?) would stop using free
> > Linux and free Apache to server their business web pages thus unfairly
> > and unfreely commercially profiting (boo, boo) from the free software.
> 
> In fact, we did and do contribute to free software, incl. some money.
> Both because it's fair, and because it helps our business. It also
> tends to follow open standards. Obviously, we haven't contributed
> in a big way. We also use GPL software where we can. We can't always,
> you describe the situation as known to you. 

> But we don't complain that makers of Free Software disallow use of
> their GPL tools and components for closed source software. If you
> need that, pay them, if you can.


I didn't do that, and I'd rather you took that back. If (which I find
doubtful) tried to follow my arguments you'll find that 

 (a) I never doubted AdaCores right to put Gnat GPL under GPL. 

 (b) Considered it harmful to the community (that is the widespread
     use of Ada as alangauge in areas presently occupied by
     C/C++/Java) that there are perhaps (which is still usbject to
     research) no modern and commerically viable Bindings to graphical
     toolkit.

 (c) Answered to you morally founded argument that users of free tools
     should produce free software that 

The answer that I'm only "complain that makers of Free Software
disallow use of their GPL tools (..)" goes a bit far. Apart from the
fact that Gnat is based on GCC which doesn't have that usage
restriction, I seriously doubt that anybody here has been aware of the
GtkAda licensing change before. (which is the primary point in (b)).

I'm not complaining: I did try to clarify a situation and was met (by
you) with probably irrelevant and rather philosphical arguments about
what should and what should not.

I actively resent being then drawn into a discussion about the
morality of profiting from free software and getting replies that
start with "your business should" all the time. I didn't ask your
advice on that.

I've had a rather longish point to point answer to your last one, but
after I found myself reiteratiing the same points again and again,
decided that it is useless. Since I can't see your point in all of
that, this discussion is ended from my side.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

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

> George Shapovalov writes:
> > M E Leypold wrote:
> >> > If anybody has any further questions I strongly suggest contacting
> >> > AdaCore directly and making another case to clean-up their sources..
> >>
> >> I might just do that. I'll keep all informed.
> > 
> > Right, I hope they finally clean the sources up and we don't have these
> > questions coming up every month, like t is now :).
> 
> I have sent them an email and I'm awaiting a response, which I will
> use for the future Debian packages.
> 
> Please feel free to ask for yourself, too.  Receiving many emails

I will.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Hi George,

> > I dare say if one follows an argument
> > based on a definition that freedom is absence of lock-ins and
> > dependencies (...) and ask where and when dependencies arise, one
> > would come to a different result.
> 
> The GPL is a deliberate choice that deliberately ignores arguments
> that (have to) prefer independence of the GPL terms.

As I said, im not interested in continuing that discussion. But for
the record I find it necessary to state that the "dependencies" I have
been talking about are not "dependencies from the GPL" but
dependencies from vendor decisions. I've been talking about library
licenses specfically and the vendor decision I'm refering to is of
course the (not very well documented) stripping of a linking exception
from the license.

This is not a personal complaint, just the observation, that people
might be a tad surprised after relying on (continued) GMGPL for GtkAda
(continued because a change in free software licensing is rather
unusual) that the license actually changed a long time ago (which
seems not to be common knowledge). Here is a vendor decision that
influences not only the future but also the present standing of a
project (be it commercial or free in any sense). In my book that
spells "dependency". Freedom can be taken back: Not only in the
future, but also in the present.

How that does impact my work should not be your concern, so please
spare me any advice of the kind "I think that your business's need
(...)". I'll cope, but I dare to say that the community will be
slightly surprised. Which illustrates my point rather well.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> But we don't complain that makers of Free Software disallow use of
> their GPL tools and components for closed source software. If you
> need that, pay them, if you can.

George, 

I know I'm not consistent, but I'd like to point you to
http://groups.google.de/group/comp.lang.ada/browse_thread/thread/3e26dfa741e64e5f/bbff925a8e74f167?lnk=st&q=Ada+Whalen+Dewar&rnum=4&hl=de#bbff925a8e74f167

which has a rather good exegesis of Steve Whalen on the point wether
GNAT is ACT's to do with it what they want, and also a relevant rather
lengthy Robert Dewar quote. After reading that you'll perhaps
understand that the situation is not as cut and dried as it seems to
you and also why I resent to being portrayed by you in public as
someone just trying to sponge off ACTs God Given Property
(i.e. Gnat). Gnat is not "their tool". They're licensing the code base
under GPL themselves and the original DOD grants as well as the
history of the last decade always imply some obligation to keep Gnat
free, presumably for all users not only in the "free for free software
sense".

So any complaint of mine in the direction that they didn't follow that
obligation any more, would have been fully justified.

My issue actually was elsewhere (since a the FSF compiler is
available) and than got bogged down in lots of (IMHO inappropriate) GPL
philosophy. Nonetheless, thanks for the accusations.

Regards  -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:

> Georg Bauhaus <bauhaus@futureapps.de> writes:
> 
> > But we don't complain that makers of Free Software disallow use of
> > their GPL tools and components for closed source software. If you
> > need that, pay them, if you can.
> 
> George, 
> 
> I know I'm not consistent, but I'd like to point you to
> http://groups.google.de/group/comp.lang.ada/browse_thread/thread/3e26dfa741e64e5f/bbff925a8e74f167?lnk=st&q=Ada+Whalen+Dewar&rnum=4&hl=de#bbff925a8e74f167

Wrong post. The right one is:

http://groups.google.de/group/comp.lang.ada/browse_thread/thread/24ac770ebf312b7a/e4fdf88d2a6768a6?lnk=st&q=Whalen+Dewar+happen&rnum=1&hl=de#e4fdf88d2a6768a6

I just happen to see that the "wrong quote" actually answers a posting
by you. Since that didn't convince you then (you didn't answer on
Steve Whalens comments), it probably won't convince you now. But
please stop to imply that other people try to rip off ACT :-). If
there is not enough community to get a commercially
also-for-closed-source Ada environment (that means appropriate libs
also) going, all the people that would want that will go elsewhere and
there will be peace in c.l.a. If the community actually can get such
an environment going and maintain it, the better.

But now way are Gnat nor the libraries a one sided gracious gift of
ACT to the unwashed masses. I acknowledge their contribution (I also
acknowledge that the got paid for it).

Another relevant quote is 

http://groups.google.de/group/comp.lang.ada/browse_frm/thread/2c7b0b777188b7c4/5a960e6166cc5f3a?tvc=1&q=Ada+Whalen+Dewar&hl=de#5a960e6166cc5f3a

  "if you weren't reading comp.lang.ada and following GNAT back then,
   please use Google groups and go back and read what the ACT/AdaCore
   people were saying back then.  They took great pains to explain
   that exactly what they are now doing would be "bad"."

I don't have the intention to revive that thread (please not). But
what you accuse me of ("complain") seems to have been a pretty common
occurrence some months ago and as it seems not all out of greed to get
everything for cheap, but out of a rather different understanding of
the implied obligations ACT had to the community.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

> Steve Whalens comments), it probably won't convince you now. But
> please stop to imply that other people try to rip off ACT :-). If
> there is not enough community to get a commercially

That should have read "commercially viable".

> also-for-closed-source Ada environment (that means appropriate libs
> also) going, all the people that would want that will go elsewhere and


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:

> 	Slipping in from the side... A non-support, run-time distribution,
> license in the $200 range would probably get a lot of takers (price
> range varying depending upon upgrade procedure 

I'd be one, let's say somewhere below 1000ￜ for the GPL Edition with
GMGPL licensing. In fact I asked AdaCore for such a beast. Doesn't
exist an probably never will.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:

> My _original_ problem was, that ACT doesn't state the copyright
> situation on their libre-site, and that the available sources have the
> GMGPL header in the ada files but not mentioning of a linking
> exception in COPYING or README (compare that with florist (3.15p or
> prior), which states the linking exception very clearly in the README,
> which would be the traditional place to do).

Just downloaded GtkAda 2.2.1 which listed as GMGPL at freshmeat. The
README says:

Disclaimer
==========

GtkAda  is free software;  you  can redistribute  it and/or  modify it
under terms of the GNU General Public License as published by the Free
Software Foundation; either  version 2, or  (at your option) any later
version.  GtkAda is  distributed in the  hope that it  will be useful,
but WITHOUT ANY   WARRANTY;   without even  the implied  warranty   of
MERCHANTABILITY -- or  FITNESS FOR A PARTICULAR  PURPOSE.  See the GNU
General Public  License for more details.  You  should have received a
copy of the GNU General -- Public License distributed with GtkAda; see
file  COPYING.  If  not,  write to  the  Free  Software Foundation, 59
Temple Place - Suite 330, Boston, -- MA 02111-1307, USA.


Looks like this hasn't changed a bit since version 0.1 of GtkAda. So
the questions is: has there ever been a public GMGPL version of
GtkAda? Has anyone a written legally binding piece of paper signed by
Robert Dewar to prove it? And does he have a proof that Dewar is
authorized to sign such a paper? After all the AUTHORS file says:

AUTHORS
=======

The authors of GtkAda are

Emmanuel Briot <briot@gnat.com>
Joel Brobecker <brobecker@gnat.com>
Arnaud Charlet <charlet@gnat.com>

And by the way: how am I supposed to know if GNAT GPL Edition is
really GPLed software? Did anyone who downloaded it get a piece of
paper signed by Robert Dewar which says so?

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Georg Bauhaus]

M E Leypold wrote:
>

First let my clarify why I got the impression you are in fact
complaining about how the GPL defines "free" (or how issues got
once again nailed down):

 Georg> "State that you want the freedom of deriving closed source
 Georg> programs from free software libraries ..."
 Markus> "Yes, and why not?"

Because from the perspective of the GPL, wanting to derive closed
from GPLed is opposed to the intent of the GPL. The typical way out
is by a contract which involves payment, and an exception.
Unfortunately, also a substantial sum in this case.


> (you didn't answer on
> Steve Whalens comments), it probably won't convince you now.

I didn't answer his comment because  he has good points, and
all sides are arguable endlessly. Things are in constant flux,
obviously, even though from a _formal_ perspective nothing
much has changed (same license, just no more exception, as I
had said).

In fact, I find it sad that there is no more community/public
funding for more "intervendor" Ada stuff.
And it is a pity that AdaCore uses 'Img everywhere something
could be useful without their compiler. But they have said more
than once that they are not in the business of getting Ada 
software compile with other vendors' compilers. They've got
a point, though reading back in Ada history I find this a little
disappointing.


> But
> please stop to imply that other people try to rip off ACT :-).

Well, rip off, no, but sure we are all happily using good software
initially made by a well-funded university spin-off, AFAICT. Of
course they have now done a great deal themselves, who would doubt
this. So if the likes of us need GtkAda GMGPLed now, we'll need to
reinvent that wheel, as we can now profit from AdaCore's software
if and only if we can write GPLed software, or use it in some other
way permitted by the GPL.

In the light of the discussion you found, there were important
arguments that show how some software efforts more or less depend
on what AdaCore does, as a matter of fact. (Many interesting GNAT
projects found a home within AdaCore.) If this dependence weren't
there, there would have been less need to determine the licensing
state of GNAT related Ada software for Debian, for example, as
AdaCore is indeed the party to decide that e.g. AWS is now GPL.

In the light of other discussions you will find a real need for a
GMGPL GNAT low cost edition. Limited support has been suggested,
of the kind that is available from some vendors:

- AdaCore software, components in particular, is just what you need
- AdaCore software usually depends on GNAT
- AdaCore has explained why they choose to have these dependences

So basically, if you choose a more viable compiler (licensing, cost)
together with (now) AdaCore (formerly independent GMGPL) components,
you won't be able to compile anything without making changes to the
components, trivial as they may be. GMGPL or GPL does not matter,
then ...

There were comments stating that other parties could try to
provide this kind of support for GNAT software.
Reportedly, ACT had tried this kind of support contract, too,
and it didn't work well for them.
See also Michael Bode's posting of tonight.



> out of a rather different understanding of
> the implied obligations ACT had to the community.

IIUC, Steve Whalen emphasized that anyway the tax
payer community should have a word in the licensing to the effect that
it be more GMGPL, to be generally useful for those who did originally
pay for it. "Presumably for all users not only in the 'free for
free software sense'", as you have put it. This was about the compiler
and the run time system. I'm all in favor of some more
community based funding of generally useful Ada components.
But the GNAT contract seems to have been about just an Ada compilation
system.

Arguing about GPL AWS, GtkAda and so on is then a very different
understanding of any obligations, alleged or implied, because, formally,
the compiler/run time licensing needn't apply to general purpose software
components.
And yes, obviously this tends to be an occasion for, well, complaints. 

I for one will be happy to join the moaning, even though in my
case this has not so much to do with projects, or money.

-- Georg 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Samuel Tardieu]

>>>>> "MEL" == M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:

MEL> Good question. The FSF (whose point of view actually has zero
MEL> impact on the discussion about GPL software from ACT, since it's
MEL> not the FSF forcing ACT in any way to use GPL ...) has, with
MEL> regard to the GPL, probably the opinion that linking means
MEL> "derive a work".

Most of the GNAT runtime is owned by the FSF. A small part if owned by
AdaCore.

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Georg Bauhaus <bauhaus@futureapps.de> writes:

> M E Leypold wrote:
> >
> 
> First let my clarify why I got the impression you are in fact
> complaining about how the GPL defines "free" (or how issues got
> once again nailed down):
> 
>  Georg> "State that you want the freedom of deriving closed source
>  Georg> programs from free software libraries ..."

>  Markus> "Yes, and why not?"


Sorry, I see nothing mentioning the GPL up to that point. As I
repeatedly pointed out, LGPL and GMGPL also qualify for free. So there
is no contradiction per se in dereiving closed source from free software. 

As an aside I'd like to hint that I see a number of cases where GPL
bindings have been put on Top of LGPL libraries. Its open do debate
wether the intentions of the authors of this LGPL libraries have been
resepected here (legally no problem, morally: I don't know).


> In fact, I find it sad that there is no more community/public
> funding for more "intervendor" Ada stuff.

> And it is a pity that AdaCore uses 'Img everywhere something
> could be useful without their compiler. But they have said more
> than once that they are not in the business of getting Ada software
> compile with other vendors' compilers. They've got
> a point, though reading back in Ada history I find this a little
> disappointing.

I can can agree with you here. But sometimes I find your attitude of
submissiveness to ACTs policies a bit disturbing.

> 
> > But
> > please stop to imply that other people try to rip off ACT :-).
> 
> Well, rip off, no, but sure we are all happily using good software
> initially made by a well-funded university spin-off, AFAICT. 

> Of
> course they have now done a great deal themselves, who would doubt
> this. So if the likes of us need GtkAda GMGPLed now, 

"need ... now" is wrong. We need it to stay GMGPLed. Taking back
licenses is not good. 

> we'll need to reinvent that wheel, as we can now profit from
> AdaCore's software if and only if we can write GPLed software, or
> use it in some other way permitted by the GPL.

Or we migrate to another language where the community hasn't got all
this problems and lock-ins ... -- But as I said elsewhere: Everyone
who recommends this as a solution has no business to bemoan the
decline of Ada as a language nor how bad it is that C/C++/Java/C# get
used in a lot projects despite their deficiencies. 

(May I, as an asidem quote from "Software Engineering with Ada: Ada is
 NOT a Four-Letter Word" by Colin Pratt at
 http://www.chips.navy.mil/archives/92_apr/file3.htm:

 "From a manager's point of view, there is much to be gained by
  settling on one primary language for system development. (...)  Ada
  is the logical choice for a primary language. Suited for business
  oriented systems as well as for C3 and embedded control systems, Ada
  can enable a programming staff to expand its range of support to all
  three disciplines, further capitalizing on valuable technical
  assets. NCTAMS LANT's experience with Ada has been positive."

 Thus where the high hopes with which Ada began. Technically (as a
 language definition) it still has the potential to replace all lesser
 imperative compiled languages (well, almost). Ada does many things
 right, that other languages only do not ore half-heartedly.  Ada also
 has a really readable and understandable RM which certain other
 languages don't.

 But success of a language is not just made by a good and
 undertsnadable language definition but also by availability of
 bindings to surrounding environment. And that is where the situation
 for Ada first looked quite OK and then started to detoriate when a
 lot of bindings to LGPL libraries where put under GPL.

 I hope that this situation can me remedied somehow. As it is, who
 starts with Ada either as a language to develop for others or as a
 language to develop internally, either has the choice to be rstricted
 (from the very beginning) to more simple tools (which perhaps even
 in-house don't meet aesthetical standards of non-tech users) or risk
 being later faced by a choice between a huge entry barrier into the
 market or open source the codebase accumulated until then.
)
 

> In the light of the discussion you found, there were important
> arguments that show how some software efforts more or less depend
> on what AdaCore does, as a matter of fact. 

> (Many interesting GNAT projects found a home within AdaCore.)

And where switched to GPL after some maintenance, disregarding, that a
number of people submitted patches and/or contributions to the
projects in a more liberal license situation. 

> If this dependence weren't there, there would have been less need to
> determine the licensing state of GNAT related Ada software for
> Debian, for example, as AdaCore is indeed the party to decide that
> e.g. AWS is now GPL.


<snipped

> So basically, if you choose a more viable compiler (licensing, cost)
> together with (now) AdaCore (formerly independent GMGPL) components,
> you won't be able to compile anything without making changes to the
> components, trivial as they may be. GMGPL or GPL does not matter,
> then ...

Which "begs the question" (hey, yes, I know :-)) how free are these
libraries then. 

> There were comments stating that other parties could try to
> provide this kind of support for GNAT software.
> Reportedly, ACT had tried this kind of support contract, too,
> and it didn't work well for them.

Well. In a number of other communities (python, ocaml, even php) there
is a level of, let's say, free style community supporr, below the
heavy weight "pay $N-1000/year" contracts. If you look there, you'll
find, that there are a lot of professionals actually making money from
using or even modyfying free software and giving back to the community
by incrementally continuiong devolpment or by debugging or documenting
(both very valuable as I have explained elsewhere).

Admittedly my hope is/was, that the Ada community is not to small to
have/develop this kind of community process. As it is just now, I have
the impression, that the interest group for a GMGPL environment
(i.e. collecting GMGPL source and meintaining it towork with the FSF
compiler) is not large enough and doesn't carry enough man power to do
that work (well, to be open, I can name the activists: Ludovico,
Martin, Jeffrey -- those three are doing that as a hobby, and I could
name myself and are very willing to open as any source as possible --
like the flowing-text-label to GTKAda I developed recently -- but
that's it AFAICS).

> > out of a rather different understanding of
> > the implied obligations ACT had to the community.

> IIUC, Steve Whalen emphasized that anyway the tax
> payer community should have a word in the licensing to the effect that
> it be more GMGPL, to be generally useful for those who did originally
> pay for it. "Presumably for all users not only in the 'free for
> free software sense'", as you have put it. This was about the compiler
> and the run time system. 

Yes, we started with that (since you came repeatedly back to ACTs
right to change the license to "their tools" to GPL), and as I said
already, I don't want to take issue with that. As I see it, ACT has
discharged their obligation to the community by hepling to reintegrate
the GNAT code base into the FSF tree. Since there is now a GMGPL
compiler I don't want to talk about GPL Gnat any more. Let's be silent
about that.

The issue I want to take, that with the present creeping license
changes of a number of libraries and support software from more
liberal licenses to GPL, the licensing situation of the supporting
environment has been, well, sort of borgified: Again "free for free
software only", apart from what you say, that more and more
dependencies on the ACT compiler creep in.

I take issue here, while I admit that ACT probably has even more
(legal) right to change those library licenses than the had the
(moral) right to change the license of Gnat. 

But this is not about ACTs "right" to do things, but simply about the
effects on the community (read: our programming environment of choice)
and what can be done now to avert negaitve effects (which probably
don't hit all of us, but only a specfic sub group).

Now, that would be the point to make some kind of appeal to the
community, like

  - If you write libraries, don't write them for ACT Gnat only.

  - If you choose a license, please choose GMGPL, even if you compile
    with Gnat GPL. This is possible and legal (AFAIK, IANAL).

  - Try to salvage the GMGPL versions from the archives and maintain
    and develop with the FSF compiler.

but I'm not at the point yet to know, wether I really want to make
that appeal. I'm not given to Don Quichote style adventures, not sure
wether the way of least resistance would not just lead elsewhere, and
wether there would be enough activists whose interests align with
mine.


> I'm all in favor of some more community based funding of generally
> useful Ada components.

Say "supported" instead of "funded" and we agree here. I don't think,
that presently you could get somebody to fund comeponents development
exactly because the situation is so desolate (perhaps ttoo strong a
word, but you get the meaning).

> But the GNAT contract seems to have been about just an Ada
> compilation system.

Exactly what I said above. As I said: ACT has probably discharged its
obligations there and we can at least agree that we can assume that it
has. The issue is not ACT Gnat but the components/library situation
(and rather specifically: The bindings to operating system service and
GUI components).


> Arguing about GPL AWS, GtkAda and so on is then a very different
> understanding of any obligations, alleged or implied, because, formally,
> the compiler/run time licensing needn't apply to general purpose software
> components.

Right. 

Looking backward: What about all the contributions that have been give
under the other licenses? Just as long as we discuss principles that
question should be answered. But I don't want to discuss principles
(you did :-)).

Looking forward: What can be done? What can be salvaged from the
situation? Partly also: What _is_ the situation, actually?

> And yes, obviously this tends to be an occasion for, well,
> complaints. I for one will be happy to join the moaning, even though

Please do so. :-) 

> in my case this has not so much to do with projects, or money.


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Simon Wright]

Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:

> 	Slipping in from the side... A non-support, run-time distribution,
> license in the $200 range would probably get a lot of takers (price
> range varying depending upon upgrade procedure -- a
> one-time/major-version fee could be higher than one that must be paid
> just to download a bug-patch release)

The GNAT compiler depends fairly intimately on the runtime, I think,
so it would be hard to produce something independent; the compiler
team could at any time change the interface, there's no independent
interface definition.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Samuel Tardieu <sam@rfc1149.net> writes:

> >>>>> "MEL" == M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:
> 
> MEL> Good question. The FSF (whose point of view actually has zero
> MEL> impact on the discussion about GPL software from ACT, since it's
> MEL> not the FSF forcing ACT in any way to use GPL ...) has, with
> MEL> regard to the GPL, probably the opinion that linking means
> MEL> "derive a work".
> 

> Most of the GNAT runtime is owned by the FSF. 

Oops yes. Oversight on my side. I should say: The FSF didn't force ACT
to drop the linking exception. But this was only an aside.

The discussion surrounding my answer had already drifted in wider
areas, which I'd like to paraphrase as "how can linking constitute a
derived work". To which I just craved to give some opinions of my own
(a - that I don't think it should be in all cases, b - from where I
think that position of the GPL originally came).

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Marc A. Criley]

Michael Bode wrote:
> I'd be one, let's say somewhere below 1000ￜ for the GPL Edition with
> GMGPL licensing. In fact I asked AdaCore for such a beast. Doesn't
> exist an probably never will.

Just trying to find some vaguely relevant post to hang this on...    :-)


A fair amount of the discussion in this thread was covered when AdaCore 
first announced GNAT GPL 2005, and there was much wailing and gnashing of 
teeth at that time.

I thought a lot about this back then, and came to a few realizations about 
developing Ada software with GNAT GPL 200x and trying to make money selling 
fully GPLed software, which were posted in some form back then, but let me 
try to summarize them here:

1) So you develop a GPLed application and sell it, either over the internet 
or in a shrinkwrapped box.  Okay, so what's the problem?  You're making 
money.  Unless it's some trivial program, or your target market is 
developers, _nobody_cares_ that the source code is available.  Put it on 
the CD or DVD--the accountant who bought your tax prep program isn't going 
to rebuild the executables.  Worse, it's in Ada, who's going to set up an 
Ada development environment to rebuild it?  :-)

2) So you develop a GPLed application and sell it to customers with whom 
you establish an ongoing support relationship.  If it's a non-technical 
product, see (1).  If it's technical and they are interested in source 
code, verbally _request_ that they not redistribute the distribution, e.g., 
AdaCore customers.  Alternatively, verbally notify them that if they 
redistribute the source you will not renew their support contract, i.e., 
Red Hat Enterprise customers. Doing this IS legal! I requested 
clarification directly from the FSF and got this response from the FSF 
Licensing Team:

"There is nothing in the license to prevent this; in fact, Red Hat does
exactly this with customers of their support services. This is not
quite the same thing as placing restrictions on the software -- after
all, the customers are still legally able to exercise all their rights
under the GPL. There is little we can do to prevent them from putting
themselves in that position."

The FSF clearly does not like this practice, but it is legal, because the 
customer can still exercise all aspects of the GPL, they just may lose the 
ability to get support from you the developer.  A business cannot be forced 
to provide or extend a support contract to someone it doesn't want to.

Hope this muddies the waters.

-- Marc A. Criley
-- McKae Technologies
-- www.mckae.com
-- DTraq - XPath In Ada - XML EZ Out

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

"Marc A. Criley" <mcNOSPAM@mckae.com> writes:

> Just trying to find some vaguely relevant post to hang this on...    :-)

I'll answer anyway :-)

I work for a company, so my boss decides over licensing, not me. I
would have no problem with GPL.

> 1) So you develop a GPLed application and sell it, either over the
> internet or in a shrinkwrapped box.  Okay, so what's the problem?
> You're making money.  Unless it's some trivial program, or your target
> market is developers, _nobody_cares_ that the source code is
> available.  Put it on the CD or DVD--the accountant who bought your
> tax prep program isn't going to rebuild the executables.  Worse, it's
> in Ada, who's going to set up an Ada development environment to
> rebuild it?  :-)

The problem here might not be the source code but the redistribution
of the binaries. Sell 1 copy, have 1 million users. Of course the same
applies to EULAed software, then it's called piracy.

> 2) So you develop a GPLed application and sell it to customers with
> whom you establish an ongoing support relationship.  If it's a
> non-technical product, see (1).  If it's technical and they are
> interested in source code, verbally _request_ that they not
> redistribute the distribution, e.g., AdaCore customers.
> Alternatively, verbally notify them that if they redistribute the
> source you will not renew their support contract, i.e., Red Hat
> Enterprise customers. 

How do you know whose support contract is it that you won't renew?

My main problem and the only thing I feel I have reason to complain
about is that it seems AdaCore might be changing licenses secretly,
deceiving users by false statements in file headers and maybe doing
this in a retroactive way (GtkAda). At least I'm not the only one who
thought GtkAda was GMGPL simply because this is explicitely written in
every single spec file of the package (except 3 files belonging to
gate).

Also I find the position that any statements about licensing in files
distributed with software or in file headers are invalid more than
disturbing. If we subscribe to this, GPL is totally fucked up. Imagine
Linus signing GPL certificate letters to Linux users for the rest of
his life.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Marc A. Criley" <mcNOSPAM@mckae.com> writes:

> Michael Bode wrote:
> > I'd be one, let's say somewhere below 1000ￜ for the GPL Edition with
> > GMGPL licensing. In fact I asked AdaCore for such a beast. Doesn't
> > exist an probably never will.
> 
> Just trying to find some vaguely relevant post to hang this on...    :-)
> 
> 
> A fair amount of the discussion in this thread was covered when
> AdaCore first announced GNAT GPL 2005, and there was much wailing and
> gnashing of teeth at that time.

Indeed. At least as far as GNAT GPL is concerned.

> I thought a lot about this back then, and came to a few realizations
> about developing Ada software with GNAT GPL 200x and trying to make
> money selling fully GPLed software, which were posted in some form
> back then, but let me try to summarize them here:
> 
> 1) So you develop a GPLed application and sell it, either over the
> internet or in a shrinkwrapped box.  Okay, so what's the problem?
> You're making money.  Unless it's some trivial program, or your target
> market is developers, _nobody_cares_ that the source code is
> available.  Put it on the CD or DVD--the accountant who bought your
> tax prep program isn't going to rebuild the executables.  Worse, it's
> in Ada, who's going to set up an Ada development environment to
> rebuild it?  :-)

Completely right. But: There is also the case when the customer wants
the developer(s) to sign an NDA. Usually (happened too me) these
customers are deeply uneasy with the GPL around. They fear (and I
don't even know wether they are right) that they cannot bind the
developers into not disclosing the code written under contract, and
they fear that contact with GPL code "frees" their own code with all
their methods and concepts. They are not sure and in a competitive
situation (meaining multiple bidders) that means they don't touch the
offer where they are not sure.

> 
> 2) So you develop a GPLed application and sell it to customers with
> whom you establish an ongoing support relationship.  If it's a
> non-technical product, see (1).  If it's technical and they are
> interested in source code, verbally _request_ that they not
> redistribute the distribution, e.g., AdaCore customers.
> Alternatively, verbally notify them that if they redistribute the
> source you will not renew their support contract, i.e., Red Hat
> Enterprise customers. Doing this IS legal! I requested clarification
> directly from the FSF and got this response from the FSF Licensing
> Team:
> 
> "There is nothing in the license to prevent this; in fact, Red Hat does
> exactly this with customers of their support services. This is not
> quite the same thing as placing restrictions on the software -- after
> all, the customers are still legally able to exercise all their rights
> under the GPL. There is little we can do to prevent them from putting
> themselves in that position."


On can also just deliver mangled code to the customer with stripped
comments and w/o documentation. One can do all that, but actually it's
bending the spirit of GPL and obeying the letter only.

Note that now I'm talking about libraries, not about GPL Gnat.

Personally I'd prefer to have LGPL or GMGPL on libraries and release
substantial portions of my reusable code into the community when and
where possible and also being able to write closed source when the
necessity arises (meaning that I can ensure the customer that I won't
disclose his modules to third parties) instead of not releasing
anything to the community and giving compilable but otherwise unusable
code to the customer.

But as Michael Bode pointed out, GPL vs LGPL/GMGPL is not the real
question here. The issue at hand is somewhat mor complicated,
specifically finding out, what certain licenses are and since when
:-).

> The FSF clearly does not like this practice, but it is legal,
> because the customer can still exercise all aspects of the GPL, they
> just may lose the ability to get support from you the developer.  A
> business cannot be forced to provide or extend a support contract to
> someone it doesn't want to.
> 
> Hope this muddies the waters.

It does.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Simon Wright]

"george" <gshapovalov@gmail.com> writes (quoting an AdaCore person):

> Right, I do not think this is worth doing. The mention inside source
> files have no real legal value anyway.

I don't suppose that has been tested in court either.  Here are all of
us non-lawyers assuming that a company such as AdaCore, led by an
expert in software copyright issues, would take the trouble to get its
publicly-visible statements related to copyright consistent.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Simon Wright]

M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:

> But the FSF, after all, has nothing todo with the GPL-releases of Gnat.

Quite a lot of the sources of the GNAT runtime say (c) FSF. Not that
you can place any reliance on that, of course.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Simon Wright <simon@pushface.org> writes:

> "george" <gshapovalov@gmail.com> writes (quoting an AdaCore person):
> 
> > Right, I do not think this is worth doing. The mention inside source
> > files have no real legal value anyway.
> 
> I don't suppose that has been tested in court either.  Here are all of
> us non-lawyers assuming that a company such as AdaCore, led by an
> expert in software copyright issues, would take the trouble to get its
> publicly-visible statements related to copyright consistent.

One would also think, that they would state the copyright _clearly_ at
the distribution site, and state the copyright in the README file. 

One is mistaken in thinking so.

A good example of stating copyrights clearly everyhwere is florist
3.15p where the linking exception is not only in the file headers, but
again in the README.

It is also usual if someone packages files with different licenses, to
give a clear explanation (usually in the README file) that single
files are under a different license ("refer to the file headers").

One would also expect that it is a big no-no to strip parts of
copyright headers where one didn't change anything (apart from
removing a default keyword) and to add a rather all-inclusive
copyright to files that weren't changed.

One is also mistaken in thinking so.

One could stipulate that all this muddying of the situational waters,
so to say, is actually done on some purpose. One could speculate on
the purpose. 

:-]

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Simon Wright <simon@pushface.org> writes:

> M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> writes:
> 
> > But the FSF, after all, has nothing todo with the GPL-releases of Gnat.
> 
> Quite a lot of the sources of the GNAT runtime say (c) FSF. Not that

See on of my other posts on that. What I wanted to say: It was not the
FSF forcing a license change from GMGPL to GPL. So no need to point to
the FSF as a reason for that. But didn't I say I won't discuss Gnat
GPL any more? So ...

> you can place any reliance on that, of course.

Nooo, nooo. Of course, not. 

But let's not forget: That was only a rumour coming from a quote
probably quoted a bit outside of its original context. I'd like to see
anyone (ACT especially) try to pull that: That the file header
wouldn't mean anything and the files are actually God-knows-whose. I
don't think that is likely to happen or even have success, so I'm not
going to get all steamed up on that one.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

M E Leypold writes:
[about distributing a commercial program under GPL]
> Completely right. But: There is also the case when the customer wants
> the developer(s) to sign an NDA. Usually (happened too me) these
> customers are deeply uneasy with the GPL around. They fear (and I
> don't even know wether they are right) that they cannot bind the
> developers into not disclosing the code written under contract, and
> they fear that contact with GPL code "frees" their own code with all
> their methods and concepts. They are not sure and in a competitive
> situation (meaining multiple bidders) that means they don't touch the
> offer where they are not sure.

Suppose you write a program and license it to a customer. You are the
author, you own the copyright, therefore you can sell licenses to other
customers as well, and you can chose the license terms for each
customer. No law forces you to use the same terms for all customers,
but law allows you to sell or give away as many licenses as you like.
Therefore, even if you sell a proprietary, binary-only license to the
first customer, you can *still* sell GPL licenses, BSD licenses or
other licenses for the same program to other customers, and you can
*still* place your program in the public domain, post it on
SourceForge, or do whatever you want with it, since you own the
copyright.

The only thing the first customer can do to prevent that is to sign a
contract with you, whereby you assign your copyright to him (which as I
understand is not legal in Germany), or you agree not to distribute
your program to anyone else.  At this point, the GPL gives the four
freedoms to the (sole) licensee, while at the same time the contract
restricts *your* freedom to redistribute. If I were a customer, I would
systematically insist on such terms when I bought custom-made software.

> On can also just deliver mangled code to the customer with stripped
> comments and w/o documentation. One can do all that, but actually it's
> bending the spirit of GPL and obeying the letter only.

I agree, but as I explained, that's what you would do if you wanted to
prevent your customers from modifying your program. But as I understand
it, that's not the problem you have now; the problem is that your
customer wants to prevent *you* from distributing your programs to
their competitors, right? Only a contract can do that.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:

> On 25 Jun 2006 22:49:48 +0200, M E Leypold
> <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> declaimed
> the following in comp.lang.ada:
> 
> > 
> > Completely right. But: There is also the case when the customer wants
> > the developer(s) to sign an NDA. Usually (happened too me) these
> > customers are deeply uneasy with the GPL around. They fear (and I

> 	Don't know how cost effective it would be -- but, from my viewpoint,

It wouldn't.

> at least, if a customer wants an NDA in the contract negotiations, the
> developer could bring in the cost of the GnatPro license (with compiler
> system with license treated as part of the deliverables, or obtained by
> the customer with developer working onsite)

Well. The calculation would be (as I have shown in another post) that
the customer would want to maintain the software for something from 3
to 10 years which with your method would equate to $45.000 to
$150.000. Obviously that only has now impact to the customers view on
cost effectiveness of the project if that amount is negligible only if
the project volume is much large than this amount of money, which
would certainly be > $1.000.000.

Let's face it: We have a restriction here: Go GPL or pay.

I'd wish that people would stop inventing ingenious methods of either
circumventing this restrictiction (distributing mangled source or
tying the customer into other dependencies that make him refraining
from distributing further) or methods of hiding the costs (let the
customer pay the compiler support).

There is a restriction. It stands to reason that it kills a number of
projects that would otherwise been economically viable (to the
satisfaction of developer and customer) and can now not be done ore
with costs that make them non viable.

That much, I think, is undisputed. What is disputed, though, is,
wether it should be like this. I cannot say anything on that topic
anymore: There are 2 different schools here. One holds that a free
tool allows to produce works with the tool that can be sold, even as
"closed works". The other holds that "free can only begat free" and
that using free tools anyone should not expect to be able to do more
than to produce free tools and probably free services. Both schools
found their respective view on different evaluations of the situation
and probably on different underlying ethical concepts as well (there
is always a strong ethical undercurrent in the arguments there). I
don't thing it will be possible to mediate between these two schools
of thought. 

Personally I think that "free can only begat free" is contradictory:
ACT (i.e.) bases their services on a free compiler, their services are
not free though. Same applies to closed source businesses using
Apache, Linux, MySQL for working tools. Nobody seriously suggest, that
only non profit organisation should be allowed the use of GPL software
from now on. All that as nothing to do with copyright, admittedly, but
it makes understandable why it is necessary to draw a line between
"producing a derived work" (copyright applies, GPL can be enforced)
and "using a tool" (that the tool incorporates parts of itself in the
product and thus makes it a derived work would be an unwanted effect
in this case, hence the birth of LGPL). All in all I don't think that
the GPL is being well served by not drawing that line clearly: Much of
the FUD around the "viral nature of GPL", like "use a GPL tool and
your source must be free" is based on exactly these fringe cases where
either the authors of some software didn't care enough to ensure
clarify the case or did purposely use the GPL to lock their software
away from what they considered unfair profiting.

But back to the main theme:

I stipulate that the projects killed by the restriction are actually
those whose killing harm the case of proliferating Ada most: They take
place in the area, where people shift from purely hobbyist or
educational use to the first commercial applications of Ada and start
to build their infrastructure for future projects.

Let me also add: Having tought software engineering courses at german
universities I wouldn't try to introduce Ada as a teaching language
there today: Whereas the language has all it should have, the skill
set acquired by the students would be only usable in very restricted
scenarios and this in a small market anyway. It simply wouldn't be
responsible to restrict the student's future to the following choices:

  - Work for DASA or ESA or for the defense industry.

  - Try to build your own pure Ada shob (w/o previous experience: hard)
    with whopping fixed costs for the Ada part of your shop.
 
  - Restrict yourself to producing GPLed software only in your shops
    "primary system language".


> 	The GPL curse applies to small independent developers seeking
> to produce generally marketable products -- the next Quicken, or
> VersaCheck, for example... Given the prohibitive cost of obtaining a
> closed-distribution capability for GNAT, they might easily chose to
> substitute some other language -- M$ C++ compiler without the
> VisualStudio environment appears to be available for the download;
> Sun's Java seems to be behind a number of applications...

That sums it up nicely. You might substitute "they might easily chose"
by "they chose", because that is what the are regularly doing. How do
you think that programming languages are chosen? I suggest the
following considerations:

 - Above all: How expensive is an environment with "closed
   distribution capability" (CDC)?

 - Can I use the language as single system language, i.e. even if my
   current application isn't graphical in nature, would I be able to
   program graphical applications? Would I be able to iterface to the
   Operating system?

 - Which libraries (with CDC) are available?

 - Can I find developers / programmers / collaborators in that
   language?

 - How many compilers / build systems are there in the market? Would I
   be able to fall back to another vendors build system in some years?
   Would I be able to fall back to community supported software for at
   least the last years of low intensity maintenance?

 - And what are the language advantages / disadvantages?

The last point creates a kind of vicious circle actually: No CDC ->
Skill set not attractive for students -> no Programmers -> no use of
the language in small shops --> skill set not attractive for students
-> ...

You get the picture. 

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> The only thing the first customer can do to prevent that is to sign a
> contract with you, whereby you assign your copyright to him (which as I
> understand is not legal in Germany), or you agree not to distribute
> your program to anyone else.  At this point, the GPL gives the four
> freedoms to the (sole) licensee, while at the same time the contract
> restricts *your* freedom to redistribute. If I were a customer, I would
> systematically insist on such terms when I bought custom-made software.

Exactly. Two complications arise:

 - I don't want to hand over exclusively my container and my widget
   library :-).

 - The customer, as I have written, is not sure, wether such a
   contract with me is binding. After all: I have the source, the libs
   are under GPL -- doesn't that entitle me to distribute under GPL (a
   right I cannot waive w/o the GPL lapsing for the libs).  So the
   customer is not sure. "Not sure" in business translates as
   "risk". Instead of jumping through the hoops to consult a lawyer,
   who might or might not give a definite answer to that, they decide
   to go with another offer that looks less fraught with ifs.

> 
> > On can also just deliver mangled code to the customer with stripped
> > comments and w/o documentation. One can do all that, but actually it's
> > bending the spirit of GPL and obeying the letter only.
> 
> I agree, but as I explained, that's what you would do if you wanted to
> prevent your customers from modifying your program. But as I understand
> it, that's not the problem you have now; the problem is that your
> customer wants to prevent *you* from distributing your programs to
> their competitors, right? Only a contract can do that.

Can it? And can the customer be convinced that it can? That is where
lawyerly territory starts ("Quick, the spell of protection, say it
now: 'IANAL, IANAL, IANAL'") and customers feel unwell in that
territory as well. Especially for small projects, so they tend to
avoid complications of that kind as a general rule.

Back to libraries like GtkAda and the posibbly future GPL Florist
which are my real issue in all that discussion: I'd prefer a simple
and clear cut situation (like a well documented GMGPL) to a number of
confusing fringe case and legal workarounds.

It doesn't equate to freedom in my book to have to refer to a lawyer
all the time. Quite the contrary.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

M E Leypold writes :
> Ludovic Brenta writes:
> > The only thing the first customer can do to prevent that is to sign a
> > contract with you, whereby you assign your copyright to him (which as I
> > understand is not legal in Germany), or you agree not to distribute
> > your program to anyone else.  At this point, the GPL gives the four
> > freedoms to the (sole) licensee, while at the same time the contract
> > restricts *your* freedom to redistribute. If I were a customer, I would
> > systematically insist on such terms when I bought custom-made software.
>
> Exactly. Two complications arise:
>
>  - I don't want to hand over exclusively my container and my widget
>    library :-).

You can stipulate that in the contract; you can say anything you want
in a contract. For example:

ARTICLE 1

The developer licenses the Program P and the Libraries L1, L2, of which
he is the author, to the customer, under the terms of the GPL.

ARTICLE 2

The developer agrees not to distribute the Program P to any third party
without written consent of the customer, but remains free to distribute
the Libraries to other parties.

ARTICLE 3

The customer agrees to pay the sum of X to the developer.

Signed, etc.

>  - The customer, as I have written, is not sure, wether such a
>    contract with me is binding. After all: I have the source, the libs
>    are under GPL -- doesn't that entitle me to distribute under GPL (a
>    right I cannot waive w/o the GPL lapsing for the libs).  So the
>    customer is not sure. "Not sure" in business translates as
>    "risk". Instead of jumping through the hoops to consult a lawyer,
>    who might or might not give a definite answer to that, they decide
>    to go with another offer that looks less fraught with ifs.

IANAL either, but I understand that a signed contract overrides any
unsigned license, provided that:
- the contract is legal (does not force any party to do illegal things)
- there is a compensation clause (i.e. you receive money or other form
of retribution; a contract has to have obligations for both parties)

If you sign a contract whereby you agree not to distribute your program
to third parties, then your customer has full assurance, and can sue
you later for breach of contract. To reinforce the assurances, you can
even stipulate which court is competent should the need arise.

This remains true even if you license your program under a non-free,
closed-source, proprietary EULA. So, if your customer is really
concerned, they should have signed a contract with you already, even
before discussing the license terms.

> It doesn't equate to freedom in my book to have to refer to a lawyer
> all the time. Quite the contrary.

You need a lawyer to assist you in legal matters, just like you need a
physician to assist you in health matters and a butcher to assist you
in meat procurement matters, or like your customer needs a software
engineer to assist him in computer-related matters. This is orthogonal
to freedom.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> M E Leypold writes :
> > Ludovic Brenta writes:
> > > The only thing the first customer can do to prevent that is to sign a
> > > contract with you, whereby you assign your copyright to him (which as I
> > > understand is not legal in Germany), or you agree not to distribute
> > > your program to anyone else.  At this point, the GPL gives the four
> > > freedoms to the (sole) licensee, while at the same time the contract
> > > restricts *your* freedom to redistribute. If I were a customer, I would
> > > systematically insist on such terms when I bought custom-made software.
> >
> > Exactly. Two complications arise:
> >
> >  - I don't want to hand over exclusively my container and my widget
> >    library :-).
> 
> You can stipulate that in the contract; you can say anything you want
> in a contract. For example:


Of course. Only wanted to hint that things are more complicated than
in your scenario.


> ARTICLE 3
> 
> The customer agrees to pay the sum of X to the developer.

This obviously is the most important part :-).



> 
> Signed, etc.
> 
> >  - The customer, as I have written, is not sure, wether such a
> >    contract with me is binding. After all: I have the source, the libs
> >    are under GPL -- doesn't that entitle me to distribute under GPL (a
> >    right I cannot waive w/o the GPL lapsing for the libs).  So the
> >    customer is not sure. "Not sure" in business translates as
> >    "risk". Instead of jumping through the hoops to consult a lawyer,
> >    who might or might not give a definite answer to that, they decide
> >    to go with another offer that looks less fraught with ifs.
> 
> IANAL either, but I understand that a signed contract overrides any
> unsigned license, provided that:

AFAIK there is no concept like "license" in, i.e. German
law. everything is based on contract law (implied contracts where
necessary) or rights that are granted one sided by a party to the
general public by some declaration (like "hereby I gift XYZ to the
public domain).

I'm not sure I got that right, perhaps it's also splitting
hairs. IANAL, you know.

> - the contract is legal (does not force any party to do illegal things)
> - there is a compensation clause (i.e. you receive money or other form
> of retribution; a contract has to have obligations for both parties)
> 
> If you sign a contract whereby you agree not to distribute your program
> to third parties, then your customer has full assurance, and can sue
> you later for breach of contract. To reinforce the assurances, you can
> even stipulate which court is competent should the need arise.


> This remains true even if you license your program under a non-free,
> closed-source, proprietary EULA. So, if your customer is really
> concerned, they should have signed a contract with you already, even
> before discussing the license terms.

But doesn't the GPL (on the libs) lapse then if/when the customer
tries to restrict my distribution rights :-)?

> > It doesn't equate to freedom in my book to have to refer to a lawyer
> > all the time. Quite the contrary.
> 
> You need a lawyer to assist you in legal matters, just like you need a

The keyword is "all the time". 

> physician to assist you in health matters and a butcher to assist you

If I have to refer to a physician all the time, I'm very ill. If I
have to refere to a lawyer all the time (not only for checking/writing
the written contract but also for simple forward planning or even for
checking wether COPYING is binding, ahem, ahem), something is amiss in
the freedom department.

> in meat procurement matters, or like your customer needs a software
> engineer to assist him in computer-related matters. This is orthogonal
> to freedom.


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Martin Krischik]

Simon Wright wrote:

> M E Leypold <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de>
> writes:
> 
>> But the FSF, after all, has nothing todo with the GPL-releases of Gnat.
> 
> Quite a lot of the sources of the GNAT runtime say (c) FSF. Not that
> you can place any reliance on that, of course.

The GNAT sources themself are not the problem as gcc 4.1.1 is pretty good as
well. It's the other libraries which AdaCore supplies.

And a little reminder: They are all separate products at AdaCore. If your
pay up your $/ￜ 15.000 then you get one year/five programmer support for
the compiler only. No ASIS, No Florist, No PolyORB, no XML/Ada - they all
cost extra.

Martin
-- 
mailto://krischik@users.sourceforge.net
Ada programming at: http://ada.krischik.com

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Marc A. Criley]

M E Leypold wrote:

>  - Above all: How expensive is an environment with "closed
>    distribution capability" (CDC)?

Actually, once we in comp.lang.ada get off our fixation with GNAT, the 
answer to the above is...not necessarily that much, especially for a 
hobbyist trying to convert their hobby and advocacy into a small business 
(so long as Windows is the target).

The problem is that because AdaCore distributed a free Ada compiler that 
could be used to build close-sourced applications for awhile, and now their 
only free, maintained, distribution is for a GPLed compiler, we're all 
stuck on this notion that we should have access to a free, maintained 
compiler with which we can build closed-source applications.

Move on.

If you can pay the $15K, or work with the GPL, then fine, otherwise look 
elsewhere.

I'm sure Randy Brukardt must have just given up in frustration after years 
of pointing out that Janus Ada (www.rrsoftware.com) has long been available 
and is priced competitively with other commercial programming environments.

According to the website: the Professional Edition, which appear to include 
CLAW, is $495 ($300/yr maintenance), while the Personal Edition (no CLAW 
and some other tools) is $195 ($95/yr maintenance), though you can get a 
Personal Edition combo for $295 ($200/yr maintenance).

Or if you want to go the Aonix route (www.aonix.com), the quote I got last 
year from them was:

Windows: $1495 + $1495/yr maintenance
Linux:   $5000 + $1000/yr maintenance

So if you're looking to go the small business route, look elsewhere than 
AdaCore, and maybe help out another small business (RR) in the process.

-- Marc A. Criley
-- McKae Technologies
-- www.mckae.com
-- DTraq - XPath In Ada - XML EZ Out

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

Dennis Lee Bieber wrote:
> 
> 	The GPL curse applies to small independent developers seeking to
> produce generally marketable products -- the next Quicken, or
> VersaCheck, for example... Given the prohibitive cost of obtaining a
> closed-distribution capability for GNAT, they might easily chose to
> substitute some other language -- M$ C++ compiler without the
> VisualStudio environment appears to be available for the download; Sun's
> Java seems to be behind a number of applications...

And GNAT is hardly the only low-cost compiler for those targets.

-- 
Jeff Carter
"People called Romanes, they go the house?"
Monty Python's Life of Brian
79

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> 
> Let's face it: We have a restriction here: Go GPL or pay.

Hardly. You can use one of the fine GMGPL compilers available free 
(3.4.X, 4.1), so need neither go GPL nor pay. You can also use a 
low-cost compiler from another vendor (RR SW, AONIX), which does not 
involve going GPL and does not require nearly the investment of GNAT Pro.

-- 
Jeff Carter
"People called Romanes, they go the house?"
Monty Python's Life of Brian
79

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Marc A. Criley" <mcNOSPAM@mckae.com> writes:

> M E Leypold wrote:
> 
> >  - Above all: How expensive is an environment with "closed
> >    distribution capability" (CDC)?
> 
> Actually, once we in comp.lang.ada get off our fixation with GNAT, the 

Yes. :-)

> answer to the above is...not necessarily that much, especially for a 
> hobbyist trying to convert their hobby and advocacy into a small business 

> (so long as Windows is the target).

Well ... this is too restrictive a future to me.

> The problem is that because AdaCore distributed a free Ada compiler that 
> could be used to build close-sourced applications for awhile, and now their 

Not quite. I think the history of Gnat and the obligations involved is
a bit more complex than "because AdaCore distributed a free Ada
compiler".

> only free, maintained, distribution is for a GPLed compiler, we're all 
> stuck on this notion that we should have access to a free, maintained 
> compiler with which we can build closed-source applications.
> 
> Move on.

Right. That probably means to another language, that has all this, and
where I/we can pay our due with contributions to an evinronment, where
I don't have to watch my legal back all the time. Freedom is also
about this: That I'm reasonably sure that things stay as they are and
library licenses (GtkAda, i.e.) don't change over night and without
notice.

> If you can pay the $15K, or work with the GPL, then fine, otherwise look 
> elsewhere.

I hope you're not trying to shoe me away. :-). You know I've not been
arguing about Gnat now for a while, but about the unknown licence of
GtkAda and the danger of Florist going completely GPL if the community
relies on the maintencance supllied by AdaCore.

A compiler, that doesn't have operating system bindings (or which are
more restrictively licensed than the rest of the runtime) is worth
nothing in my opinion and they could just as well remove Ada from Gcc
tree again.


> I'm sure Randy Brukardt must have just given up in frustration after years 
> of pointing out that Janus Ada (www.rrsoftware.com) has long been available 
> and is priced competitively with other commercial programming environments.

I know his compilers. I appreciate his contributions to c.l.a. But the
compiler doesn't seem available for Linux or BSD. 

> According to the website: the Professional Edition, which appear to include 
> CLAW, is $495 ($300/yr maintenance), while the Personal Edition (no CLAW 
> and some other tools) is $195 ($95/yr maintenance), though you can get a 
> Personal Edition combo for $295 ($200/yr maintenance).

I know all that. Believe me, I researched the Ada compiler market over
more than a year ago very thoroughly.

Generally I'm not asking for advice here. Every discussion here seems
to gravitate pretty fast to sentences like "If you cannot afford, you
should or could...".

As I've been reiterating ad nauseam, I'm not intrested in AdaCore Gnat
anymore. But doesn't "the community" have any issues with the GtkAda
license changes? If so, there are remedies to that.


> Or if you want to go the Aonix route (www.aonix.com), the quote I got last 
> year from them was:
> 
> Windows: $1495 + $1495/yr maintenance
> Linux:   $5000 + $1000/yr maintenance

I got the same quote over a year ago.

> So if you're looking to go the small business route, look elsewhere than 
> AdaCore, and maybe help out another small business (RR) in the process.

You know, I understand that very well.

I'd have, but the RR product line looks a bit on the way out of the
market. Also, there are no bindings to a portable graphical tools kit
available.

What you're actually spelling out is the message: Ada is dead as a
universal language. It's a language for big players and specialists
only now. I've actually been prepared to hear that for some time now,
but it makes me sad anyway and I've been hoping for a community (of
non hobbyists) that is interested to keep a crossplatform enviroment
for Ada alive. I've always had a fondness for this kind of languages
(from Pascal, Modula, Ada, Oberon), but it probably makes of course no
sense "in business" to decide on a this kind of gut feeling.

I don't want rely on Language where there is only 1 affordable vendor
(even if the bindings I need where available). And I'd like to have
the option to go back on community supported tools

All that is available with C/C++/Ocaml/Java/whatever. So, if for a
moment we go away from "my case" (which I never thought exemplary
anyway), are there actually any open questions why Ada is not the
language of choice for all the small startups? 

This question, why there is so much (supposedly bug ridden and
unreliable software) is produced in other languages which do not have
the rigor of Ada, has turned up repeatedly here.  Here might be some
answer.

I've to admit, that I'm a bit tired of the discussion, since I'm
repeating myself. Just to recap my main points:

 - I'm not asking for advice. I know well what I'm doing :-). I'm
   asking for information.

   As I personal statement: Relying on commercially available compiler
   is in my eyes not an option (YMMV). Generally I'd prefer to change
   to another language as my primary system language. Which wouldn't
   preclude me from doing consulting and contract work in Ada, so
   what :-).

 - Some licensing questions are still open. Apart from the
   consequences certain license have or would have the first step
   should be to find out, what those licenses actually are.

 - I'm not interested in discussing what AdaCore or whoever has done
   or should have done. That's their business, but I'd certainly like
   to know what they are actually doing licensewise: This is in the
   dark presently as far as GtkAda and some other sources are
   concerned.

 - I wonder, how many people are here, that would be interested in
   trying to collect and maintain a community supported Ada
   environment.



Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> Dennis Lee Bieber wrote:
> > 
> > 	The GPL curse applies to small independent developers seeking to
> > produce generally marketable products -- the next Quicken, or
> > VersaCheck, for example... Given the prohibitive cost of obtaining a
> > closed-distribution capability for GNAT, they might easily chose to
> > substitute some other language -- M$ C++ compiler without the
> > VisualStudio environment appears to be available for the download; Sun's
> > Java seems to be behind a number of applications...
> 
> And GNAT is hardly the only low-cost compiler for those targets.


Exactly. There are other languages. All complainers now go away and
will leave the Ada community (which has as such no issues with highly
prices compilers) alone.

That was irony. 

But please don't do any Ada advocacy from now on. There are reasons
why Ada has been falling into disuse slowly (excapt for defense and
avionics) despit its technical merits.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.or> writes:

> M E Leypold wrote:
> > 
> > Let's face it: We have a restriction here: Go GPL or pay.
> 
> Hardly. You can use one of the fine GMGPL compilers available free 
> (3.4.X, 4.1), so need neither go GPL nor pay. You can also use a 
> low-cost compiler from another vendor (RR SW, AONIX), which does not 
> involve going GPL and does not require nearly the investment of GNAT Pro.

Please follow my other posts :-). Regarding what I'd call "the
environment", like bindings to a portable GUI toolkit or the OS.

I know I've been posting a lot these days, and it's difficult to read
all that. But I also do not want to rehash everything for the
umpteenth time.

As far as a general group of answers here go: Answers like "you can
use another language" are hardly adequate to problem. But I'll try a
simpler approach now:

 - I love Ada.
 - I'd like to see a commercially viable (GMGPL or LGPL) environment.
 - Anybody else?

A number of people here seem to interpret requests for informations
here as cries for help (sort of "my business is dying, get me a free
whatever"). This is all muddled with often not very well thought out
philosophy what free means or should mean. Perhaps I _am_ a victim of
a language barrier here. Never mind. But as far as I'm concerned my
requests only want to

  (a) Clarify the state of some licenses and their supposed interaction
      with other licenses.

  (b) Find out wether there is still a commercially viable avenue in
      developing portable consumer software with Gnat (FSF) and/or 


What "portable", "consumer" and "viable" mean in the case I'm trying
to make, is basically my definition (and my right to define here). So
if one hasn't tried to do exactly that in 2006, most advice is bound
to fall short of the real issues here. I.e availability of a compiler,
alone, is the least problem one faces (on the other side it already
spells desaster if the choice of compiler limits the number of
available libraries / bindings).

For me, it's not an emergency. I've a number of other options that
have nothing to do with Ada.


Regards -- Markus

Teaching languages (was: can one build commercial applications with latest gnat and other licenses related questions...)

[Björn Persson]

M E Leypold wrote:
> Let me also add: Having tought software engineering courses at german
> universities I wouldn't try to introduce Ada as a teaching language
> there today: Whereas the language has all it should have, the skill
> set acquired by the students would be only usable in very restricted
> scenarios and this in a small market anyway.

If they were software engineering courses I really hope you taught your 
students something more than a programming language. I was taught 
Scheme, Pascal and Java in university courses. At work I currently use 
C++, and a little Python, Perl and Bash. Yet I do use what I learned in 
those courses: structured programming, information hiding, algorithms, 
data structures and all that. Universities do teach programming 
languages, because you can't teach programming without a language, but 
the language isn't the skill set that the students are supposed to 
acquire. Pascal was designed specifically for teaching, and I doubt my 
teachers expected that I'd ever write commercial applications in Scheme. 
Give your students a good grasp of the paradigms and concepts of 
computer science, and introduce them to a few different languages (the 
more different the better), and they'll be able to apply their knowledge 
to new languages as needed.

-- 
Bj�rn Persson                              PGP key A88682FD
                    omb jor ers @sv ge.
                    r o.b n.p son eri nu

Re: Teaching languages (was: can one build commercial applications with latest gnat and other licenses related questions...)

[M E Leypold]

Bj�rn Persson <spam-away@nowhere.nil> writes:

> M E Leypold wrote:
> > Let me also add: Having tought software engineering courses at german
> > universities I wouldn't try to introduce Ada as a teaching language
> > there today: Whereas the language has all it should have, the skill
> > set acquired by the students would be only usable in very restricted
> > scenarios and this in a small market anyway.
> 
> If they were software engineering courses I really hope you taught
> your students something more than a programming language. I was taught

I hope so. We explicitely focused on SE (what was possible in the
restricted time) and tried to work on the language independent
concepts. And at least on an understanding of the problems involved in
constructiong software systematically.

> Scheme, Pascal and Java in university courses. At work I currently use
> C++, and a little Python, Perl and Bash. Yet I do use what I learned
> in those courses: structured programming, information hiding,
> algorithms, data structures and all that. 

> Universities do teach programming languages, because you can't teach
> programming without a language, but the language isn't the skill set
> that the students are supposed to acquire.

I agree. My comment was more in the direction of: A language is a
skill set (few companies today are willing to hire anyone for
programming work who has not at least some experience in the language
they are going to use: Industry these days doesn't believe in the
portability of knowledge -- more in trained monkey reflexes). It would
be a divident if the student can actually use that skill set in
practice. 

> Pascal was designed specifically for teaching,
> and I doubt my teachers expected that I'd ever write commercial
> applications in Scheme. 

> Give your students a good grasp of the

For me that's past tense for some time. :-).

> paradigms and concepts of computer science, and introduce them to a
> few different languages (the more different the better), and they'll

Exactly my saying. Unfortunately students of CS in Germany are rather
unwilling to learn more than one language and the universities don't seem
to see this as necessary.

> be able to apply their knowledge to new languages as needed.

Regards  -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ole-Hjalmar Kristensen]

Completely agree. The current licensing model of GNAT does very little
to promote the growth of Ada, but brings a lot of FUD. Once again we
are shooting ourselves in the foot. 

>>>>> "DLB" == Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:

    DLB> On 25 Jun 2006 22:49:48 +0200, M E Leypold
    DLB> <development-2006-8ecbb5cc8aREMOVETHIS@ANDTHATm-e-leypold.de> declaimed
    DLB> the following in comp.lang.ada:

    >> 
    >> Completely right. But: There is also the case when the customer wants
    >> the developer(s) to sign an NDA. Usually (happened too me) these
    >> customers are deeply uneasy with the GPL around. They fear (and I

    DLB> 	Don't know how cost effective it would be -- but, from my viewpoint,
    DLB> at least, if a customer wants an NDA in the contract negotiations, the
    DLB> developer could bring in the cost of the GnatPro license (with compiler
    DLB> system with license treated as part of the deliverables, or obtained by
    DLB> the customer with developer working onsite)

    DLB> 	The GPL curse applies to small independent developers seeking to
    DLB> produce generally marketable products -- the next Quicken, or
    DLB> VersaCheck, for example... Given the prohibitive cost of obtaining a
    DLB> closed-distribution capability for GNAT, they might easily chose to
    DLB> substitute some other language -- M$ C++ compiler without the
    DLB> VisualStudio environment appears to be available for the download; Sun's
    DLB> Java seems to be behind a number of applications...
    DLB> -- 
    DLB> 	Wulfraed	Dennis Lee Bieber		KD6MOG
    DLB> 	wlfraed@ix.netcom.com		wulfraed@bestiaria.com
    DLB> 		HTTP://wlfraed.home.netcom.com/
    DLB> 	(Bestiaria Support Staff:		web-asst@bestiaria.com)
    DLB> 		HTTP://www.bestiaria.com/

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Dmitry A. Kazakov]

On 27 Jun 2006 09:26:26 +0200, Ole-Hjalmar Kristensen wrote:

> Completely agree. The current licensing model of GNAT does very little
> to promote the growth of Ada, but brings a lot of FUD. Once again we
> are shooting ourselves in the foot. 

Are you an ACT employee? (:-)) Otherwise, agree, probably they do
themselves...

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ole-Hjalmar Kristensen]

No. Sorry about the 'we', I was thinking about the Ada community as a
whole. My employer is rather actively pushing Java :-)


>>>>> "DAK" == Dmitry A Kazakov <mailbox@dmitry-kazakov.de> writes:

    DAK> On 27 Jun 2006 09:26:26 +0200, Ole-Hjalmar Kristensen wrote:
    >> Completely agree. The current licensing model of GNAT does very little
    >> to promote the growth of Ada, but brings a lot of FUD. Once again we
    >> are shooting ourselves in the foot. 

    DAK> Are you an ACT employee? (:-)) Otherwise, agree, probably they do
    DAK> themselves...

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

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Pascal Obry]

Markus,

I've some hard time to understand you. You have a business, you ask for
an Ada compiler for free (GMGPL or GPL). This is asking for using
somebody else toolset to develop closed-source softwares without paying!

So why not use the GPL for your softwares? Why are you asking some
company to give-for-free years of work when you are not considering to
do the same? Please this is nothing personal, I'm just trying to understand.

I know that lot of companies are using GCC, Emacs, OpenOffice, Firebird,
Thunderbird... daily and they are not even considering supporting the
open source movement. I know some companies are using the Open Source to
obtain better price for concurrent applications.

One way to contribute to the Open Source movement is to develop GPL
softwares. Another way is to support the Open Source by paying support
or licenses. But isn't asking for a "commercially viable" free (as free
beer) LGPL compiler too much?

Right this is possible for some projects like GCC, but I don't see this
as fair.

Pascal.

-- 

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Pascal Obry <pascal@obry.net> writes:

> Right this is possible for some projects like GCC, but I don't see this
> as fair.

Is it fair to change the past?

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Pascal Obry]

Michael Bode a �crit :

> Is it fair to change the past?

You can never change the past. So I do not understand what you are saying.

Pascal.

-- 

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> "Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> 
>> And GNAT is hardly the only low-cost compiler for those targets.
> 
> Exactly. There are other languages. All complainers now go away and
> will leave the Ada community (which has as such no issues with highly
> prices compilers) alone.

This being c.l.a, "compiler" means an Ada compiler. If I intend a 
compiler for another language, I state that explicitly.

-- 
Jeff Carter
"Many times we're given rhymes that are quite unsingable."
Monty Python and the Holy Grail
57

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Pascal Obry <pascal@obry.net> writes:

> You can never change the past. So I do not understand what you are saying.

I'd bet GtkAda 2.4 once was GMGPL. Ludovic Brenta thought that too. M
E Leypold had some quotes to that effect in
<ctirmmu0fc.fsf@hod.lan.m-e-leypold.de>.

And now we learn that GtkAda 2.4 never was under GMGPL but GPL. That's
what I call changing the past.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> 
> Please follow my other posts :-). Regarding what I'd call "the
> environment", like bindings to a portable GUI toolkit or the OS.

OK. I have 2 GMGPL versions of GtkAda (2.2.0 and 2.4.0) and a GMGPL 
version of GNAT (MinGW 3.4.2) on my computer at the moment. (These are 
not the only compilers or bindings, just the ones that seem relevant to 
this discussion.) GtkAda is a portable GUI toolkit and GNAT comes with 
GMGPL bindings to the OS. All were free in terms of money. So I don't 
see what your complaint is.

It's also clear that ACT cannot legally impose the GPL on their 
compilers the way they have. The DOD contract that resulted in GNAT was 
to create an Ada-95 compiler that is free (in both senses of the word) 
and may be used to create un-free programs. (These requirements resulted 
in the GMGPL.) As there is no Ada-0X standard yet, there can be no 
Ada-0X compilers yet, and in fact the ACT GPL compilers are Ada-95 
compilers that implement some of the features of the Ada-0X draft only 
as an option. It seems ACT is calling them "200[5|6]" to try to get 
around the requirements of the contract, but it's clear that these are 
Ada-95 compilers with some extra features optionally available, and so 
in violation of the contract. (I wonder who at the DOD one would contact 
about this, since the AJPO no longer exists.)

Of course, IANAL.

-- 
Jeff Carter
"Many times we're given rhymes that are quite unsingable."
Monty Python and the Holy Grail
57

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Michael Bode <m.g.bode@web.de> writes:

> I'd bet GtkAda 2.4 once was GMGPL. Ludovic Brenta thought that too. M
> E Leypold had some quotes to that effect in
> <ctirmmu0fc.fsf@hod.lan.m-e-leypold.de>.

The wayback machine is a great tool:

http://web.archive.org/web/20050203211706/http://libre.act-europe.fr/GtkAda/

-- 
Michael Bode

community supported Ada environment (was: can one build commercial applications with latest gnat and other licenses related questions...)

[Björn Persson]

M E Leypold wrote:
>  - I wonder, how many people are here, that would be interested in
>    trying to collect and maintain a community supported Ada
>    environment.

Here's one. I could only do a small part of the work and I have little 
hope that enough manpower can be found, but if such a project could be 
started then I'd want to join it.

-- 
Bj�rn Persson                              PGP key A88682FD
                    omb jor ers @sv ge.
                    r o.b n.p son eri nu

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

Michael Bode <m.g.bode@web.de> writes:

> Pascal Obry <pascal@obry.net> writes:
>
>> You can never change the past. So I do not understand what you are saying.
>
> I'd bet GtkAda 2.4 once was GMGPL. Ludovic Brenta thought that too. M
> E Leypold had some quotes to that effect in
> <ctirmmu0fc.fsf@hod.lan.m-e-leypold.de>.
>
> And now we learn that GtkAda 2.4 never was under GMGPL but GPL. That's
> what I call changing the past.

No, we do not learn that.  AdaCore was quite specific in saying
they're not saying anything at all about the past.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> It's also clear that ACT cannot legally impose the GPL on their
> compilers the way they have. The DOD contract that resulted in GNAT
> was to create an Ada-95 compiler that is free (in both senses of the
> word) and may be used to create un-free programs. (These requirements
> resulted in the GMGPL.) As there is no Ada-0X standard yet, there can
> be no Ada-0X compilers yet, and in fact the ACT GPL compilers are
> Ada-95 compilers that implement some of the features of the Ada-0X
> draft only as an option. It seems ACT is calling them "200[5|6]" to
> try to get around the requirements of the contract, but it's clear
> that these are Ada-95 compilers with some extra features optionally
> available, and so in violation of the contract. (I wonder who at the
> DOD one would contact about this, since the AJPO no longer exists.)
>
> Of course, IANAL.

What if the contract had a fixed term, and has now expired?
Conjectures, hypotheses and speculation.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Pascal Obry <pascal@obry.net> writes:

> Markus,

Pascal,

I answered that in private mail (since I didn't see immediately that
you also posted to c.l.a) with an additional question about the past
and present license status of AWS.

Quntessence of my answer was, that I'm repeating myself and tried
again to clarify my point which has nothing to do with expecting
"others" to get me something "for free". OS doesn't work like this
IMHO. I'd like to refer anyone who is reading this to all my othe
postings for more details. As I said: I'm repeating myself.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Pascal Obry <pascal@obry.net> writes:

> Michael Bode a �crit :
> 
> > Is it fair to change the past?
> 
> You can never change the past. So I do not understand what you are saying.

You're not up to date Pascal :-).

Regards -- Markus
 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Michael Bode <m.g.bode@web.de> writes:

> Pascal Obry <pascal@obry.net> writes:
> 
> > You can never change the past. So I do not understand what you are saying.
> 
> I'd bet GtkAda 2.4 once was GMGPL. Ludovic Brenta thought that too. M
> E Leypold had some quotes to that effect in
> <ctirmmu0fc.fsf@hod.lan.m-e-leypold.de>.
> 
> And now we learn that GtkAda 2.4 never was under GMGPL but GPL. That's
> what I call changing the past.

That's what I call Orwellian. :-)

Regards -- Markus (unhappy)

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

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

> Michael Bode <m.g.bode@web.de> writes:
> 
> > Pascal Obry <pascal@obry.net> writes:
> >
> >> You can never change the past. So I do not understand what you are saying.
> >
> > I'd bet GtkAda 2.4 once was GMGPL. Ludovic Brenta thought that too. M
> > E Leypold had some quotes to that effect in
> > <ctirmmu0fc.fsf@hod.lan.m-e-leypold.de>.
> >
> > And now we learn that GtkAda 2.4 never was under GMGPL but GPL. That's
> > what I call changing the past.
> 
> No, we do not learn that.  AdaCore was quite specific in saying
> they're not saying anything at all about the past.

Let me paraphrase:

 - "We are distributing GtkAda 2.4.0 as GPL!"

 - "What? GtkAda 2.4.0 in 2005/02? GMGPL? We don't know anything about
   that!"

 - "What? The versions 2.2.0, 1.3.x etc? From our site? The are GPL!"
 
 - "What? They were distributed as GMGPL? We don't know anything about
   that?"

Selective loss of memory or the will to remember does come very close
here to a denial that the past ever existed. Indeed it is the
strongest mode to deny the past without outright saying so (and get
caught, which would be inevitable here).

I wonder wether they remember how and when they acquired any right to
the sources or wether the selective amnesia also covers that areas.

Ludovico: If GtkAda 2.4.0 has been acquired before 2005/02, presumably
from libre2.act-europe.fr, and the past has not changed, why don't you
let the license as GMGPL?


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> M E Leypold wrote:
> > "Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> >
> >> And GNAT is hardly the only low-cost compiler for those targets.
> > Exactly. There are other languages. All complainers now go away and
> > will leave the Ada community (which has as such no issues with highly
> > prices compilers) alone.
> 
> This being c.l.a, "compiler" means an Ada compiler. If I intend a
> compiler for another language, I state that explicitly.

Oops, sorry. This was really an unintended mixup with another
thread. Context overspill :-).

OK. Let me rephrase: There is exactly 1 low cost compiler, that from
RR sofware. AFAIS there is no Linux target (which in my context of the
discussion was always a given necessity). But one compiler hardly
justifies plural.

Or did I overlook other offers? I'm seriously intrested in the
information apart from the discussion in this thread which is already
a dead horse in my eyes. Let's stop flogging it. The compiler is not
an issue, there is FSF Gnat.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> M E Leypold wrote:
> > Please follow my other posts :-). Regarding what I'd call "the
> > environment", like bindings to a portable GUI toolkit or the OS.
> 
> OK. I have 2 GMGPL versions of GtkAda (2.2.0 and 2.4.0) and a GMGPL
> version of GNAT (MinGW 3.4.2) on my computer at the moment. (These are
> not the only compilers or bindings, just the ones that seem relevant
> to this discussion.) GtkAda is a portable GUI toolkit and GNAT comes
> with GMGPL bindings to the OS. All were free in terms of money. So I
> don't see what your complaint is.

That ACT denies they can remember that ever GtkAda was distributed a
GMGPL. That is about all and at the core of all that. You understand
that now? Even the older versions from the ACT site are now supposed
to be under GPL ONLY, even if they didn't change from the way they
very (as we remember) distributed as GMGPL.

> It's also clear that ACT cannot legally impose the GPL on their
> compilers the way they have. The DOD contract that resulted in GNAT
> was to create an Ada-95 compiler that is free (in both senses of the

Make me laugh. I got flamed repeatedly because I only suggested that
the GMGPL -> GPL transition might not have been quite such a clear cut
right of ACT to do and because I wanted to create "unfree programs"
with "AdaCore's tools".

> word) and may be used to create un-free programs. (These requirements
> resulted in the GMGPL.) As there is no Ada-0X standard yet, there can
> be no Ada-0X compilers yet, and in fact the ACT GPL compilers are
> Ada-95 compilers that implement some of the features of the Ada-0X
> draft only as an option. It seems ACT is calling them "200[5|6]" to
> try to get around the requirements of the contract, but it's clear
> that these are Ada-95 compilers with some extra features optionally
> available, and so in violation of the contract. (I wonder who at the
> DOD one would contact about this, since the AJPO no longer exists.)

I rejoice, that you of all people here have found such clear words on
that issue. I've a comment and a suggestion though:

 (1) The present headaches revolve around the changing license for
     libraries like GtkAda and Florist (to just name 2). The FSF
     compiler, if maintained properly, would suffice for most people
     as Ada 95 compiler. So the pressure to argue against Gnat GPL has
     somewhat diminished from a purely practical point of view.

 (2) I'm not sure wether this is only rumor, but I seem to remember
     something, that ACT has discharged its obliagation of keeping the
     compiler (a compiler) free by helping the FSF getting it back
     into the Gcc source tree.

If of course you would find somebody at the DOD to contact, I think it
would at least be intresting to see what exactly will happen. (This is
out of my league, but sometimes it is entertaining an edifying to see
the big players move).

> 
> Of course, IANAL.

IANAL2.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

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

> "Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> > It's also clear that ACT cannot legally impose the GPL on their
> > compilers the way they have. The DOD contract that resulted in GNAT
> > was to create an Ada-95 compiler that is free (in both senses of the
> > word) and may be used to create un-free programs. (These requirements
> > resulted in the GMGPL.) As there is no Ada-0X standard yet, there can
> > be no Ada-0X compilers yet, and in fact the ACT GPL compilers are
> > Ada-95 compilers that implement some of the features of the Ada-0X
> > draft only as an option. It seems ACT is calling them "200[5|6]" to
> > try to get around the requirements of the contract, but it's clear
> > that these are Ada-95 compilers with some extra features optionally
> > available, and so in violation of the contract. (I wonder who at the
> > DOD one would contact about this, since the AJPO no longer exists.)
> >
> > Of course, IANAL.
> 
> What if the contract had a fixed term, and has now expired?
> Conjectures, hypotheses and speculation.

What if it didn't? Me thinks the expiry date was "when there is no Ada
95 anymore in Gnat".

Regards .. Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> 
> That ACT denies they can remember that ever GtkAda was distributed a
> GMGPL. That is about all and at the core of all that. You understand
> that now? Even the older versions from the ACT site are now supposed
> to be under GPL ONLY, even if they didn't change from the way they
> very (as we remember) distributed as GMGPL.

Anything they distributed under the GMGPL remains under the GMGPL. 
They're certainly allowed to change the license they use for GtkAda for 
other distributions, but that doesn't affect the license under which 
they distributed it to you. Florist is maintained and distributed 
elsewhere, it seems, so that doesn't seem such an issue.

But if you're worried that the Florist license might change, get it now 
so you'll have it with the current license if it does.

-- 
Jeff Carter
"Many times we're given rhymes that are quite unsingable."
Monty Python and the Holy Grail
57

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> 
> OK. Let me rephrase: There is exactly 1 low cost compiler, that from
> RR sofware. AFAIS there is no Linux target (which in my context of the
> discussion was always a given necessity). But one compiler hardly
> justifies plural.

Aonix is 10% of the cost of an ACT support contract, so that seems to 
qualify as low cost by comparison. I think they have a free version, 
too. I guess it depends on the definition of low cost.

-- 
Jeff Carter
"Many times we're given rhymes that are quite unsingable."
Monty Python and the Holy Grail
57

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> M E Leypold wrote:
> > That ACT denies they can remember that ever GtkAda was distributed a
> > GMGPL. That is about all and at the core of all that. You understand
> > that now? Even the older versions from the ACT site are now supposed
> > to be under GPL ONLY, even if they didn't change from the way they
> > very (as we remember) distributed as GMGPL.
> 
> Anything they distributed under the GMGPL remains under the
> GMGPL. They're certainly allowed to change the license they use for
> GtkAda for other distributions, but that doesn't affect the license
> under which they distributed it to you. 

Read the other threads. I know that it should be like that, I even
think that it is, but still I'd call that demonstrative amnesia of ACT
unsettling. An attitude that shouldn't be forgiven by the community --
at least I will remember it.

> Florist is maintained and distributed elsewhere, it seems, so that
> doesn't seem such an issue.

Let me ask -- are they allowed to change the license in files in
florist which they hardly changed (only stripping in specifiers from
the procedure definitions) and to which they don't hold the original
copyright AFAICS?

I repeat: My needs are hardly at issue here. This has more and more
become a question of principle. I can't stand FUD.

> But if you're worried that the Florist license might change, get it
> now so you'll have it with the current license if it does.

I did. I have the obsessive habit of mirroring anything of interest,
since I know how things tend to rot away. Like libre.act-europe.fr
did.


Regards --Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Simon Wright]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> Aonix is 10% of the cost of an ACT support contract, so that seems to
> qualify as low cost by comparison. I think they have a free version,
> too. I guess it depends on the definition of low cost.

The free version supports up to (n ~= 35) units, which won't get you
very far!

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

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

> No, we do not learn that.  AdaCore was quite specific in saying
> they're not saying anything at all about the past.

They were also quite specific in saying anything from their libre2
site *including* CVS is GPL. That means any old version from CVS is
GPL too. There never was such a thing as GMGPL.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

M E Leypold wrote :
> Ludovico: If GtkAda 2.4.0 has been acquired before 2005/02, presumably
> from libre2.act-europe.fr, and the past has not changed, why don't you
> let the license as GMGPL?

But I do; the switch to pure GPL will only affect future versions.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> M E Leypold wrote :
> > Ludovico: If GtkAda 2.4.0 has been acquired before 2005/02, presumably
> > from libre2.act-europe.fr, and the past has not changed, why don't you
> > let the license as GMGPL?
> 
> But I do; the switch to pure GPL will only affect future versions.

Ah! I understand. I misunderstood you because you also refered to
Gentoo which who already have switched to GPL fro 2.4.0.

So -- purely hypothetically -- if a fork of 2.4.0 (from your present
version turns) up it might be possible to keep the GPL? Of course
there is a number of issues here (maintenance vs. GMGPL, is anybody
intrested in maintaining old 2.4.0 etc) but the question is purely
hypothetical so far.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

Simon Wright <simon@pushface.org> writes:

> "Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
> 
> > Aonix is 10% of the cost of an ACT support contract, so that seems to
> > qualify as low cost by comparison. I think they have a free version,
> > too. I guess it depends on the definition of low cost.
> 
> The free version supports up to (n ~= 35) units, which won't get you
> very far!

That too. I stopped reading after "is 10% of the cost of an ACT"
since 5000 is not 10% of 15000.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

"Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:

> M E Leypold wrote :
>> Ludovico: If GtkAda 2.4.0 has been acquired before 2005/02, presumably
>> from libre2.act-europe.fr, and the past has not changed, why don't you
>> let the license as GMGPL?
>
> But I do; the switch to pure GPL will only affect future versions.

I just want to know your opinion: Sarge contains GtkAda 2.4.0 which
you obtained at a time when AdaCore called it GMGPL software. There is
no new released version of GtkAda but now 2.4.0 is called GPL. What
version will be in Etch? The same 2.4.0 as in Sarge but under GPL
(which could also be GMGPL), 2.4.0 as downloaded today from
libre2... (which then is GPL anyway) or some newer version from CVS?

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Niklas Holsti]

Michael Bode wrote:
> Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:
> 
> 
>>	Slipping in from the side... A non-support, run-time distribution,
>>license in the $200 range would probably get a lot of takers (price
>>range varying depending upon upgrade procedure 
> 
> 
> I'd be one, let's say somewhere below 1000ￜ for the GPL Edition with
> GMGPL licensing. 

I second (or third :-) that.

-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
      .      @       .

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

Simon Wright wrote:
> 
> The free version supports up to (n ~= 35) units, which won't get you
> very far!

True. But it still qualifies as low cost!

-- 
Jeff Carter
"Ditto, you provincial putz?"
Blazing Saddles
86

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

Michael Bode <m.g.bode@web.de> writes:

> "Ludovic Brenta" <ludovic@ludovic-brenta.org> writes:
>
>> M E Leypold wrote :
>>> Ludovico: If GtkAda 2.4.0 has been acquired before 2005/02, presumably
>>> from libre2.act-europe.fr, and the past has not changed, why don't you
>>> let the license as GMGPL?
>>
>> But I do; the switch to pure GPL will only affect future versions.
>
> I just want to know your opinion: Sarge contains GtkAda 2.4.0 which
> you obtained at a time when AdaCore called it GMGPL software. There is
> no new released version of GtkAda but now 2.4.0 is called GPL. What
> version will be in Etch? The same 2.4.0 as in Sarge but under GPL
> (which could also be GMGPL), 2.4.0 as downloaded today from
> libre2... (which then is GPL anyway) or some newer version from CVS?

A newer version from CVS, under GPL.  That will be version 2.8.1,
which has yet to be tagged in the repository.

Since yo asked for my opinion, here it is: you can use version 2.4.0
(currently in Sarge) under the terms of the GMGPL.  The change of
license from AdaCore does not affect the software you get from Debian;
it only affects the software I will get from AdaCore in the future.
Even if that is not true, I doubt very much that AdaCore would sue
anyone over this issue.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

M E Leypold writes:

> Ludovic Brenta <ludovic@ludovic-brenta.org> writes:
>
>> "Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:
>> > It's also clear that ACT cannot legally impose the GPL on their
>> > compilers the way they have. The DOD contract that resulted in GNAT
>> > was to create an Ada-95 compiler that is free (in both senses of the
>> > word) and may be used to create un-free programs. (These requirements
>> > resulted in the GMGPL.) As there is no Ada-0X standard yet, there can
>> > be no Ada-0X compilers yet, and in fact the ACT GPL compilers are
>> > Ada-95 compilers that implement some of the features of the Ada-0X
>> > draft only as an option. It seems ACT is calling them "200[5|6]" to
>> > try to get around the requirements of the contract, but it's clear
>> > that these are Ada-95 compilers with some extra features optionally
>> > available, and so in violation of the contract. (I wonder who at the
>> > DOD one would contact about this, since the AJPO no longer exists.)
>> >
>> > Of course, IANAL.
>> 
>> What if the contract had a fixed term, and has now expired?
>> Conjectures, hypotheses and speculation.
>
> What if it didn't? Me thinks the expiry date was "when there is no Ada
> 95 anymore in Gnat".
>
> Regards .. Markus

More conjectures, hypotheses and speculation.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Ludovic Brenta]

Niklas Holsti <nobody@nowhere.fi> writes:
> Michael Bode wrote:
>> Dennis Lee Bieber <wlfraed@ix.netcom.com> writes:
>> 
>> 
>>>	Slipping in from the side... A non-support, run-time distribution,
>>>license in the $200 range would probably get a lot of takers (price
>>>range varying depending upon upgrade procedure 
>> 
>> 
>> I'd be one, let's say somewhere below 1000€ for the GPL Edition with
>> GMGPL licensing. 
>
> I second (or third :-) that.

I suggest y'all file support requests and join the GNU Ada project.

-- 
Ludovic Brenta.

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

http://www.oss-in-atm.info/20051207/09-gasperoni.php


  "There is a case in the community: wouldn't it be nice if AdaCore
   provided support for a single developer for 500 euro, but we know
   that isn't going to fly. And we said, "Fine, if somebody else wants
   to take it up, fine, we don't think that business will survive for
   a long time."

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> Simon Wright wrote:
> > The free version supports up to (n ~= 35) units, which won't get you
> > very far!
> 
> True. But it still qualifies as low cost!

That at least is indisputable.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Jeffrey R. Carter]

M E Leypold wrote:
> 
> That too. I stopped reading after "is 10% of the cost of an ACT"
> since 5000 is not 10% of 15000.

Sorry, I remembered $1500 (which is 10% of $15000), but that's only for 
the Windows version. The Linux version is unaccountably much more. I 
guess they think people who use a free OS have more money available for 
compilers.

-- 
Jeff Carter
"I unclog my nose towards you."
Monty Python & the Holy Grail
11

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Jeffrey R. Carter" <spam.not.jrcarter@acm.not.spam.org> writes:

> M E Leypold wrote:
> > That too. I stopped reading after "is 10% of the cost of an ACT"
> > since 5000 is not 10% of 15000.
> 
> Sorry, I remembered $1500 (which is 10% of $15000), but that's only

Never mind. I should also have read you statement to the end. Lame
excuse: It was late and I was already tired from bombing c.l.a with
mails on licensing issues.

> for the Windows version. The Linux version is unaccountably much
> more. I guess they think people who use a free OS have more money
> available for compilers.

Oh certainly.

May I add one thing (not as reproach on what you said, but rather
continuing the free OS theme here): My skill set is primarily in
unixish systems especially as far as the enviroment (available tools,
shell, scripting) goes. In this ongoing discussion I got offered (more
than once) compilers as subsitute which would only run on windows,
have certain important libraries only for windows or are substantially
more expensive on non windows platforms. But I hope one can understand
that as a Unix animal I'm not looking for a windows compiler.

The big advantage of Ada could be that it is really portable. I did
sucessfully develop programs under Linux and just compiled them on
windows. This is big plus. Or it could be. Because if you look at the
compiler market, the choices become very restricted (or expensive) if
one wants to compile on both platforms, develop primarily with Linux
and wants more than the Ada standard library.


Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Marc A. Criley]

Jeffrey R. Carter wrote:
> M E Leypold wrote:
> 
>>
>> That too. I stopped reading after "is 10% of the cost of an ACT"
>> since 5000 is not 10% of 15000.
> 
> 
> Sorry, I remembered $1500 (which is 10% of $15000), but that's only for 
> the Windows version. The Linux version is unaccountably much more. I 
> guess they think people who use a free OS have more money available for 
> compilers.

Though the annual maintenance [for Aonix Ada on Linux] is cheaper:  $1000 
vs $1495.  The break-even point is in seven years.  I agree that it is an 
odd pricing structure.

-- Marc A. Criley
-- McKae Technologies
-- www.mckae.com
-- DTraq - XPath In Ada - XML EZ Out

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Steve]

"Marc A. Criley" <mcNOSPAM@mckae.com> wrote in message 
news:g39pg.434$PE1.242@newsread2.news.pas.earthlink.net...
> Jeffrey R. Carter wrote:
>> M E Leypold wrote:
>>
>>>
>>> That too. I stopped reading after "is 10% of the cost of an ACT"
>>> since 5000 is not 10% of 15000.
>>
>>
>> Sorry, I remembered $1500 (which is 10% of $15000), but that's only for 
>> the Windows version. The Linux version is unaccountably much more. I 
>> guess they think people who use a free OS have more money available for 
>> compilers.
>
> Though the annual maintenance [for Aonix Ada on Linux] is cheaper:  $1000 
> vs $1495.  The break-even point is in seven years.  I agree that it is an 
> odd pricing structure.
>

One more data point.  The last time I enquired the cost was either $10000 or 
$12000 for 10 developer seats, and that was the minimum.  But that was many 
years ago.  We wound up going with ObjectAda at the time which was closer to 
$500 per developer with no minimum.  At the time we also received a discount 
from Aonix by purchasing 5 licenses.

I am in the camp of individuals who think that AdaCore should have a lower 
minimum, although to be fair I haven't asked recently.

Steve
(The Duck)

> -- Marc A. Criley
> -- McKae Technologies
> -- www.mckae.com
> -- DTraq - XPath In Ada - XML EZ Out
> 

Re: can one build commercial applications with latest gnat and other licenses related questions...

[M E Leypold]

"Steve" <nospam_steved94@comcast.net> writes:

> "Marc A. Criley" <mcNOSPAM@mckae.com> wrote in message 
> news:g39pg.434$PE1.242@newsread2.news.pas.earthlink.net...
> > Jeffrey R. Carter wrote:
> >> M E Leypold wrote:
> >>
> >>>
> >>> That too. I stopped reading after "is 10% of the cost of an ACT"
> >>> since 5000 is not 10% of 15000.
> >>
> >>
> >> Sorry, I remembered $1500 (which is 10% of $15000), but that's only for 
> >> the Windows version. The Linux version is unaccountably much more. I 
> >> guess they think people who use a free OS have more money available for 
> >> compilers.
> >
> > Though the annual maintenance [for Aonix Ada on Linux] is cheaper:  $1000 
> > vs $1495.  The break-even point is in seven years.  I agree that it is an 
> > odd pricing structure.
> >
> 
> One more data point.  The last time I enquired the cost was either $10000 or 
> $12000 for 10 developer seats, and that was the minimum.  But that was many 

You're talking about ACT's offer now? We have been talking about Aonix.

> years ago.  We wound up going with ObjectAda at the time which was closer to 
> $500 per developer with no minimum.  At the time we also received a discount 
> from Aonix by purchasing 5 licenses.

I've asked ACT and Aonix for price quotes some time around the
beginning officiating 2005. That was when we did a bid on some
projects where we thought the ability to fall back to a commercial
compiler would perhaps come handy at least in the development stage
(admittedly the idea was to use a commercial compiler during 2 years
of development and maintenance and fall back to an open source
compiler for maintenance when the 

I might remember wrongly, so take all that cautiously:

 - ACT just returned with a number of question (for how many people,
   size of your shoes etc), no price was quoted. Since I aimed for a
   standard package and also wanted a standard price list to give a
   customer an idea how much the transition to Ada (instead of C++
   ...)  would cost, I left it there. I gathered the > 10K fee at that
   time from usenet and that was enough as a lower bracket to remove
   ACT from the considerations in this case.

 - Aonix (I think) actually deigned to answer and at that time you had
   to buy support for 5 developers minimum (at ~ 1000.- EUR per
   seat). Since I'm working freelance most of the time (with changing
   affiliations) and don't have a splut personality I felt somewhat
   cheated here.

By the way: Today, when I went to Aonix' site I find a PDF
http://www.aonix.de/pdf/adt.pdf (which is supposedly the "AonixADT
Product Line Overview"). Surprisingly I need a password to open that
PDF. Seems you need a support contract to see Aonix' product line.

I think both ACT and Aonix could do better in the CRM department. :-/

But to wrap it up:

 - ACT charges 10K - 15K minimum Since pricing seems to be highly
   individual (they talk about a matrix they have to decide on the
   price), you have to ask in every single case.

 - Aonix charges ~ $1000 per developer on Linux but you have to buy
   the windows license separately and the minimum number of seats to
   buy is 5 (makes it ~5000 for the smallest license on Linux).

As I said I could be wrong: Those people are not very forthcoming with
pricing information and I have a hard time finding any pricing list on
their web pages -- perhaps they are much too used to individual
government and big company contracts.

This is also one of the reasons I'd really recommend RR Software
(Janus Ada): They have a price list on their site, which is exactly
what Mr Joe Small Developer surfing along would like to have.

If representatives of the afore mentioned companies read this post I'd
suggest that they just correct any mistakes and misunderstandings.

Regards -- Markus

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Pascal Obry]

M E Leypold a �crit :

> I might remember wrongly, so take all that cautiously:
> 
>  - ACT just returned with a number of question (for how many people,
>    size of your shoes etc), no price was quoted. Since I aimed for a
>    standard package and also wanted a standard price list to give a
>    customer an idea how much the transition to Ada (instead of C++
>    ...)  would cost, I left it there. I gathered the > 10K fee at that
>    time from usenet and that was enough as a lower bracket to remove
>    ACT from the considerations in this case.

And this is your mistake. If AdaCore asked questions it is probably to
find the best arrangement between you and them! So staying with the
"usenet gatrhered > 10k" is just a plain mistake...

AdaCore is really listening to customers, tell them something!

After that we hear so much non sense on this thread...

Pascal.

-- 

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Pascal Obry <pascal@obry.net> writes:

> And this is your mistake. If AdaCore asked questions it is probably to
> find the best arrangement between you and them! So staying with the
> "usenet gatrhered > 10k" is just a plain mistake...

I asked them in 10/2005 and the price for Gnat Pro was 14000ￜ for
Windows and another 7000ￜ if I wanted Linux too. GtkAda was an
additional 7000ￜ. Then I could get 2 additional packages (compiler for
another target, Glade, AWS, ...) at no additional cost. That is for 5
developers, but if we don't have 5 developers, that's out fault.

So the minimal cost for the first year for Windows + Linux + GtkAda is
28000ￜ.

> AdaCore is really listening to customers, tell them something!

I told them that I would be ok with less than stellar support, like any
other Visual C++ or Delphi customer gets, if I could get the packages
for a more Microsoft- or Borland-like price. They told me their
support is far better than Microsoft's or Borland's. That's it.

-- 
Michael Bode

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Pascal Obry]

Michael Bode a ï¿œcrit :

> I told them that I would be ok with less than stellar support, like any
> other Visual C++ or Delphi customer gets, if I could get the packages
> for a more Microsoft- or Borland-like price. They told me their
> support is far better than Microsoft's or Borland's. That's it.

Again I'm not on the maketing side, but to me you are just comparing
apples and oranges. Microsoft or Borland are selling CD. Duplicating a
CD cost just nothing (almost right?, the cost is on dev process) hence
the low price for VC++ for example. AdaCore is selling support. There is
no way to "duplicate" a support, each customer is different, has
different needs and requirement. The cost can't be as low as VC++.

Now what do you want ?

A product without support as for VC++, then use the GNAT GPL, GNAT FSF,
Aonix ObjectAda, Rational Ada, RR Software Janus Ada... What's wrong
with that ? Use the product that fits your needs.

Pascal.

-- 

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

Re: can one build commercial applications with latest gnat and other licenses related questions...

[Michael Bode]

Pascal Obry <pascal@obry.net> writes:

> Now what do you want ?

You said:

> AdaCore is really listening to customers, tell them something!

And I told them something. The problem obviously is: I'm not a customer.

> A product without support as for VC++, then use the GNAT GPL, GNAT
> FSF,

VC++ can produce non GPL binaries, GNAT GPL not. GNAT FSF is ok, but
there is no portable GUI library for GNAT FSF. And even using GtkAda
2.4.0 is problematic as I can't prove that the GtkAda 2.4.0 I use was
downloaded before day X where X is the unknown day when GtkAda 2.4.0
started to be licensed under GPL.

> Aonix ObjectAda, Rational Ada, RR Software Janus Ada... What's wrong
> with that ? Use the product that fits your needs.

As I said: Windows + Linux + Gtk(Ada) (or other portable GUI
lib). None of the above AFAIK can offer that.

Of course AdaCore can do what they want: if they're only interested in
customers from 14000ￜ/year upwards, that's fine. But then I have to
look for things like GNAT FSF and old versions of GtkAda for which I
hope AdaCore won't sue me if I use them and consider other options for
the future (Java might be one).

-- 
Michael Bode