Ada_Arrays Project

Ada_Arrays Project

[M?rio Amado Alves]

(Sorry for duplicates. I've sent this message before via
comp.lang.ada@ada-france.org but for some reason it does not seem to
have reached the newsgroup. I'm using Google now, so my replies will
be tardy.)

I'm entertaining the idea of a generic package Ada_Arrays and children
where Ada_Arrays is like Ada.Strings except the element type is
generic instead of Character.

Anybody besides me foresees the great usefulness of Ada_Arrays?
Anybody did it?
Anybody interested in collaborating?

I'm considering doing it by program(med) transformation of existing
source-code for Ada.Strings. My first studies indicate semantic
analysis is required, so this might configure an ASIS application.

(I'm also interested in this project as a test case for open source
software licensing. I'd like to release it under a "commercial open
source" license like one of those endorsed by the Software Developers
Cooperative (www.softdevelcoop.org). Specially if done by way of
program transformation from the GPLed GNAT, the licensing of
Ada_Arrays is a complex issue.)

Re: Ada_Arrays Project

[Preben Randhol]

On 2003-11-27, M?rio Amado Alves <amado.alves@netcabo.pt> wrote:
> Anybody besides me foresees the great usefulness of Ada_Arrays?

No, but please explain why you find it useful :-)

> (I'm also interested in this project as a test case for open source
> software licensing. I'd like to release it under a "commercial open
> source" license like one of those endorsed by the Software Developers
> Cooperative (www.softdevelcoop.org). Specially if done by way of

Why? What is wrong with GMGPL?

> program transformation from the GPLed GNAT, the licensing of
> Ada_Arrays is a complex issue.)


-- 
"Saving keystrokes is the job of the text editor, not the programming
 language."

Re: Ada_Arrays Project

[M?rio Amado Alves]

> > Anybody besides me foresees the great usefulness of Ada_Arrays?
> No, but please explain why you find it useful :-)

Standard fixed, bounded and unbounded arrays of *any* element type
sounds yummy to me. I know there are some libraries out there that
provide things like this. Advantages of generating it from Ada.Strings
would include:
- interface and semantics already very well known by users
- validation carried over from Ada.Strings
- added facilities e.g. Find_Token
- fun ASIS project

But if "no"--*nobody* foresees the usefulness--then I'll accept I'm
not seeing right and I'll drop the project on the spot and say thanks
for the eye operation and I'm sorry for the all the fuss.

> > (I'm also interested in this project as a test case for open source
> > software licensing. I'd like to release it under a "commercial open
> > source" license...
> Why? What is wrong with GMGPL?

Nothing "wrong", it just isn't commercial. If you don't want to make
money then GMGPL is fine. The basic idea of commercial open source is
that if an (open source) component ends up being part of a lucrative
setup then the authors of the component receive a fair reward. Before
objecting please review the material indexed on the SDC site, for
example my essay Open Source Business Found Parasitic, Bob Leif's Ada
Developers Cooperative License, and the Commented Conditions of Use of
SDC Artifacts. And I suggest we don't discuss the general issue here
(it's OT). You're very welcome to do it on the SDC forum of course
(www.softdevelcoop.org).

Re: Ada_Arrays Project

[Georg Bauhaus]

M?rio Amado Alves <amado.alves@netcabo.pt> wrote:
:> > Anybody besides me foresees the great usefulness of Ada_Arrays?
:> No, but please explain why you find it useful :-)
: 
: Standard fixed, bounded and unbounded arrays of *any* element type
: sounds yummy to me.

For a start, you could turn the non-Character items into satellite
data orbitting a Character value?


-- Georg

Re: Ada_Arrays Project

[M?rio Amado Alves]

> : Standard fixed, bounded and unbounded arrays of *any* element type
> : sounds yummy to me.
> 
> For a start, you could turn the non-Character items into satellite
> data orbitting a Character value?

It's a possibility, specially if you use wide characters for 16-bit
addressing because 8-bit addressing is clearly too short. But anyway I
don't like it because eventually you would have to do memory
management and that is precisely one of the things that I want
Ada_Arrays to encapsulate, namely in the magical
Ada_Arrays.Unbounded.Unbounded_String (Unbounded_Array?) type.

(I'm using Google to read/write CLA, but it's too slow.
comp.lang.ada@ada-france.org does not seem to be working for me.
Anyone can recommend a fix or an alternate gateway? Thanks.)

Re: Ada_Arrays Project

[Jeff C,]

"M?rio Amado Alves" <amado.alves@netcabo.pt> wrote in message
news:17eddf9f.0311271236.2a1ded9f@posting.google.com...
> > Why? What is wrong with GMGPL?
>
> Nothing "wrong", it just isn't commercial. If you don't want to make
> money then GMGPL is fine. The basic idea of commercial open source is
> that if an (open source) component ends up being part of a lucrative
> setup then the authors of the component receive a fair reward. Before


Quick..Someone call ACT and tell them that all that money people have been
paying them is not for a commercial venture.

Re: Ada_Arrays Project

[M?rio Amado Alves]

"Jeff C," <nolongersafeto@userealemailsniff.com> wrote:
> > Nothing "wrong", it just isn't commercial. If you don't want to make
> > money then GMGPL is fine. The basic idea of commercial open source is
> > that if an (open source) component ends up being part of a lucrative
> > setup then the authors of the component receive a fair reward. Before
> 
> 
> Quick..Someone call ACT and tell them that all that money people have been
> paying them is not for a commercial venture.

Please read my paragraph in full.

Re: Ada_Arrays Project

[Preben Randhol]

On 2003-11-27, M?rio Amado Alves <amado.alves@netcabo.pt> wrote:
> Nothing "wrong", it just isn't commercial. If you don't want to make
> money then GMGPL is fine. The basic idea of commercial open source is
> that if an (open source) component ends up being part of a lucrative
> setup then the authors of the component receive a fair reward. 

Yes, exactly, a big fat *IF*.

> Before objecting please review the material indexed on the SDC site,
> for example my essay Open Source Business Found Parasitic, Bob Leif's
> Ada Developers Cooperative License, and the Commented Conditions of
> Use of SDC Artifacts. And I suggest we don't discuss the general issue
> here (it's OT). You're very welcome to do it on the SDC forum of
> course (www.softdevelcoop.org).

Well, whatever. I wouldn't touch the library if it has some strange
license. If you are talking about a *program* things are different, but
a *library*...

-- 
"Saving keystrokes is the job of the text editor, not the programming
 language."

Re: Ada_Arrays Project

[Marius Amado Alves]

Preben Randhol wrote:
> On 2003-11-27, M?rio Amado Alves <amado.alves@netcabo.pt> wrote:
> 
>>Nothing "wrong", it just isn't commercial. If you don't want to make
>>money then GMGPL is fine. The basic idea of commercial open source is
>>that if an (open source) component ends up being part of a lucrative
>>setup then the authors of the component receive a fair reward. 
> 
> Yes, exactly, a big fat *IF*.

Yes, exactly, and fat is good in this case: the user only owes when 
revenue is generated. Please see the discussions.

>>Before objecting please review the material indexed on the SDC site,
>>for example my essay Open Source Business Found Parasitic, Bob Leif's
>>Ada Developers Cooperative License, and the Commented Conditions of
>>Use of SDC Artifacts. And I suggest we don't discuss the general issue
>>here (it's OT). You're very welcome to do it on the SDC forum of
>>course (www.softdevelcoop.org).
 >
> Well, whatever. I wouldn't touch the library if it has some strange
> license. If you are talking about a *program* things are different, but
> a *library*...

I know some users dislike "strange" licences, but this attitude is often 
irrational. It is in this case. The definition of program vs. library is 
one of the complex issues aluded. So much so that it is a problem in 
GPL, and even on the LGPL which was done to overcome that (but strictly 
did not). Some of us have been trying to solve the problem with the 
concept of "use of a component". Again, please see the discussions. I 
won't argue here anymore unless you mark this thread OT and people don't 
complaint.

SVG to PNG [was: Ada_Arrays Project]

[Nick Roberts]

M?rio Amado Alves wrote:

> I'm entertaining the idea of a generic package Ada_Arrays and children
> where Ada_Arrays is like Ada.Strings except the element type is
> generic instead of Character.
> 
> Anybody besides me foresees the great usefulness of Ada_Arrays?
> Anybody did it?

I think it's the sort of thing that has been done before by lots of people, 
generally not in a very organised or uniform way. I would guess the biggest 
problem would be that people would be reluctant to pay for this sort of 
thing; they would think "Crikey, I could just as easily write that myself," 
even if they actually were wrong.

> I'm considering doing it by program(med) transformation of existing
> source-code for Ada.Strings. My first studies indicate semantic
> analysis is required, so this might configure an ASIS application.

I suggest that much of the interfaces of the Ada.Strings.*_Strings packages 
are specific to the Character (component) type, and not sensibly applicable 
to a generic array interface. I think this sort of package would be sailing 
perilously close to the treacherous waters of the 'containers controversy'.

> (I'm also interested in this project as a test case for open source
> software licensing. I'd like to release it under a "commercial open
> source" license like one of those endorsed by the Software Developers
> Cooperative (www.softdevelcoop.org). Specially if done by way of
> program transformation from the GPLed GNAT, the licensing of
> Ada_Arrays is a complex issue.)

Allow me to suggest an alternative project, which I believe has a gap in 
the market at the moment, and could be commercially attractive.

I am not aware of any commercial program at the moment which can convert a 
standard SVG source file into a suitable image format file (I suggest PNG 
as the first target to aim at). There is the SVG viewer offered by Adobe, 
but it has problems (bugs, deviation from the standard, certain standard 
facilities not supported, slowness/inefficiency), and I'm not sure that it 
can actually genarate image files anyway. There may be some high-end 
products taht are very expensive, but if so that seems to leave a gap at 
the lower-cost end.

I think this would be a suitable product to be written in Ada, because it 
would require little or no platform-specific code, and it would suit Ada's 
strengths: efficiency and reliability. It would be impossible to 
accommodate any interactive features, of course, and animation would lead 
us to the complexities of animated GIFs or MNG files (but might be possible).

I understand there are some Ada XML packages available (I'm not sure on 
what terms), one of which could be used as a starting point for reading the 
SVG source files. I believe there is an Ada PNG package (again I'm not sure 
about its availablity).

There are further possibilities with conversions between HTML, XSL-FO, 
PS/EPS/PDF, and other presentation formats.

 > Anybody interested in collaborating?

If you're interested in something like this, Mario, I might be interested 
in collaborating. It would be important to plan the project carefully.

-- 
Nick Roberts

Re: SVG to PNG [was: Ada_Arrays Project]

[Rodrigo Garcia]

You already have an open source product for this (in Java): Batik from the
Apache Software Foundation.
"Another possibility is to use Batik's modules to convert SVG to various
formats, such as raster images (JPEG,PNG or Tiff)."

More in:
http://xml.apache.org/batik/index.html

Rodrigo

"Nick Roberts" <nick.roberts@acm.org> wrote in message
news:bq5r6r$1uos2e$1@ID-25716.news.uni-berlin.de...
> M?rio Amado Alves wrote:
>
> > I'm entertaining the idea of a generic package Ada_Arrays and children
> > where Ada_Arrays is like Ada.Strings except the element type is
> > generic instead of Character.
> >
> > Anybody besides me foresees the great usefulness of Ada_Arrays?
> > Anybody did it?
>
> I think it's the sort of thing that has been done before by lots of
people,
> generally not in a very organised or uniform way. I would guess the
biggest
> problem would be that people would be reluctant to pay for this sort of
> thing; they would think "Crikey, I could just as easily write that
myself,"
> even if they actually were wrong.
>
> > I'm considering doing it by program(med) transformation of existing
> > source-code for Ada.Strings. My first studies indicate semantic
> > analysis is required, so this might configure an ASIS application.
>
> I suggest that much of the interfaces of the Ada.Strings.*_Strings
packages
> are specific to the Character (component) type, and not sensibly
applicable
> to a generic array interface. I think this sort of package would be
sailing
> perilously close to the treacherous waters of the 'containers
controversy'.
>
> > (I'm also interested in this project as a test case for open source
> > software licensing. I'd like to release it under a "commercial open
> > source" license like one of those endorsed by the Software Developers
> > Cooperative (www.softdevelcoop.org). Specially if done by way of
> > program transformation from the GPLed GNAT, the licensing of
> > Ada_Arrays is a complex issue.)
>
> Allow me to suggest an alternative project, which I believe has a gap in
> the market at the moment, and could be commercially attractive.
>
> I am not aware of any commercial program at the moment which can convert a
> standard SVG source file into a suitable image format file (I suggest PNG
> as the first target to aim at). There is the SVG viewer offered by Adobe,
> but it has problems (bugs, deviation from the standard, certain standard
> facilities not supported, slowness/inefficiency), and I'm not sure that it
> can actually genarate image files anyway. There may be some high-end
> products taht are very expensive, but if so that seems to leave a gap at
> the lower-cost end.
>
> I think this would be a suitable product to be written in Ada, because it
> would require little or no platform-specific code, and it would suit Ada's
> strengths: efficiency and reliability. It would be impossible to
> accommodate any interactive features, of course, and animation would lead
> us to the complexities of animated GIFs or MNG files (but might be
possible).
>
> I understand there are some Ada XML packages available (I'm not sure on
> what terms), one of which could be used as a starting point for reading
the
> SVG source files. I believe there is an Ada PNG package (again I'm not
sure
> about its availablity).
>
> There are further possibilities with conversions between HTML, XSL-FO,
> PS/EPS/PDF, and other presentation formats.
>
>  > Anybody interested in collaborating?
>
> If you're interested in something like this, Mario, I might be interested
> in collaborating. It would be important to plan the project carefully.
>
> -- 
> Nick Roberts
>

Re: SVG to PNG [was: Ada_Arrays Project]

[Dr Steve Sangwine]

On Thu, 27 Nov 2003 21:43:45 +0000, Nick Roberts
<nick.roberts@acm.org> wrote:

>M?rio Amado Alves wrote:
>
>> I'm entertaining the idea of a generic package Ada_Arrays and children
>> where Ada_Arrays is like Ada.Strings except the element type is
>> generic instead of Character.
>> 
>> Anybody besides me foresees the great usefulness of Ada_Arrays?
>> Anybody did it?
....................
>
>Allow me to suggest an alternative project, which I believe has a gap in 
>the market at the moment, and could be commercially attractive.
>
..................
>I understand there are some Ada XML packages available (I'm not sure on 
>what terms), one of which could be used as a starting point for reading the 
>SVG source files. I believe there is an Ada PNG package (again I'm not sure 
>about its availablity).

Indeed there is an Ada PNG package, called PNG_IO. I wrote it, and it
is GPL Open Source. The website is:
http://privatewww.essex.ac.uk/~sjs/png_io/png_io.html

>
>There are further possibilities with conversions between HTML, XSL-FO, 
>PS/EPS/PDF, and other presentation formats.
>
> > Anybody interested in collaborating?
>
>If you're interested in something like this, Mario, I might be interested 
>in collaborating. It would be important to plan the project carefully.

Re: SVG to PNG [was: Ada_Arrays Project]

[Preben Randhol]

On 2003-11-28, Dr Steve Sangwine <sjs@essex.ac.uk> wrote:
>
> Indeed there is an Ada PNG package, called PNG_IO. I wrote it, and it
> is GPL Open Source. The website is:
> http://privatewww.essex.ac.uk/~sjs/png_io/png_io.html

Not GMGPL?

-- 
"Saving keystrokes is the job of the text editor, not the programming
 language."

Re: SVG to PNG [was: Ada_Arrays Project]

[Simon Wright]

Preben Randhol <randhol+valid_for_reply_from_news@pvv.org> writes:

> On 2003-11-28, Dr Steve Sangwine <sjs@essex.ac.uk> wrote:
> >
> > Indeed there is an Ada PNG package, called PNG_IO. I wrote it, and it
> > is GPL Open Source. The website is:
> > http://privatewww.essex.ac.uk/~sjs/png_io/png_io.html
> 
> Not GMGPL?

I saw that, but it's a university that has decided to allow the
software to be released under the GPL .. who knows if it's a
deliberate decision to go GPL or just the state of play on the day ..

-- 
Simon Wright                               100% Ada, no bugs.

Re: SVG to PNG [was: Ada_Arrays Project]

[Preben Randhol]

On 2003-12-01, Simon Wright <simon@pushface.org> wrote:
> Preben Randhol <randhol+valid_for_reply_from_news@pvv.org> writes:
>
>> On 2003-11-28, Dr Steve Sangwine <sjs@essex.ac.uk> wrote:
>> >
>> > Indeed there is an Ada PNG package, called PNG_IO. I wrote it, and it
>> > is GPL Open Source. The website is:
>> > http://privatewww.essex.ac.uk/~sjs/png_io/png_io.html
>> 
>> Not GMGPL?
>
> I saw that, but it's a university that has decided to allow the
> software to be released under the GPL .. who knows if it's a
> deliberate decision to go GPL or just the state of play on the day ..

Would still be GPL, only a bit looser :-)

Well, I guess it isn't to hard to make a new binding.

-- 
"Saving keystrokes is the job of the text editor, not the programming
 language."

Re: SVG to PNG [was: Ada_Arrays Project]

[M?rio Amado Alves]

> > I'm entertaining the idea of a generic package Ada_Arrays and children
> > where Ada_Arrays is like Ada.Strings except the element type is
> > generic instead of Character.
> > 
> > Anybody besides me foresees the great usefulness of Ada_Arrays?
> > Anybody did it?
> 
> I think it's the sort of thing that has been done before by lots of people, 
> generally not in a very organised or uniform way. I would guess the biggest 
> problem would be that people would be reluctant to pay for this sort of 
> thing; they would think "Crikey, I could just as easily write that myself," 
> even if they actually were wrong.

I would be willing to pay--under the referenced conditions.