GNADE still alive?

GNADE still alive?

[Alan Jump]

I've come across a porting project that needs to talk to an existing SQLite database. A quick search brings me to Sourceforge and the GNADE project...but I see no activity in that project since 2008. Is it still being developed? Or should I look elsewhere for an existing package?

Re: GNADE still alive?

[Vadim Godunko]

On Apr 24, 3:47 am, Alan Jump <alan.j...@gmail.com> wrote:
> I've come across a porting project that needs to talk to an existing SQLite database. A quick search brings me to Sourceforge and the GNADE project...but I see no activity in that project since 2008. Is it still being developed? Or should I look elsewhere for an existing package?

There are several bindings to SQLite, look here:

http://en.wikibooks.org/wiki/Ada_Programming/Libraries/Database

Re: GNADE still alive?

[Thomas Løcke]

On 04/24/2012 01:47 AM, Alan Jump wrote:
> I've come across a porting project that needs to talk to an existing SQLite database. A quick search brings me to Sourceforge and the GNADE project...but I see no activity in that project since 2008. Is it still being developed? Or should I look elsewhere for an existing package?


You could give GNATCOLL.SQL a try. It works with SQLite out of the box.

http://docs.adacore.com/gnatcoll-docs/sql.html

-- 
Thomas L�cke | thomas@12boo.net | http://12boo.net

Re: GNADE still alive?

[Ludovic Brenta]

Alan Jump wrote on comp.lang.ada:
> I've come across a porting project that needs to talk to an existing
> SQLite database. A quick search brings me to Sourceforge and the GNADE
> project...but I see no activity in that project since 2008. Is it still
> being developed? Or should I look elsewhere for an existing package?

GNADE is mature and stable and it has the advantage of being available
as a precompiled binary package, with long-term support, in Debian. Other
solutions require you to compile them yourself.

-- 
Ludovic Brenta.

Re: GNADE still alive?

[Thomas Løcke]

On 04/24/2012 09:08 AM, Ludovic Brenta wrote:
> GNADE is mature and stable and it has the advantage of being available
> as a precompiled binary package, with long-term support, in Debian. Other
> solutions require you to compile them yourself.



GNADE appears pretty dead, at least as far as updates to the code goes,
unless of course I'm looking in all the wrong places...  :o)

Is the Debian package build using the vanilla sources from the project
website, or do you guys add some of your own magic sauce?

GNADE might be the best thing since sliced bread, but looking at it
from the outside, it sure do appears to be somewhat stale.

-- 
Thomas L�cke | thomas@12boo.net | http://12boo.net

Re: GNADE still alive?

[Dmitry A. Kazakov]

On Mon, 23 Apr 2012 16:47:04 -0700 (PDT), Alan Jump wrote:

> I've come across a porting project that needs to talk to an existing
> SQLite database. A quick search brings me to Sourceforge and the GNADE
> project...but I see no activity in that project since 2008. Is it still
> being developed?

GNADE is more or less stable, but lacks maintenance.

I am using only GNADE ODBC, because to me it does not make sense to run any
DBMS-specific ones. Except for SQLite, which is arguably not a DBMS in
traditional sense. GNADE ODBC is packaged for Debian, but still absent in
Fedora. So I had to package it with my stuff for Fedora. Under Windows it
looks worse. I am using GNADE ODBC sources patched to be compatible with
Ada 2005.

> Or should I look elsewhere for an existing package?

I created medium-thick SQLite bindings specifically for static linking to
the SQLite amalgation:

http://www.dmitry-kazakov.de/ada/components.htm#SQLite

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

Re: GNADE still alive?

[Ludovic Brenta]

Thomas Løcke wrote on comp.lang.ada:
> Ludovic Brenta wrote:
>> GNADE is mature and stable and it has the advantage of being available
>> as a precompiled binary package, with long-term support, in Debian. Other
>> solutions require you to compile them yourself.
> 
> GNADE appears pretty dead, at least as far as updates to the code goes,
> unless of course I'm looking in all the wrong places...  :o)

Ada appears dead to quite a few people, too, but it's still working very
well, thank you very much.

The truth is, Free Software never dies.  GNADE may not have received much
love in the past few years but it is the basis of GNATCOLL and the sources
live on :)  For that matter, it would be nice if someone would invest a few
dozen man-hours in packaging GNATCOLL for Debian, see
http://bugs.debian.org/640532.

> Is the Debian package build using the vanilla sources from the project
> website, or do you guys add some of your own magic sauce?

Only a little fairy dust to make the package compliant to the Debian Policy
for Ada.  We did make some changes that we contributed upstream.  Stephe
may have more to say about this.

> GNADE might be the best thing since sliced bread, but looking at it
> from the outside, it sure do appears to be somewhat stale.

The MySQL and PostgreSQL parts of GNADE are definitely stale and not
in the binary packages in Debian.  The ODBC and SQLite bindings work
and will only become stale when the ODBC or SQLite break backward
compatibility.

Note: I am not an author of GNADE, I handed over maintenance of the
Debian package to Stephe years ago and I have no interest in promoting
GNADE over other solutions.  I'm only providing information that might
be relevant for the OP.  As you know, I am of the opinion that
binary packaging matters a lot, especially to beginners and professionals,
maybe less to enthusiasts who like to recompile everything for themselves.

BTW, there is also APQ in Debian but it only has a PostgreSQL binding
(not stale) at the moment, so I didn't mention it at first.

-- 
Ludovic Brenta.

Re: GNADE still alive?

[Thomas Løcke]

On 04/24/2012 10:07 AM, Ludovic Brenta wrote:
> Ada appears dead to quite a few people, too, but it's still working very
> well, thank you very much.



I know. YAY!  \o/



> The truth is, Free Software never dies.  GNADE may not have received much
> love in the past few years but it is the basis of GNATCOLL and the sources
> live on :)  For that matter, it would be nice if someone would invest a few
> dozen man-hours in packaging GNATCOLL for Debian, see
> http://bugs.debian.org/640532.



Yea, that would be pretty nice.

Work is underway towards getting GNATColl packaged for Slackware:

https://github.com/qunying/ada-4-slackware

Personally I'm one of those compile nuts, who just grab the latest svn
and go from there.


-- 
Thomas L�cke | thomas@12boo.net | http://12boo.net

Re: GNADE still alive?

[Stephen Leake]

Alan Jump <alan.jump@gmail.com> writes:

> I've come across a porting project that needs to talk to an existing
> SQLite database. A quick search brings me to Sourceforge and the GNADE
> project...but I see no activity in that project since 2008. Is it
> still being developed? Or should I look elsewhere for an existing
> package?

<rant>
Why do people care if something is being "developed"? I'd much more care
if it _works_, and then if there is support for bug fixing and new
feature adding, or if I can understand the code well enough to fix
it/improve it myself.

People still rely on CVS; it hasn't been under development for decades!

Windows is constantly under developement; would you choose that for a
long-term critical project?
</rant>

I worked on GNADE some; I believe it works well enough to be a starting
point, but you should worry about support for both bugs and new features
that you might need. There are no active GNADE developers, so you'd have
to do that yourself. That should be easier than any of the other
database front ends I'm aware of.

I think the pro versions of GNATcoll include SQLite, but not the current
public version. I don't like the way AdaCore writes code (although I do
like their compiler :), so I would find it hard to maintain/improve
GNATcoll. But if you can afford an AdaCore support contract, that is
definitely the way to go.

QtAda has a database abstraction, and I think it can use an SQLite
backend, but I didn't get very far trying to use it. There's a big C++
intermediate between the Ada and the SQLite API, so it would be a pain
to understand and maintain.

-- 
-- Stephe

Re: GNADE still alive?

[Georg Bauhaus]

On 24.04.12 14:36, Stephen Leake wrote:

> <rant>
> Why do people care if something is being "developed"?...
> </rant>

Individuals rely on collective intelligence and "collective"
is always important, since human beings are social beings;
a project that will make an individual appear to be a loner is not
attractive (unless the individual wishes to be a lonesome hero
on her way to disrupt the market).

In any case, the solution to the perceptibility problem is easy:

1. Add a note to file README stating the software can be compiled
   on current systems, naming them.

2. Check in the new README which adds credible signs of project
   activity.

Re: GNADE still alive?

[Thomas Løcke]

On 04/24/2012 02:36 PM, Stephen Leake wrote:
> I think the pro versions of GNATcoll include SQLite, but not the current
> public version. I don't like the way AdaCore writes code (although I do
> like their compiler :), so I would find it hard to maintain/improve
> GNATcoll. But if you can afford an AdaCore support contract, that is
> definitely the way to go.



There's SQLite support in the current dev version of GNATColl:

svn co http://svn.eu.adacore.com/anonsvn/Dev/trunk/gps/gnatlib/

Whether or not Alan is willing to risk using the latest SVN checkout
is another matter entirely. Personally I've never had any issues with
GNATColl dev.

re. your rant: I care about a project being actively developed because
I don't trust my own abilities. I get a certain feeling of safety and
comfort from knowing that there are people constantly working on fixing,
maintaining and improving on software X.

If I _feel_ software has been abandoned, I'm much less inclined to start
using it, unless I'm certain I can actually manage bug fixes and mainte-
nance myself, in which case I would much rather try to bring the project
back to life and hopefully entice other developers to join, as opposed
to just using and maintaining it by myself in my own walled garden.

Did that even make sense?  :)

-- 
Thomas L�cke | thomas@12boo.net | http://12boo.net

Re: GNADE still alive?

[Alan Jump]

On Tuesday, April 24, 2012 7:00:46 AM UTC-7, Thomas Løcke wrote:
> On 04/24/2012 02:36 PM, Stephen Leake wrote:
> > I think the pro versions of GNATcoll include SQLite, but not the current
> > public version. I don't like the way AdaCore writes code (although I do
> > like their compiler :), so I would find it hard to maintain/improve
> > GNATcoll. But if you can afford an AdaCore support contract, that is
> > definitely the way to go.
> 
> 
> 
> There's SQLite support in the current dev version of GNATColl:
> 
> svn co http://svn.eu.adacore.com/anonsvn/Dev/trunk/gps/gnatlib/
> 
> Whether or not Alan is willing to risk using the latest SVN checkout
> is another matter entirely. Personally I've never had any issues with
> GNATColl dev.
> 
> re. your rant: I care about a project being actively developed because
> I don't trust my own abilities. I get a certain feeling of safety and
> comfort from knowing that there are people constantly working on fixing,
> maintaining and improving on software X.
> 
> If I _feel_ software has been abandoned, I'm much less inclined to start
> using it, unless I'm certain I can actually manage bug fixes and mainte-
> nance myself, in which case I would much rather try to bring the project
> back to life and hopefully entice other developers to join, as opposed
> to just using and maintaining it by myself in my own walled garden.
> 
> Did that even make sense?  :)

Made sense to me. :-)

I realize now that I made a glaring omission or three when I mentioned the project. First, the project itself is FOSS. That makes my operating and development budget precisely zero dollars, which precludes the "pro" versions of GNATcoll and the like, which is why I first searched Sourceforge. I'm further limited, budget-wise, in that my development machine is also my personal WinXX machine, and due to age and lack of processor power, I can't virtualize a Linux development environment. I have the GNAT GPL environment running, and running well, and unless I've been terribly misinformed, I can do cross-platform releases from it, so I'm even more reluctant to try and switch horses.

If GNADE is considered stable enough for routine use, I'll give it a go. The project I'm looking at is itself about four years old and doesn't appear to be actively worked on for at least two years, which is why I'm looking at dabbling with it. It should prove...educational. Doubly so, since I am nearly a complete Ada neophyte.

I think that's enough curve balls to throw for one morning.

Re: GNADE still alive?

[Jacob Sparre Andersen]

Thomas L�cke wrote:

> If I _feel_ software has been abandoned, [...]

That puts all Ada software at a disadvantage, since there are fewer bugs
to correct.

I'm perfectly happy riding metro line 14 in Paris, even if they still
are running on version 1.0 of the control software.

Correct software is good software - even if it is a decade or two old -
IMO.

Greetings,

Jacob
-- 
"Never trust a statistic you have not falsified yourself."

Re: GNADE still alive?

[Jeffrey Carter]

On 04/24/2012 07:00 AM, Thomas L�cke wrote:
>
> re. your rant: I care about a project being actively developed because
> I don't trust my own abilities. I get a certain feeling of safety and
> comfort from knowing that there are people constantly working on fixing,
> maintaining and improving on software X.

There's a little program out there called Mine Detector 
(http://pragmada.x10hosting.com/mindet.html). It hasn't been changed since 2010 
Jan, because it's working fine. It's not dead, simply mature and stable.

-- 
Jeff Carter
"Apart from the sanitation, the medicine, education, wine,
public order, irrigation, roads, the fresh water system,
and public health, what have the Romans ever done for us?"
Monty Python's Life of Brian
80

Re: GNADE still alive?

[Adam Beneschan]

On Tuesday, April 24, 2012 4:06:40 PM UTC-7, Jeffrey Carter wrote:
> On 04/24/2012 07:00 AM, Thomas Løcke wrote:
> >
> > re. your rant: I care about a project being actively developed because
> > I don't trust my own abilities. I get a certain feeling of safety and
> > comfort from knowing that there are people constantly working on fixing,
> > maintaining and improving on software X.
> 
> There's a little program out there called Mine Detector 
> (http://pragmada.x10hosting.com/mindet.html). It hasn't been changed since 2010 
> Jan, because it's working fine. It's not dead, simply mature and stable.

Yeah, although you have to admit that the list of feature requests for a program like Mine Detector tends to be somewhat smaller than for something like GNADE.  Not to mention that the ISO/ANSI standard for SQL probably gets revised a little more often than the ISO/ANSI standard for mine detection programs.  :)

                     -- Adam

Re: GNADE still alive?

[Thomas Løcke]

On 04/24/2012 08:45 PM, Jacob Sparre Andersen wrote:
> Thomas L�cke wrote:
>
>> If I _feel_ software has been abandoned, [...]
>
> That puts all Ada software at a disadvantage, since there are fewer bugs
> to correct.
>
> I'm perfectly happy riding metro line 14 in Paris, even if they still
> are running on version 1.0 of the control software.



That is obviously true, but equally obvious is the fact that the control
software for metro line 14 in Paris does not have to link against fast
moving targets.

SQLite is not the same today as it were back in 2007 and neither is
PostgreSQL.

Software is not born equal.

I fully expect the french engineers will keep their metro line control
software updated according to the various physical systems it controls,
it just so happens that trains and tracks are updated and changed a lot
less than something like SQLite.

I do _not_ think something like GNADE can be compared to the control
software for metro line 14 in Paris. It is two completely different
worlds and two completely different beasts - apples and oranges, if you
will.



> Correct software is good software - even if it is a decade or two old -
> IMO.



I fully agree.


-- 
Thomas L�cke | thomas@12boo.net | http://12boo.net

Re: GNADE still alive?

[Vadim Godunko]

On Apr 24, 4:36 pm, Stephen Leake <stephen_le...@stephe-leake.org>
wrote:
>
> I think the pro versions of GNATcoll include SQLite, but not the current
> public version. I don't like the way AdaCore writes code (although I do
> like their compiler :), so I would find it hard to maintain/improve
> GNATcoll. But if you can afford an AdaCore support contract, that is
> definitely the way to go.
>
> QtAda has a database abstraction, and I think it can use an SQLite
> backend, but I didn't get very far trying to use it. There's a big C++
> intermediate between the Ada and the SQLite API, so it would be a pain
> to understand and maintain.
>
Another alive option is SQL module of Matreshka, its SQLite3 backend
is deployed on hundreds of machines.

Re: GNADE still alive?

[Stephen Leake]

Georg Bauhaus <rm.dash-bauhaus@futureapps.de> writes:

> On 24.04.12 14:36, Stephen Leake wrote:
>
>> <rant>
>> Why do people care if something is being "developed"?...
>> </rant>
>
> Individuals rely on collective intelligence and "collective"
> is always important, since human beings are social beings;
> a project that will make an individual appear to be a loner is not
> attractive (unless the individual wishes to be a lonesome hero
> on her way to disrupt the market).

So you are more interested in how many people are knowledgeable about
using GNADE (and are willing to share that knowledge); still not
directly related to whether it is being developed (CVS is the best
example here).

-- 
-- Stephe

Re: GNADE still alive?

[Stephen Leake]

Thomas Løcke <thomas@12boo.net> writes:

> On 04/24/2012 02:36 PM, Stephen Leake wrote:
>> I think the pro versions of GNATcoll include SQLite, but not the current
>> public version. I don't like the way AdaCore writes code (although I do
>> like their compiler :), so I would find it hard to maintain/improve
>> GNATcoll. But if you can afford an AdaCore support contract, that is
>> definitely the way to go.
>
>
>
> There's SQLite support in the current dev version of GNATColl:
>
> svn co http://svn.eu.adacore.com/anonsvn/Dev/trunk/gps/gnatlib/
>
> Whether or not Alan is willing to risk using the latest SVN checkout
> is another matter entirely. Personally I've never had any issues with
> GNATColl dev.
>
> re. your rant: I care about a project being actively developed because
> I don't trust my own abilities. I get a certain feeling of safety and
> comfort from knowing that there are people constantly working on fixing,
> maintaining and improving on software X.
>
> If I _feel_ software has been abandoned, I'm much less inclined to start
> using it, unless I'm certain I can actually manage bug fixes and mainte-
> nance myself, in which case I would much rather try to bring the project
> back to life and hopefully entice other developers to join, as opposed
> to just using and maintaining it by myself in my own walled garden.
>
> Did that even make sense?  :)

That is pretty much what I said; any project needs support for bugs and
enhancements. If you can't provide that support yourself, you care about
project developers.

-- 
-- Stephe

Re: GNADE still alive?

[Stephen Leake]

Thomas Løcke <thomas@12boo.net> writes:

> On 04/24/2012 08:45 PM, Jacob Sparre Andersen wrote:
> SQLite is not the same today as it were back in 2007 and neither is
> PostgreSQL.

As long as any the API is backward-compatible, GNADE is still usefull. 

It might be _more_ useful if it was extended to cover any new features
in the backends.

When I adopted the Debian GNADE package, I looked at the changes in the
back-end APIs over time (several releases for each). MySQL often had
non-backward compatible changes, while SQLite and PostgreSQL did not. So
I dropped MySQL, because I did not have time to maintain it.

That said, I have not reviewed any recent changes in these backends to
look for backward compatibility issues.

That is a big problem with hand-generated Ada bindings for libraries. We
really should switch GNADE to use --fdump-ada-spec; that would automate
this check.

-- 
-- Stephe

Re: GNADE still alive?

[Graham Stark]

I've been using GNADE quite a lot recently and it works quite well. If there was to be work put into any Ada DB interface, I'd argue for Gnade on the grounds that you get support for all databases for the price of one. In particular we're unlikely to get support for the big databases like DB/2 and Oracle any other way. 

Areas that look like they need work are support for Unicode (I've never been able to get this to work properly) and support for binary objects.

Here's my attempt at a simpler ODBC interface built on top of Gnade:

http://virtual-worlds-research.com/downloads/mill/

Graham
x

Re: GNADE still alive?

[Maciej Sobczak]

On 28 Maj, 17:49, Graham Stark <graham.st...@virtual-worlds.biz>
wrote:

> In particular we're unlikely to get support for the big databases like DB/2 and Oracle any other way.

You can get support for the big databases and Oracle in particular
from the SOCI-Ada project, now fully incorporated into SOCI:

http://soci.sourceforge.net/doc/languages/ada/index.html

Note: SOCI does not use ODBC to connect to databases like Oracle or
PostgreSQL - their native client APIs are used instead, which is much
easier in terms of configuration. ODBC also exists as a separate
backend, although with the possibility to use more direct methods for
most popular databases, it is much less used.

--
Maciej Sobczak * http://www.msobczak.com * http://www.inspirel.com