Is CORBA dead for Ada

Is CORBA dead for Ada

[Laurent Pautet]

It seems to me that the CORBA mapping for Ada is almost dead. The last
official document is formal/01-10-42 for CORBA 2.3.

The ada-rtf team does not seem to be very active in its job to update
the mapping (the last closed issues are from 1999).

OIS which seems to lead this task force is very shy in its promotion
of Ada on its web site.

TopGraph'X is still promoting its Ada products. If ORB-River is
compatible with CORBA 2.6, it seems that it does not include new
features from versions greater than 2.3.

Does anyone have fresh news on the OMG activities around the Ada
mapping ?

Is CORBA dead for Ada

[Laurent Pautet]

It seems to me that the CORBA mapping for Ada is almost dead. The last
official document is formal/01-10-42 for CORBA 2.3.

The ada-rtf team does not seem to be very active in its job to update
the mapping (the last closed issues are from 1999).

OIS which seems to lead this task force is very shy in its promotion
of Ada on its web site.

TopGraph'X is still promoting its Ada products. If ORB-River is
compatible with CORBA 2.6, it seems that it does not include new
features from versions greater than 2.3.

Does anyone have fresh news on the OMG activities around the Ada
mapping ?

--
--  Laurent

Re: Is CORBA dead for Ada

[Stephen Leake]

Laurent Pautet <pautet@enst.fr> writes:

> It seems to me that the CORBA mapping for Ada is almost dead. The last
> official document is formal/01-10-42 for CORBA 2.3.

I gather from your tone that there is a later CORBA version?

-- 
-- Stephe

Re: Is CORBA dead for Ada

[Stephane Richard]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 859 bytes --]

indeed, CORBA 2.6 that I read from his message ;-)....but I couldn't say if
there's anything later than that.  CORBA and the Ada DSA are both for
distributed systems so I would much rather see the DSA pushed further. as an
"opposing force" to CORBA :-).  but that's just me.

I couldn't tell you teh discrepencies between 2.3 and 2.6. Wouldn't major
changes in inner structures and such be better suited for Corba 3.0 instead
of 2.6 ?

-- 
St�phane Richard
"Ada World" Webmaster
http://www.adaworld.com


"Stephen Leake" <Stephe.Leake@nasa.gov> wrote in message
news:uwuazlngh.fsf@nasa.gov...
> Laurent Pautet <pautet@enst.fr> writes:
>
> > It seems to me that the CORBA mapping for Ada is almost dead. The last
> > official document is formal/01-10-42 for CORBA 2.3.
>
> I gather from your tone that there is a later CORBA version?
>
> -- 
> -- Stephe

Re: Is CORBA dead for Ada

[Nick Roberts]

Stephane Richard wrote:

> indeed, CORBA 2.6 that I read from his message ;-)....but I couldn't say if
> there's anything later than that.  CORBA and the Ada DSA are both for
> distributed systems so I would much rather see the DSA pushed further. as an
> "opposing force" to CORBA :-).  but that's just me.

So would I, but the DSA as it currently stands is suitable only for
embedded systems (it has no reflection, security, transaction management,
publication, and so on), and the ARG seems to be very reluctant to expand
the DSA's role (or separately extend it).

> I couldn't tell you teh discrepencies between 2.3 and 2.6. Wouldn't major
> changes in inner structures and such be better suited for Corba 3.0 instead
> of 2.6 ?

I wouldn't be surprised if the Ada team are waiting for an indication of
how the mutual type reference problem is going to be solved in the new
revision of Ada before undertaking any major review of the Ada binding.

Nevertheless, I'm inclined to investigate this issue if I can.

-- 
Nick Roberts

Re: Is CORBA dead for Ada

[Stephane Richard]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1574 bytes --]


"Nick Roberts" <nick.roberts@acm.org> wrote in message
news:bn20as$rl85q$1@ID-25716.news.uni-berlin.de...

> So would I, but the DSA as it currently stands is suitable only for
> embedded systems (it has no reflection, security, transaction management,
> publication, and so on), and the ARG seems to be very reluctant to expand
> the DSA's role (or separately extend it).

*** I wonder why.  would be like saying let's not do an X-Windows binding
because there's already a Windows binding done?  Maybe they dont see it as
an oposing force to CORBA.  It should be.  Sure CORBA is good, but so is the
DSA and at the state where it is, because it doesn't have all these things
you mention, it's the perfect oppurtunity to elaborate it in a given
direction :-).
>
> I wouldn't be surprised if the Ada team are waiting for an indication of
> how the mutual type reference problem is going to be solved in the new
> revision of Ada before undertaking any major review of the Ada binding.
>
*** For this I'm not sure, I'm not part of the Ada team ;-)....yet...but the
problem is there and if not anything else should be resolved.  It's a very
concrete problem and should be fixable somehow :-).

> Nevertheless, I'm inclined to investigate this issue if I can.
>
*** So am I and I will too ;-).  perhaps we can exchange notes?

By the way, it's good to see a post from you Nick, been a while since I've
seen you posting (maybe I just dont read the newsgroups often enough ;-).

> -- 
> Nick Roberts
>
>


-- 
St�phane Richard
"Ada World" Webmaster
http://www.adaworld.com

Re: Is CORBA dead for Ada

[Laurent Pautet]

Nick Roberts <nick.roberts@acm.org> wrote:
> 
> So would I, but the DSA as it currently stands is suitable only for
> embedded systems (it has no reflection, security, transaction management,
> publication, and so on), and the ARG seems to be very reluctant to expand
> the DSA's role (or separately extend it).

Indeed, there is no attempt to extend DSA. The current idea is to be
less normative in order to allow DSA implementations based on CORBA.
However, my question is about CORBA and not DSA :)

> I wouldn't be surprised if the Ada team are waiting for an indication of
> how the mutual type reference problem is going to be solved in the new
> revision of Ada before undertaking any major review of the Ada binding.

This would mean that it is not possible to use CORBA > 2.3 until Ada0Y
is complete. I do not believe that.

The last CORBA revision is CORBA 3.0. The last Ada mapping is for
CORBA 2.3. In between, many important features have been introduced in
CORBA and many issues have been raised in the (old) mapping (mostly by
O. Kellogg). But the Ada mapping revision seems to be frozen for
years. My belief is that there is no real CORBA market for Ada so the
major tool vendors focus on other markets like the Java and C++ ones.

Re: Is CORBA dead for Ada

[Jean-Claude Mahieux]

> TopGraph'X is still promoting its Ada products. If ORB-River is
> compatible with CORBA 2.6, it seems that it does not include new
> features from versions greater than 2.3.
Laurent , I entirely disagree with that (see our web site).

OrbRiver/Ada supports :
Interoperable Naming Service
Corbaloc/Corbaname URLs
CORBA Messaging
RT CORBA
MIOP (with a reliable implementation to come soon)
Notification Service (entirely implemented in Ada95)

OrbRiver/C++ is approximately at the same level and OrbRiver/Java a little
bit behind


--
Jean-Claude MAHIEUX
Top Graph'X Sales Manager
Parc de la Fontaine de Jouvence
4 rue Angiboust
91460 Marcoussis - FRANCE
Tel : +33 1 69 01 53 54
Fax : +33 1 69 01 53 55
Email : jeanclaude.mahieux@topgraphx.com
URL : http://www.topgraphx.com
German Representative : orbriver-de@topgraphx.com
US Representative : sroliver@topgraphx.com

Re: Is CORBA dead for Ada

[Laurent Pautet]

>OrbRiver/Ada supports :
>Interoperable Naming Service
>Corbaloc/Corbaname URLs
[...]
>MIOP (with a reliable implementation to come soon)
>Notification Service (entirely implemented in Ada95)

Fine, Jean-Claude ! We have some of these with polyorb as well for
instance the two last ones.

But my question was about the status of the Ada mapping. CORBA/Ada is
almost unusable for a user without the mapping. Has it been updated ?

-- 
-- Laurent

Re: Is CORBA dead for Ada

[Volkert]

> My belief is that there is no real CORBA market for Ada so the
> major tool vendors focus on other markets like the Java and C++ ones.
i hope some ada (compiler, orb) vendors are clever 
enough to see the impact of middleware technologies 
for the future ada market. for the most companies using
ada now it is really important to get/keep their 
systems accessible from other sw systems created with 
modern/mainstream technologies. otherwise it will be 
harder and harder to promote ada for any future developments 
and any future investments  ...

believe me, if there is no visible progress with 
the corba standard the things will getting harder 
for some of us ... it is really counterproductive 

Volkert
we have a quite large ada system running here ...

Re: Is CORBA dead for Ada

[Ed Falis]

On 22 Oct 2003 11:20:45 -0700, Volkert <volkert@nivoba.de> wrote:

>> My belief is that there is no real CORBA market for Ada so the
>> major tool vendors focus on other markets like the Java and C++ ones.
> i hope some ada (compiler, orb) vendors are clever
> enough to see the impact of middleware technologies
> for the future ada market. for the most companies using
> ada now it is really important to get/keep their
> systems accessible from other sw systems created with
> modern/mainstream technologies. otherwise it will be
> harder and harder to promote ada for any future developments
> and any future investments  ...
>

Volkert,

You may want to check out libre.act-europe.fr/polyorb/

(note that the website appears to be down as I post this).

- Ed

Re: Is CORBA dead for Ada

[Xenos]

"Volkert" <volkert@nivoba.de> wrote in message
news:d37844cb.0310221020.2832f325@posting.google.com...
> > My belief is that there is no real CORBA market for Ada so the
> > major tool vendors focus on other markets like the Java and C++ ones.
> i hope some ada (compiler, orb) vendors are clever
> enough to see the impact of middleware technologies
> for the future ada market. for the most companies using
> ada now it is really important to get/keep their
> systems accessible from other sw systems created with
> modern/mainstream technologies. otherwise it will be
> harder and harder to promote ada for any future developments
> and any future investments  ...
>
> believe me, if there is no visible progress with
> the corba standard the things will getting harder
> for some of us ... it is really counterproductive
>
> Volkert
> we have a quite large ada system running here ...

I'm using CORBA for Ada right now....


DrX.

Re: Is CORBA dead for Ada

[Volkert]

"Xenos" <dont.spam.me@spamhate.com> wrote in message 

> > believe me, if there is no visible progress with
> > the corba standard the things will getting harder
> > for some of us ... it is really counterproductive
> >
> > Volkert
> > we have a quite large ada system running here ...
> 
> I'm using CORBA for Ada right now....
We are starting to evaluate one ada orb right now.

Volkert

Re: Is CORBA dead for Ada

[Volkert]

Ed Falis <falis@verizon.net> wrote in message news:<oprxgi33mn8wdn3j@news.verizon.net>...
> On 22 Oct 2003 11:20:45 -0700, Volkert <volkert@nivoba.de> wrote:
> 
> >> My belief is that there is no real CORBA market for Ada so the
> >> major tool vendors focus on other markets like the Java and C++ ones.
> > i hope some ada (compiler, orb) vendors are clever
> > enough to see the impact of middleware technologies
> > for the future ada market. for the most companies using
> > ada now it is really important to get/keep their
> > systems accessible from other sw systems created with
> > modern/mainstream technologies. otherwise it will be
> > harder and harder to promote ada for any future developments
> > and any future investments  ...
> >
> 
> Volkert,
> 
> You may want to check out libre.act-europe.fr/polyorb/
allready done, but i had problems with connecting java clients 
(SUN JDK 1.4.2/SUN ORB) to Ada servers. As long as polyorb 
is not a >>visible<< supported product of ACT, it is not 
interesting for us ...

Volkert

RE: Is CORBA dead for Ada

[Lionel.DRAGHI]

| -----Message d'origine-----
| De: volkert@nivoba.de [mailto:volkert@nivoba.de]
...
| believe me, if there is no visible progress with 
| the corba standard the things will getting harder 
| for some of us ... it is really counterproductive 
| 
I agree with you. 

Others post here answered about actually existing Ada ORB, and that's OK.
But original laurent's demand was about standart. If Ada ORB do not stick to
the latest CORBA version, it will cause interoperability problems, and will
be also interpreted as a further Ada influence loss.

As Volkert said, for a company involved in big systems, this may lead to Ada
drop (and exceptionnaly with an andisputable reason).

Maybe need we just that ORB vendors set our minds at rest by reaffirming
their commitment in Ada for CORBA, as do by Jean-Claude in this thread.

-- 
Lionel Draghi 

Re: Is CORBA dead for Ada

[Laurent Pautet]

Lionel.DRAGHI@fr.thalesgroup.com wrote:
>But original laurent's demand was about standart. If Ada ORB do not stick to
>the latest CORBA version, it will cause interoperability problems, and will
>be also interpreted as a further Ada influence loss.

Thanks Lionel.

Indeed my question is the current status of the mapping.

I must apologize for having started this troll but at least one CORBA
tool vendor is now on-line :) We had the advertising part about their
technology, we are now waiting for an answer about the mapping :))

-- 
-- Laurent

Re: Is CORBA dead for Ada

[Laurent Pautet]

Lionel.DRAGHI@fr.thalesgroup.com wrote:
>But original laurent's demand was about standart. If Ada ORB do not stick to
>the latest CORBA version, it will cause interoperability problems, and will
>be also interpreted as a further Ada influence loss.

Thanks Lionel.

Indeed my question is the current status of the mapping.

I must apologize for having started this troll but at least one CORBA
tool vendor is now on-line :) We had the advertising part about their
technology, we are now waiting for an answer about the mapping :))

-- 
-- Laurent

Re: Is CORBA dead for Ada

[Volkert]

pautet@antigone.enst.fr (Laurent Pautet) wrote in message news:<slrnbpbfbu.cbv.pautet@antigone.enst.fr>...
> >OrbRiver/Ada supports :
> >Interoperable Naming Service
> >Corbaloc/Corbaname URLs
>  [...]
> >MIOP (with a reliable implementation to come soon)
> >Notification Service (entirely implemented in Ada95)
> 
> Fine, Jean-Claude ! We have some of these with polyorb as well for
> instance the two last ones.
> 
> But my question was about the status of the Ada mapping. CORBA/Ada is
> almost unusable for a user without the mapping. Has it been updated ?
Strange ... Mr Pautet you are one persons behind polyorb ... polyorb leads 
to ACT. If i can get support from ACT for polyorb so why does ACT (as an
orb vendor) not participate in the maintanance of corba standard? 

Volkert

Re: Is CORBA dead for Ada

[Victor Giddings]

Laurent Pautet <pautet@enst.fr> wrote in news:bn145s$1kbk$1@avanie.enst.fr:

> It seems to me that the CORBA mapping for Ada is almost dead. The last
> official document is formal/01-10-42 for CORBA 2.3.
> 
> The ada-rtf team does not seem to be very active in its job to update
> the mapping (the last closed issues are from 1999).
> 
> OIS which seems to lead this task force is very shy in its promotion
> of Ada on its web site.

I will feed this back to our marketing team. We are not shy in our 
promotion of Ada. In the last OMG Real-Time Workshop in response to a 
comment, I was accused of being "an unrepentent Ada programmer". At which 
point, my boss exclaimed "damned right we are"!

> TopGraph'X is still promoting its Ada products. If ORB-River is
> compatible with CORBA 2.6, it seems that it does not include new
> features from versions greater than 2.3.
> 
> Does anyone have fresh news on the OMG activities around the Ada
> mapping ?
> 
> --
> --  Laurent

Dirk Craeynest was kind enough to repost my responses to a similar thread 
on comp.object.corba and inform me of this thread. I would like to expand 
on some of the earlier responses. 

There seems to be much apprehension and more than a little misunderstanding 
associated with CORBA "versions". CORBA is not a monolithic specification 
but a collection of adopted specifications that may or may not be 
consolidated into individual documents. The individual documents have 
associated Revision Task Forces (RTFs) and individual life cycles. The 
references such as CORBA 2.6 are more properly references to a particular 
version of the "CORBA Core" specification, a particular document that 
specifies the language-independent requirements of what an ORB product must 
implement. Changes to the CORBA Core specification may or may not require 
changes to the language mapping specifications, depending on whether there 
is a significant change to the IDL language. Therefore, a lot of revision 
of the CORBA Core would have been addressed by changing the last digit in 
the statement that the mapping was "aligned to CORBA version 2.x". 

There is understandable confusion about this that has been partly caused by 
the OMG itself. First of all, the CORBA Core document contains a lot of 
things, e.g. CORBA/COM Interworking, that don't have to be implemented by 
an ORB product. The OMG staff have also issued press releases that claimed 
what the contents of CORBA x.x would contain. In general, this is a problem 
that needs to be fixed. In addition to being the chair of the Ada RTF, I am 
chair of a group in the OMG called the Product Specification Definition 
(psdef) subcommittee that is trying to straighten out publication 
organization and coordination of versioning. I urge you to participate. 

The bottom line is that the fact that the current Ada Language Mapping 
specification is "aligned to CORBA 2.3" means very little. As I stated in 
the comp.object.corba post, the OMG IDL language is fairly stable, so few 
changes in the language mapping are needed. Most of the features added in 
the CORBA Core 2.4 and later versions are specified in a language 
independent manner and have not affected the language mapping. So, as Jean-
Claude Mahieux was able to report, there has been no hindrance to advancing 
Ada ORB implementations. 

That being said, the other responders are correct in that there has not 
been an active Ada Mapping RTF is almost 3 years. My only excuse for this 
is that we have been busy with other OMG specifications like Real-Time 
CORBA (1.0 and 2.0), Fault Tolerant CORBA, Data Distribution, etc., etc.. 
(I'm not sure what excuse my competitors have ;) Nevertheless, there is 
currently an active RTF attempting to deal with the backlog. Anyone may 
participate in this activity by joining the email group (ada-rtf-
request@omg.org) Non-OMG members may need to contact me to be added to the 
list. Voting membership (one per organization) in the RTF requires a 
minimal level of membership in the OMG but, in practice, most voting is 
pro-forma after a consensus has been worked out among us. Let me know if 
you are interested in formal membership, so this can be placed on the PTC 
agenda. 
 
-- 
Victor Giddings		mailto:victor.giddings@ois.com
Senior Scientist	+1 703 295 6500
Objective Interface Systems	Fax: +1 703 295 6501

Re: Is CORBA dead for Ada

[Laurent Pautet]

In article <d37844cb.0310231055.7f372bd6@posting.google.com>, Volkert wrote:
>Strange ... Mr Pautet you are one persons behind polyorb ... polyorb leads 
>to ACT. If i can get support from ACT for polyorb so why does ACT (as an
>orb vendor) not participate in the maintanance of corba standard? 

As far as I am concerned, I don't see the maintenance of the Ada
mapping as a research interest (anyway, I would need a funding to
participate to it). But in the context of schizophrenic middleware, I
am interested in an up-to-date mapping.

Concerning ACT, ask them.

-- 
-- Laurent