* Re: Polyorb abd the DSA Annex
2020-08-03 13:47 ` Shark8
@ 2020-08-03 14:16 ` Dmitry A. Kazakov
2020-08-03 14:27 ` Luke A. Guest
1 sibling, 0 replies; 5+ messages in thread
From: Dmitry A. Kazakov @ 2020-08-03 14:16 UTC (permalink / raw)
On 03/08/2020 15:47, Shark8 wrote:
> On Monday, August 3, 2020 at 5:42:02 AM UTC-6, Luke A. Guest wrote:
>> On 03/08/2020 12:04, tonyg wrote:
>>> I just tried to build the git cloned copy of polyorb (failed on the configure!) with the 2020 community version of GNAT. It said I had no gnat ada compiler. It was on the path and it pointed to the gcc compiler on the path. Are they compatible? Is there still a polyorb "enthusiast" list ?
>>>
>>
>> Distributed annex is being removed from gnat due to "lack of customer interest."
>
> This is pretty sad, and IMO, stupid; the ability to [relatively] easily make distributed applications via DSA is a killer feature and, in conjunction with Ada2020 'parallel' blocks/loops would make for a very attractive system.
>
> IOW, the "lack of customer interest" is an excuse to shoot themselves in the foot.
The reality is a bit more complex.
Distributed Annex is based on RPC.
Ada is largely used in the field applications, embedded, real-time. RPC
are pretty much useless there, as well as in massively parallel
applications.
For service-oriented sluggish applications RPC might be OK, but CORBA is
a blocker there, because static topology/configuration is too rigid for
such applications. (Static topology is less and less tolerated in the
former as well)
P.S. I have an almost ready distributed Annex implementation based on
inter process communication (no network, same box), but I have no
information how to dock it into GNAT.
--
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Polyorb abd the DSA Annex
2020-08-03 13:47 ` Shark8
2020-08-03 14:16 ` Dmitry A. Kazakov
@ 2020-08-03 14:27 ` Luke A. Guest
1 sibling, 0 replies; 5+ messages in thread
From: Luke A. Guest @ 2020-08-03 14:27 UTC (permalink / raw)
On 03/08/2020 14:47, Shark8 wrote:
> On Monday, August 3, 2020 at 5:42:02 AM UTC-6, Luke A. Guest wrote:
>> On 03/08/2020 12:04, tonyg wrote:
>>> I just tried to build the git cloned copy of polyorb (failed on the configure!) with the 2020 community version of GNAT. It said I had no gnat ada compiler. It was on the path and it pointed to the gcc compiler on the path. Are they compatible? Is there still a polyorb "enthusiast" list ?
>>>
>>
>> Distributed annex is being removed from gnat due to "lack of customer interest."
>
> This is pretty sad, and IMO, stupid; the ability to [relatively] easily make distributed applications via DSA is a killer feature and, in conjunction with Ada2020 'parallel' blocks/loops would make for a very attractive system.
>
> IOW, the "lack of customer interest" is an excuse to shoot themselves in the foot.
>
The alternative is keep a local tree with the removal patches revoked.
^ permalink raw reply [flat|nested] 5+ messages in thread