aries-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Guillaume Nodet <>
Subject Re: Initial Contribution from IBM
Date Fri, 02 Oct 2009 06:44:48 GMT
Sounds good.
I'll go ahead and import the jndi stuff into trunk, keeping the small
bundles and adding a standalone one as done in blueprint.

On Thu, Oct 1, 2009 at 22:18, Jay D. McHugh <> wrote:
> Hey all,
> I am thinking (with no real familiarity to OSGi) that it may be
> worthwhile to retain the separate bundles in JNDI since the has already
> been broken up.  Also providing (and using within Aries) a larger
> combined bundle would simplify JNDI here.  But, if one wanted to swap
> out the one namespace handler that is separate - then you could cherry-pick.
> My $.005
> Jay
> Valentin Mahrwald wrote:
>> On 1 Oct 2009, at 19:47, Guillaume Nodet wrote:
>>> I personally would like to propose the following structure:
>>>  trunk/jndi
>>>  trunk/jpa
>>> I think they should be migrated to maven, package renamed, etc...
>>> I also think that where it makes sense, having not too fine grained
>>> bundles is better.  I don't really see the point in having two
>>> separate bundles for the jndi stuff, given both parts will have to be
>>> implemented anyway.   Or at least, we should do as it's for blueprint,
>>> and provide a single bundle in addition to smaller ones if others
>>> think it might be useful.
>> The reason for splitting the JNDI bundles is to allow alternative
>> implementations of the osgi:services/ namespace handler without having
>> to redo/repackage the core RFC 142 implementation. There is a use for
>> this in the IBM code. But I can see that from the point alone of the
>> Aries code base it might be neater to stick everything in one bundle.
>>> Also, for, unless there are plans to actually use it
>>> in other bundles, i would also include it in the jpa one or at least
>>> embed this class as a private package if it will be reused.
>> There are two other places the bundle gets used in the
>> IBM code that is not part of the initial contribution, which is why we
>> didn't want to put it into the JPA code. But this argument of course
>> does not really apply anymore in the Aries code base.
>>> I'd be happy to help with the maven stuff if needed.
>>> 2009/10/1 Jeremy Hughes <>:
>>>> Hi, I have just committed r820722 - the initial contribution from IBM
>>>> [1]. There is a readme [2]. The next step is to discuss & move over to
>>>> the trunk. The Software Grant Agreement has been sent to secretary@a.o
>>>> [1]
>>>> [2]
>>>> Cheers,
>>>> Jeremy
>>> --
>>> Cheers,
>>> Guillaume Nodet
>>> ------------------------
>>> Blog:
>>> ------------------------
>>> Open Source SOA

Guillaume Nodet
Open Source SOA

View raw message