You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@aries.apache.org by da...@apache.org on 2010/02/22 12:00:41 UTC

Async communications experimental component

Hi everyone,

In the OSGi Alliance we're currently working on defining 'Asynchronous
Distributed OSGi' in various forms and what that might look like.
Currently the work is divided up into three areas:
* Bringing JMS to OSGi
* Message Driven Components
* Asynchronous Services
We're really at the requirements gathering stage right now...

In the mean time I'm looking for a place where we can do some
experimentation and collaboration in the area in order to get a better
understanding of these problems and potential solutions. I would love
to start a component in Aries for this purpose.
The three areas identified above are more or less orthogonal, but I
can see many cases where they are combined, so at least initially I
think a single experimentation area for them would be good.
Then, when the dust settles a bit we might extract a number of
independent components... Anyway, that's at least how I think we could
do it, but I would also have no big issues with three separate
experimentation areas (or maybe three submodules of an async top-level
module would be a middle ground).

Anyway, if people think this is a good idea I can start putting some
of the structure in the build system over the coming few days...

Best regards,

David

Re: Async communications experimental component

Posted by da...@apache.org.
Good point.
There is another experimental component that I started a little while
ago around JRE SPI support
(http://svn.apache.org/repos/asf/incubator/aries/trunk/spi-fly) that I
also don't think is ready for a release.

While I would have no problem waiting until the release is out of the
way, I think that it would be better to make it possible to mark
components as experimental and as such skip them from the release.

It all depends on how the release is pulled together, I guess. In
other projects I have seen a pom.xml specifically for the release that
pulls in the desired components. So in that case by default a
component is not part of a release, unless its listed in that release
pom. Sure, a 'mvn deploy' will publish everything, but the actual
release download will only contain the selected components.

Best regards,

David

On 22 February 2010 11:52, Alasdair Nottingham <no...@apache.org> wrote:
> Hi,
>
> Given that we are trying to do a release at the moment this makes me
> wonder how we should differentiate things which are experimental and
> wouldn't be part of the release, and things which are to be part of
> the release.
>
> Thoughts?
> Alasdair
>
> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>> Hi everyone,
>>
>> In the OSGi Alliance we're currently working on defining 'Asynchronous
>> Distributed OSGi' in various forms and what that might look like.
>> Currently the work is divided up into three areas:
>> * Bringing JMS to OSGi
>> * Message Driven Components
>> * Asynchronous Services
>> We're really at the requirements gathering stage right now...
>>
>> In the mean time I'm looking for a place where we can do some
>> experimentation and collaboration in the area in order to get a better
>> understanding of these problems and potential solutions. I would love
>> to start a component in Aries for this purpose.
>> The three areas identified above are more or less orthogonal, but I
>> can see many cases where they are combined, so at least initially I
>> think a single experimentation area for them would be good.
>> Then, when the dust settles a bit we might extract a number of
>> independent components... Anyway, that's at least how I think we could
>> do it, but I would also have no big issues with three separate
>> experimentation areas (or maybe three submodules of an async top-level
>> module would be a middle ground).
>>
>> Anyway, if people think this is a good idea I can start putting some
>> of the structure in the build system over the coming few days...
>>
>> Best regards,
>>
>> David
>>
>
>
>
> --
> Alasdair Nottingham
> not@apache.org
>

RE: Async communications experimental component

Posted by Timothy Ward <ti...@apache.org>.
I'm +1 for the new async component, but agree that we should have some mechanism to differentiate experimental bundles and release bundles.

Perhaps an "EXPERIMENTAL" flag on the pom version?

Regards,

Tim

> Date: Mon, 22 Feb 2010 11:52:21 +0000
> Subject: Re: Async communications experimental component
> From: not@apache.org
> To: aries-dev@incubator.apache.org; davidb@apache.org
> 
> Hi,
> 
> Given that we are trying to do a release at the moment this makes me
> wonder how we should differentiate things which are experimental and
> wouldn't be part of the release, and things which are to be part of
> the release.
> 
> Thoughts?
> Alasdair
> 
> On 22 February 2010 11:00,  <da...@apache.org> wrote:
> > Hi everyone,
> >
> > In the OSGi Alliance we're currently working on defining 'Asynchronous
> > Distributed OSGi' in various forms and what that might look like.
> > Currently the work is divided up into three areas:
> > * Bringing JMS to OSGi
> > * Message Driven Components
> > * Asynchronous Services
> > We're really at the requirements gathering stage right now...
> >
> > In the mean time I'm looking for a place where we can do some
> > experimentation and collaboration in the area in order to get a better
> > understanding of these problems and potential solutions. I would love
> > to start a component in Aries for this purpose.
> > The three areas identified above are more or less orthogonal, but I
> > can see many cases where they are combined, so at least initially I
> > think a single experimentation area for them would be good.
> > Then, when the dust settles a bit we might extract a number of
> > independent components... Anyway, that's at least how I think we could
> > do it, but I would also have no big issues with three separate
> > experimentation areas (or maybe three submodules of an async top-level
> > module would be a middle ground).
> >
> > Anyway, if people think this is a good idea I can start putting some
> > of the structure in the build system over the coming few days...
> >
> > Best regards,
> >
> > David
> >
> 
> 
> 
> -- 
> Alasdair Nottingham
> not@apache.org
 		 	   		  
_________________________________________________________________
Do you have a story that started on Hotmail? Tell us now
http://clk.atdmt.com/UKM/go/195013117/direct/01/

Re: Async communications experimental component

Posted by David Bosschaert <da...@gmail.com>.
Seeing that David Jecks has untertaken the buildsystem refactor
(excellent work, David!), I don't see the need for the async component
to be in a sandbox any more. There are a couple of reasons why I would
really prefer trunk (after having thought about it for a little
while):
- being apart from the rest of Aries, the sandbox is less visible
- the sandbox doesn't have Hudson builds, so no continuous integration
- since one of the goals of Aries is to collaborate on components that
might turn into OSGi Specifications I think that doing this
collaboration in a clearly visible place, like trunk is best.

So... if there are no objections to this I'll add an 'async' module to
trunk over the coming few days.

Best regards,

David

On 24 February 2010 11:48, David Bosschaert <da...@gmail.com> wrote:
> Ok, I guess this thread got obsoleted by the Aries Release thread a little :)
>
> Once the buildsystem refactor is done as suggested on that thread,
> there shouldn't really be a need to do experimental components in the
> sandbox, right? Or do people feel strongly that they should? I guess I
> would prefer the main trunk as they would then also be included in the
> hudson builds etc...
>
> Thanks,
>
> David
>
> On 23 February 2010 17:49, Joe Bohn <jo...@gmail.com> wrote:
>>
>> me too
>>
>> Joe
>>
>> Alasdair Nottingham wrote:
>>>
>>> Makes sense to me.
>>>
>>> Alasdair
>>>
>>> On 23 Feb 2010, at 15:04, David Bosschaert <da...@gmail.com>
>>> wrote:
>>>
>>>> Sandbox sounds fine to me.
>>>> Should I also move the SPI Fly component to the sandbox? Since that
>>>> one's experimental too...
>>>>
>>>> Cheers,
>>>>
>>>> David
>>>>
>>>> On 23 February 2010 02:58, Joe Bohn <jo...@gmail.com> wrote:
>>>>>
>>>>> Guillaume Nodet wrote:
>>>>>>
>>>>>> Yeah, I agree.  What about put it in a sandbox area for now if you
>>>>>> want to experiment and move it to trunk after the release?
>>>>>
>>>>> +1 - the sandbox seems to work well for other apache projects.
>>>>> Alternatively
>>>>> I guess we could continue to use the contrib location for this purpose.
>>>>>
>>>>> Joe
>>>>>
>>>>>
>>>>>>
>>>>>> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org>
>>>>>> wrote:
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> Given that we are trying to do a release at the moment this makes me
>>>>>>> wonder how we should differentiate things which are experimental and
>>>>>>> wouldn't be part of the release, and things which are to be part of
>>>>>>> the release.
>>>>>>>
>>>>>>> Thoughts?
>>>>>>> Alasdair
>>>>>>>
>>>>>>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>>>>>>>
>>>>>>>> Hi everyone,
>>>>>>>>
>>>>>>>> In the OSGi Alliance we're currently working on defining
>>>>>>>> 'Asynchronous
>>>>>>>> Distributed OSGi' in various forms and what that might look like.
>>>>>>>> Currently the work is divided up into three areas:
>>>>>>>> * Bringing JMS to OSGi
>>>>>>>> * Message Driven Components
>>>>>>>> * Asynchronous Services
>>>>>>>> We're really at the requirements gathering stage right now...
>>>>>>>>
>>>>>>>> In the mean time I'm looking for a place where we can do some
>>>>>>>> experimentation and collaboration in the area in order to get a
>>>>>>>> better
>>>>>>>> understanding of these problems and potential solutions. I would love
>>>>>>>> to start a component in Aries for this purpose.
>>>>>>>> The three areas identified above are more or less orthogonal, but I
>>>>>>>> can see many cases where they are combined, so at least initially I
>>>>>>>> think a single experimentation area for them would be good.
>>>>>>>> Then, when the dust settles a bit we might extract a number of
>>>>>>>> independent components... Anyway, that's at least how I think we
>>>>>>>> could
>>>>>>>> do it, but I would also have no big issues with three separate
>>>>>>>> experimentation areas (or maybe three submodules of an async
>>>>>>>> top-level
>>>>>>>> module would be a middle ground).
>>>>>>>>
>>>>>>>> Anyway, if people think this is a good idea I can start putting some
>>>>>>>> of the structure in the build system over the coming few days...
>>>>>>>>
>>>>>>>> Best regards,
>>>>>>>>
>>>>>>>> David
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Alasdair Nottingham
>>>>>>> not@apache.org
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Joe
>>>>>
>>>
>>
>>
>> --
>> Joe
>>
>

Re: Async communications experimental component

Posted by David Bosschaert <da...@gmail.com>.
Ok, I guess this thread got obsoleted by the Aries Release thread a little :)

Once the buildsystem refactor is done as suggested on that thread,
there shouldn't really be a need to do experimental components in the
sandbox, right? Or do people feel strongly that they should? I guess I
would prefer the main trunk as they would then also be included in the
hudson builds etc...

Thanks,

David

On 23 February 2010 17:49, Joe Bohn <jo...@gmail.com> wrote:
>
> me too
>
> Joe
>
> Alasdair Nottingham wrote:
>>
>> Makes sense to me.
>>
>> Alasdair
>>
>> On 23 Feb 2010, at 15:04, David Bosschaert <da...@gmail.com>
>> wrote:
>>
>>> Sandbox sounds fine to me.
>>> Should I also move the SPI Fly component to the sandbox? Since that
>>> one's experimental too...
>>>
>>> Cheers,
>>>
>>> David
>>>
>>> On 23 February 2010 02:58, Joe Bohn <jo...@gmail.com> wrote:
>>>>
>>>> Guillaume Nodet wrote:
>>>>>
>>>>> Yeah, I agree.  What about put it in a sandbox area for now if you
>>>>> want to experiment and move it to trunk after the release?
>>>>
>>>> +1 - the sandbox seems to work well for other apache projects.
>>>> Alternatively
>>>> I guess we could continue to use the contrib location for this purpose.
>>>>
>>>> Joe
>>>>
>>>>
>>>>>
>>>>> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org>
>>>>> wrote:
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Given that we are trying to do a release at the moment this makes me
>>>>>> wonder how we should differentiate things which are experimental and
>>>>>> wouldn't be part of the release, and things which are to be part of
>>>>>> the release.
>>>>>>
>>>>>> Thoughts?
>>>>>> Alasdair
>>>>>>
>>>>>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>>>>>>
>>>>>>> Hi everyone,
>>>>>>>
>>>>>>> In the OSGi Alliance we're currently working on defining
>>>>>>> 'Asynchronous
>>>>>>> Distributed OSGi' in various forms and what that might look like.
>>>>>>> Currently the work is divided up into three areas:
>>>>>>> * Bringing JMS to OSGi
>>>>>>> * Message Driven Components
>>>>>>> * Asynchronous Services
>>>>>>> We're really at the requirements gathering stage right now...
>>>>>>>
>>>>>>> In the mean time I'm looking for a place where we can do some
>>>>>>> experimentation and collaboration in the area in order to get a
>>>>>>> better
>>>>>>> understanding of these problems and potential solutions. I would love
>>>>>>> to start a component in Aries for this purpose.
>>>>>>> The three areas identified above are more or less orthogonal, but I
>>>>>>> can see many cases where they are combined, so at least initially I
>>>>>>> think a single experimentation area for them would be good.
>>>>>>> Then, when the dust settles a bit we might extract a number of
>>>>>>> independent components... Anyway, that's at least how I think we
>>>>>>> could
>>>>>>> do it, but I would also have no big issues with three separate
>>>>>>> experimentation areas (or maybe three submodules of an async
>>>>>>> top-level
>>>>>>> module would be a middle ground).
>>>>>>>
>>>>>>> Anyway, if people think this is a good idea I can start putting some
>>>>>>> of the structure in the build system over the coming few days...
>>>>>>>
>>>>>>> Best regards,
>>>>>>>
>>>>>>> David
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Alasdair Nottingham
>>>>>> not@apache.org
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Joe
>>>>
>>
>
>
> --
> Joe
>

Re: Async communications experimental component

Posted by Joe Bohn <jo...@gmail.com>.
me too

Joe

Alasdair Nottingham wrote:
> Makes sense to me.
> 
> Alasdair
> 
> On 23 Feb 2010, at 15:04, David Bosschaert <da...@gmail.com> 
> wrote:
> 
>> Sandbox sounds fine to me.
>> Should I also move the SPI Fly component to the sandbox? Since that
>> one's experimental too...
>>
>> Cheers,
>>
>> David
>>
>> On 23 February 2010 02:58, Joe Bohn <jo...@gmail.com> wrote:
>>>
>>> Guillaume Nodet wrote:
>>>>
>>>> Yeah, I agree.  What about put it in a sandbox area for now if you
>>>> want to experiment and move it to trunk after the release?
>>>
>>> +1 - the sandbox seems to work well for other apache projects. 
>>> Alternatively
>>> I guess we could continue to use the contrib location for this purpose.
>>>
>>> Joe
>>>
>>>
>>>>
>>>> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org> 
>>>> wrote:
>>>>>
>>>>> Hi,
>>>>>
>>>>> Given that we are trying to do a release at the moment this makes me
>>>>> wonder how we should differentiate things which are experimental and
>>>>> wouldn't be part of the release, and things which are to be part of
>>>>> the release.
>>>>>
>>>>> Thoughts?
>>>>> Alasdair
>>>>>
>>>>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>>>>>
>>>>>> Hi everyone,
>>>>>>
>>>>>> In the OSGi Alliance we're currently working on defining 
>>>>>> 'Asynchronous
>>>>>> Distributed OSGi' in various forms and what that might look like.
>>>>>> Currently the work is divided up into three areas:
>>>>>> * Bringing JMS to OSGi
>>>>>> * Message Driven Components
>>>>>> * Asynchronous Services
>>>>>> We're really at the requirements gathering stage right now...
>>>>>>
>>>>>> In the mean time I'm looking for a place where we can do some
>>>>>> experimentation and collaboration in the area in order to get a 
>>>>>> better
>>>>>> understanding of these problems and potential solutions. I would love
>>>>>> to start a component in Aries for this purpose.
>>>>>> The three areas identified above are more or less orthogonal, but I
>>>>>> can see many cases where they are combined, so at least initially I
>>>>>> think a single experimentation area for them would be good.
>>>>>> Then, when the dust settles a bit we might extract a number of
>>>>>> independent components... Anyway, that's at least how I think we 
>>>>>> could
>>>>>> do it, but I would also have no big issues with three separate
>>>>>> experimentation areas (or maybe three submodules of an async 
>>>>>> top-level
>>>>>> module would be a middle ground).
>>>>>>
>>>>>> Anyway, if people think this is a good idea I can start putting some
>>>>>> of the structure in the build system over the coming few days...
>>>>>>
>>>>>> Best regards,
>>>>>>
>>>>>> David
>>>>>>
>>>>>
>>>>>
>>>>> -- 
>>>>> Alasdair Nottingham
>>>>> not@apache.org
>>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> -- 
>>> Joe
>>>
> 


-- 
Joe

Re: Async communications experimental component

Posted by Alasdair Nottingham <no...@apache.org>.
Makes sense to me.

Alasdair

On 23 Feb 2010, at 15:04, David Bosschaert  
<da...@gmail.com> wrote:

> Sandbox sounds fine to me.
> Should I also move the SPI Fly component to the sandbox? Since that
> one's experimental too...
>
> Cheers,
>
> David
>
> On 23 February 2010 02:58, Joe Bohn <jo...@gmail.com> wrote:
>>
>> Guillaume Nodet wrote:
>>>
>>> Yeah, I agree.  What about put it in a sandbox area for now if you
>>> want to experiment and move it to trunk after the release?
>>
>> +1 - the sandbox seems to work well for other apache projects.  
>> Alternatively
>> I guess we could continue to use the contrib location for this  
>> purpose.
>>
>> Joe
>>
>>
>>>
>>> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham  
>>> <no...@apache.org> wrote:
>>>>
>>>> Hi,
>>>>
>>>> Given that we are trying to do a release at the moment this makes  
>>>> me
>>>> wonder how we should differentiate things which are experimental  
>>>> and
>>>> wouldn't be part of the release, and things which are to be part of
>>>> the release.
>>>>
>>>> Thoughts?
>>>> Alasdair
>>>>
>>>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>>>>
>>>>> Hi everyone,
>>>>>
>>>>> In the OSGi Alliance we're currently working on defining  
>>>>> 'Asynchronous
>>>>> Distributed OSGi' in various forms and what that might look like.
>>>>> Currently the work is divided up into three areas:
>>>>> * Bringing JMS to OSGi
>>>>> * Message Driven Components
>>>>> * Asynchronous Services
>>>>> We're really at the requirements gathering stage right now...
>>>>>
>>>>> In the mean time I'm looking for a place where we can do some
>>>>> experimentation and collaboration in the area in order to get a  
>>>>> better
>>>>> understanding of these problems and potential solutions. I would  
>>>>> love
>>>>> to start a component in Aries for this purpose.
>>>>> The three areas identified above are more or less orthogonal,  
>>>>> but I
>>>>> can see many cases where they are combined, so at least  
>>>>> initially I
>>>>> think a single experimentation area for them would be good.
>>>>> Then, when the dust settles a bit we might extract a number of
>>>>> independent components... Anyway, that's at least how I think we  
>>>>> could
>>>>> do it, but I would also have no big issues with three separate
>>>>> experimentation areas (or maybe three submodules of an async top- 
>>>>> level
>>>>> module would be a middle ground).
>>>>>
>>>>> Anyway, if people think this is a good idea I can start putting  
>>>>> some
>>>>> of the structure in the build system over the coming few days...
>>>>>
>>>>> Best regards,
>>>>>
>>>>> David
>>>>>
>>>>
>>>>
>>>> --
>>>> Alasdair Nottingham
>>>> not@apache.org
>>>>
>>>
>>>
>>>
>>
>>
>> --
>> Joe
>>

Re: Async communications experimental component

Posted by David Bosschaert <da...@gmail.com>.
Sandbox sounds fine to me.
Should I also move the SPI Fly component to the sandbox? Since that
one's experimental too...

Cheers,

David

On 23 February 2010 02:58, Joe Bohn <jo...@gmail.com> wrote:
>
> Guillaume Nodet wrote:
>>
>> Yeah, I agree.  What about put it in a sandbox area for now if you
>> want to experiment and move it to trunk after the release?
>
> +1 - the sandbox seems to work well for other apache projects. Alternatively
> I guess we could continue to use the contrib location for this purpose.
>
> Joe
>
>
>>
>> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org> wrote:
>>>
>>> Hi,
>>>
>>> Given that we are trying to do a release at the moment this makes me
>>> wonder how we should differentiate things which are experimental and
>>> wouldn't be part of the release, and things which are to be part of
>>> the release.
>>>
>>> Thoughts?
>>> Alasdair
>>>
>>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>>>
>>>> Hi everyone,
>>>>
>>>> In the OSGi Alliance we're currently working on defining 'Asynchronous
>>>> Distributed OSGi' in various forms and what that might look like.
>>>> Currently the work is divided up into three areas:
>>>> * Bringing JMS to OSGi
>>>> * Message Driven Components
>>>> * Asynchronous Services
>>>> We're really at the requirements gathering stage right now...
>>>>
>>>> In the mean time I'm looking for a place where we can do some
>>>> experimentation and collaboration in the area in order to get a better
>>>> understanding of these problems and potential solutions. I would love
>>>> to start a component in Aries for this purpose.
>>>> The three areas identified above are more or less orthogonal, but I
>>>> can see many cases where they are combined, so at least initially I
>>>> think a single experimentation area for them would be good.
>>>> Then, when the dust settles a bit we might extract a number of
>>>> independent components... Anyway, that's at least how I think we could
>>>> do it, but I would also have no big issues with three separate
>>>> experimentation areas (or maybe three submodules of an async top-level
>>>> module would be a middle ground).
>>>>
>>>> Anyway, if people think this is a good idea I can start putting some
>>>> of the structure in the build system over the coming few days...
>>>>
>>>> Best regards,
>>>>
>>>> David
>>>>
>>>
>>>
>>> --
>>> Alasdair Nottingham
>>> not@apache.org
>>>
>>
>>
>>
>
>
> --
> Joe
>

Re: Async communications experimental component

Posted by Joe Bohn <jo...@gmail.com>.
Guillaume Nodet wrote:
> Yeah, I agree.  What about put it in a sandbox area for now if you
> want to experiment and move it to trunk after the release?

+1 - the sandbox seems to work well for other apache projects. 
Alternatively I guess we could continue to use the contrib location for 
this purpose.

Joe


> 
> On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org> wrote:
>> Hi,
>>
>> Given that we are trying to do a release at the moment this makes me
>> wonder how we should differentiate things which are experimental and
>> wouldn't be part of the release, and things which are to be part of
>> the release.
>>
>> Thoughts?
>> Alasdair
>>
>> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>>> Hi everyone,
>>>
>>> In the OSGi Alliance we're currently working on defining 'Asynchronous
>>> Distributed OSGi' in various forms and what that might look like.
>>> Currently the work is divided up into three areas:
>>> * Bringing JMS to OSGi
>>> * Message Driven Components
>>> * Asynchronous Services
>>> We're really at the requirements gathering stage right now...
>>>
>>> In the mean time I'm looking for a place where we can do some
>>> experimentation and collaboration in the area in order to get a better
>>> understanding of these problems and potential solutions. I would love
>>> to start a component in Aries for this purpose.
>>> The three areas identified above are more or less orthogonal, but I
>>> can see many cases where they are combined, so at least initially I
>>> think a single experimentation area for them would be good.
>>> Then, when the dust settles a bit we might extract a number of
>>> independent components... Anyway, that's at least how I think we could
>>> do it, but I would also have no big issues with three separate
>>> experimentation areas (or maybe three submodules of an async top-level
>>> module would be a middle ground).
>>>
>>> Anyway, if people think this is a good idea I can start putting some
>>> of the structure in the build system over the coming few days...
>>>
>>> Best regards,
>>>
>>> David
>>>
>>
>>
>> --
>> Alasdair Nottingham
>> not@apache.org
>>
> 
> 
> 


-- 
Joe

Re: Async communications experimental component

Posted by Guillaume Nodet <gn...@gmail.com>.
Yeah, I agree.  What about put it in a sandbox area for now if you
want to experiment and move it to trunk after the release?

On Mon, Feb 22, 2010 at 12:52, Alasdair Nottingham <no...@apache.org> wrote:
> Hi,
>
> Given that we are trying to do a release at the moment this makes me
> wonder how we should differentiate things which are experimental and
> wouldn't be part of the release, and things which are to be part of
> the release.
>
> Thoughts?
> Alasdair
>
> On 22 February 2010 11:00,  <da...@apache.org> wrote:
>> Hi everyone,
>>
>> In the OSGi Alliance we're currently working on defining 'Asynchronous
>> Distributed OSGi' in various forms and what that might look like.
>> Currently the work is divided up into three areas:
>> * Bringing JMS to OSGi
>> * Message Driven Components
>> * Asynchronous Services
>> We're really at the requirements gathering stage right now...
>>
>> In the mean time I'm looking for a place where we can do some
>> experimentation and collaboration in the area in order to get a better
>> understanding of these problems and potential solutions. I would love
>> to start a component in Aries for this purpose.
>> The three areas identified above are more or less orthogonal, but I
>> can see many cases where they are combined, so at least initially I
>> think a single experimentation area for them would be good.
>> Then, when the dust settles a bit we might extract a number of
>> independent components... Anyway, that's at least how I think we could
>> do it, but I would also have no big issues with three separate
>> experimentation areas (or maybe three submodules of an async top-level
>> module would be a middle ground).
>>
>> Anyway, if people think this is a good idea I can start putting some
>> of the structure in the build system over the coming few days...
>>
>> Best regards,
>>
>> David
>>
>
>
>
> --
> Alasdair Nottingham
> not@apache.org
>



-- 
Cheers,
Guillaume Nodet
------------------------
Blog: http://gnodet.blogspot.com/
------------------------
Open Source SOA
http://fusesource.com

Re: Async communications experimental component

Posted by Alasdair Nottingham <no...@apache.org>.
Hi,

Given that we are trying to do a release at the moment this makes me
wonder how we should differentiate things which are experimental and
wouldn't be part of the release, and things which are to be part of
the release.

Thoughts?
Alasdair

On 22 February 2010 11:00,  <da...@apache.org> wrote:
> Hi everyone,
>
> In the OSGi Alliance we're currently working on defining 'Asynchronous
> Distributed OSGi' in various forms and what that might look like.
> Currently the work is divided up into three areas:
> * Bringing JMS to OSGi
> * Message Driven Components
> * Asynchronous Services
> We're really at the requirements gathering stage right now...
>
> In the mean time I'm looking for a place where we can do some
> experimentation and collaboration in the area in order to get a better
> understanding of these problems and potential solutions. I would love
> to start a component in Aries for this purpose.
> The three areas identified above are more or less orthogonal, but I
> can see many cases where they are combined, so at least initially I
> think a single experimentation area for them would be good.
> Then, when the dust settles a bit we might extract a number of
> independent components... Anyway, that's at least how I think we could
> do it, but I would also have no big issues with three separate
> experimentation areas (or maybe three submodules of an async top-level
> module would be a middle ground).
>
> Anyway, if people think this is a good idea I can start putting some
> of the structure in the build system over the coming few days...
>
> Best regards,
>
> David
>



-- 
Alasdair Nottingham
not@apache.org