[DISCUSS] JAXWS-API 2.2 spec

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

[DISCUSS] JAXWS-API 2.2 spec

Guillaume Nodet
Administrator
It seems the JAXWS-API 2.2 spec uses a quite different mechanism than all
other specs.
It embeds the Geronimo osgi bits for specs, while all other usually use the
ServiceMix one (the osgi locator).
In addition to be incoherent, those specs also use a slightly different
mechanism, as providers have to opt-in by adding an osgi header.

Is there any reason for not using the same mechanism as for other specs ?

--
------------------------
Guillaume Nodet
------------------------
Blog: http://gnodet.blogspot.com/
------------------------
Open Source SOA
http://fusesource.com
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Daniel  Kulp
On Monday, October 24, 2011 3:51:02 PM Guillaume Nodet wrote:
> It seems the JAXWS-API 2.2 spec uses a quite different mechanism than all
> other specs.
> It embeds the Geronimo osgi bits for specs, while all other usually use the
> ServiceMix one (the osgi locator).
> In addition to be incoherent, those specs also use a slightly different
> mechanism, as providers have to opt-in by adding an osgi header.
>
> Is there any reason for not using the same mechanism as for other specs ?

To provide some context around this issue:

http://irclogs.dankulp.com/logs/irclogger_log/cxf?date=2011-10-24,Mon&sel=56#l52


--
Daniel Kulp
[hidden email]
http://dankulp.com/blog
Talend - http://www.talend.com
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Gert Vanthienen
Administrator
In reply to this post by Guillaume Nodet
L.S.,


Personally, I don't think this was intentional - at least, I don't recall
any discussion about switching back to the Geronimo implementation so I'm
guessing we just picked the wrong implementation when we noticed the two
OSGi activator implementations in the resulting JAR.  Unless Freeman or
someone else knows about a good reason to stick with the Geronimo
implementation, it looks like it would be both more consistent and easier
(i.e. not opt-in required) to switch to the ServiceMix implementation for
everything.


Regards,

Gert Vanthienen
------------------------
FuseSource
Web: http://fusesource.com
Blog: http://gertvanthienen.blogspot.com/


On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]> wrote:

> It seems the JAXWS-API 2.2 spec uses a quite different mechanism than all
> other specs.
> It embeds the Geronimo osgi bits for specs, while all other usually use the
> ServiceMix one (the osgi locator).
> In addition to be incoherent, those specs also use a slightly different
> mechanism, as providers have to opt-in by adding an osgi header.
>
> Is there any reason for not using the same mechanism as for other specs ?
>
> --
> ------------------------
> Guillaume Nodet
> ------------------------
> Blog: http://gnodet.blogspot.com/
> ------------------------
> Open Source SOA
> http://fusesource.com
>
Regards,

Gert Vanthienen
------------------------
Open Source SOA: http://fusesource.com
Blog: http://gertvanthienen.blogspot.com/
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Freeman-2
Hi Guys,

IIRC there's no special reason that we use the Geronimo locator.
At the moment when I added jaxws-api 2.2 bundle, I found that geronimo-
jaxws_2.2_spec.jar is already OSGi friendly and has the locator thing  
there,  but  import javax.xml.namespace with version="1.0", which  
prevent us to use the one from system bundle, so I simply wrap it to  
fix the javax.xml.namespace with version="1.0" issue.

I'll change that bundle to use SMX OSGi locator to keep consistent.

Freeman
On 2011-10-24, at 下午10:24, Gert Vanthienen wrote:

> L.S.,
>
>
> Personally, I don't think this was intentional - at least, I don't  
> recall
> any discussion about switching back to the Geronimo implementation  
> so I'm
> guessing we just picked the wrong implementation when we noticed the  
> two
> OSGi activator implementations in the resulting JAR.  Unless Freeman  
> or
> someone else knows about a good reason to stick with the Geronimo
> implementation, it looks like it would be both more consistent and  
> easier
> (i.e. not opt-in required) to switch to the ServiceMix  
> implementation for
> everything.
>
>
> Regards,
>
> Gert Vanthienen
> ------------------------
> FuseSource
> Web: http://fusesource.com
> Blog: http://gertvanthienen.blogspot.com/
>
>
> On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]>  
> wrote:
>
>> It seems the JAXWS-API 2.2 spec uses a quite different mechanism  
>> than all
>> other specs.
>> It embeds the Geronimo osgi bits for specs, while all other usually  
>> use the
>> ServiceMix one (the osgi locator).
>> In addition to be incoherent, those specs also use a slightly  
>> different
>> mechanism, as providers have to opt-in by adding an osgi header.
>>
>> Is there any reason for not using the same mechanism as for other  
>> specs ?
>>
>> --
>> ------------------------
>> Guillaume Nodet
>> ------------------------
>> Blog: http://gnodet.blogspot.com/
>> ------------------------
>> Open Source SOA
>> http://fusesource.com
>>

---------------------------------------------
Freeman Fang

FuseSource
Email:[hidden email]
Web: fusesource.com
Twitter: freemanfang
Blog: http://freemanfang.blogspot.com









Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Daniel  Kulp
In reply to this post by Gert Vanthienen
On Monday, October 24, 2011 4:24:36 PM Gert Vanthienen wrote:

> L.S.,
>
>
> Personally, I don't think this was intentional - at least, I don't recall
> any discussion about switching back to the Geronimo implementation so I'm
> guessing we just picked the wrong implementation when we noticed the two
> OSGi activator implementations in the resulting JAR.  Unless Freeman or
> someone else knows about a good reason to stick with the Geronimo
> implementation, it looks like it would be both more consistent and easier
> (i.e. not opt-in required) to switch to the ServiceMix implementation for
> everything.

I certainly like the SMX locator more.  However, now that I understand what is
going on with the Geronimo version, it's less of an issue to me.   CXF now has
the "opt in" header and I've updated the features.xml to pull in the required
Geronimo registry.   Thus, I now have it working.   So consider this a "low
priority, but certainly would be nice to have" kind of thing.

Dan


>
>
> Regards,
>
> Gert Vanthienen
> ------------------------
> FuseSource
> Web: http://fusesource.com
> Blog: http://gertvanthienen.blogspot.com/
>
> On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]> wrote:
> > It seems the JAXWS-API 2.2 spec uses a quite different mechanism than
> > all
> > other specs.
> > It embeds the Geronimo osgi bits for specs, while all other usually use
> > the ServiceMix one (the osgi locator).
> > In addition to be incoherent, those specs also use a slightly different
> > mechanism, as providers have to opt-in by adding an osgi header.
> >
> > Is there any reason for not using the same mechanism as for other specs
> > ?
> >
> > --
> > ------------------------
> > Guillaume Nodet
> > ------------------------
> > Blog: http://gnodet.blogspot.com/
> > ------------------------
> > Open Source SOA
> > http://fusesource.com
--
Daniel Kulp
[hidden email]
http://dankulp.com/blog
Talend - http://www.talend.com
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Gert Vanthienen
Administrator
L.S.,


Since Freeman already committed a fix this morning, how about adding a specs
1.9.0 release to the ServiceMix 4.4.0 release plan?  There's hardly any work
involved in doing the specs release so we might as well include this
improvement while we're at it...


Regards,

Gert Vanthienen
------------------------
FuseSource
Web: http://fusesource.com
Blog: http://gertvanthienen.blogspot.com/


On Tue, Oct 25, 2011 at 5:26 AM, Daniel Kulp <[hidden email]> wrote:

> On Monday, October 24, 2011 4:24:36 PM Gert Vanthienen wrote:
> > L.S.,
> >
> >
> > Personally, I don't think this was intentional - at least, I don't recall
> > any discussion about switching back to the Geronimo implementation so I'm
> > guessing we just picked the wrong implementation when we noticed the two
> > OSGi activator implementations in the resulting JAR.  Unless Freeman or
> > someone else knows about a good reason to stick with the Geronimo
> > implementation, it looks like it would be both more consistent and easier
> > (i.e. not opt-in required) to switch to the ServiceMix implementation for
> > everything.
>
> I certainly like the SMX locator more.  However, now that I understand what
> is
> going on with the Geronimo version, it's less of an issue to me.   CXF now
> has
> the "opt in" header and I've updated the features.xml to pull in the
> required
> Geronimo registry.   Thus, I now have it working.   So consider this a "low
> priority, but certainly would be nice to have" kind of thing.
>
> Dan
>
>
> >
> >
> > Regards,
> >
> > Gert Vanthienen
> > ------------------------
> > FuseSource
> > Web: http://fusesource.com
> > Blog: http://gertvanthienen.blogspot.com/
> >
> > On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]>
> wrote:
> > > It seems the JAXWS-API 2.2 spec uses a quite different mechanism than
> > > all
> > > other specs.
> > > It embeds the Geronimo osgi bits for specs, while all other usually use
> > > the ServiceMix one (the osgi locator).
> > > In addition to be incoherent, those specs also use a slightly different
> > > mechanism, as providers have to opt-in by adding an osgi header.
> > >
> > > Is there any reason for not using the same mechanism as for other specs
> > > ?
> > >
> > > --
> > > ------------------------
> > > Guillaume Nodet
> > > ------------------------
> > > Blog: http://gnodet.blogspot.com/
> > > ------------------------
> > > Open Source SOA
> > > http://fusesource.com
> --
> Daniel Kulp
> [hidden email]
> http://dankulp.com/blog
> Talend - http://www.talend.com
>
Regards,

Gert Vanthienen
------------------------
Open Source SOA: http://fusesource.com
Blog: http://gertvanthienen.blogspot.com/
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Guillaume Nodet
Administrator
Sounds like a good plan, especially asa few things have been fixed in the
spec locator since the last release:

[SMX4-949] use Servicemix OSGi Locator for jaxws 2.2 api bundle
[SMX4-921] Jaxb api throws jaxb exception when using the jre implementation
[SMX4-860] Add a configurable timeout
[SMX4-860] ServiceMix specs should wait for an implementation for a certain
amount of time when in OSGi
[SMX4-859] ServiceMix Specs should check class space consistency before
using an implementation


On Tue, Oct 25, 2011 at 10:45, Gert Vanthienen <[hidden email]>wrote:

> L.S.,
>
>
> Since Freeman already committed a fix this morning, how about adding a
> specs
> 1.9.0 release to the ServiceMix 4.4.0 release plan?  There's hardly any
> work
> involved in doing the specs release so we might as well include this
> improvement while we're at it...
>
>
> Regards,
>
> Gert Vanthienen
> ------------------------
> FuseSource
> Web: http://fusesource.com
> Blog: http://gertvanthienen.blogspot.com/
>
>
> On Tue, Oct 25, 2011 at 5:26 AM, Daniel Kulp <[hidden email]> wrote:
>
> > On Monday, October 24, 2011 4:24:36 PM Gert Vanthienen wrote:
> > > L.S.,
> > >
> > >
> > > Personally, I don't think this was intentional - at least, I don't
> recall
> > > any discussion about switching back to the Geronimo implementation so
> I'm
> > > guessing we just picked the wrong implementation when we noticed the
> two
> > > OSGi activator implementations in the resulting JAR.  Unless Freeman or
> > > someone else knows about a good reason to stick with the Geronimo
> > > implementation, it looks like it would be both more consistent and
> easier
> > > (i.e. not opt-in required) to switch to the ServiceMix implementation
> for
> > > everything.
> >
> > I certainly like the SMX locator more.  However, now that I understand
> what
> > is
> > going on with the Geronimo version, it's less of an issue to me.   CXF
> now
> > has
> > the "opt in" header and I've updated the features.xml to pull in the
> > required
> > Geronimo registry.   Thus, I now have it working.   So consider this a
> "low
> > priority, but certainly would be nice to have" kind of thing.
> >
> > Dan
> >
> >
> > >
> > >
> > > Regards,
> > >
> > > Gert Vanthienen
> > > ------------------------
> > > FuseSource
> > > Web: http://fusesource.com
> > > Blog: http://gertvanthienen.blogspot.com/
> > >
> > > On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]>
> > wrote:
> > > > It seems the JAXWS-API 2.2 spec uses a quite different mechanism than
> > > > all
> > > > other specs.
> > > > It embeds the Geronimo osgi bits for specs, while all other usually
> use
> > > > the ServiceMix one (the osgi locator).
> > > > In addition to be incoherent, those specs also use a slightly
> different
> > > > mechanism, as providers have to opt-in by adding an osgi header.
> > > >
> > > > Is there any reason for not using the same mechanism as for other
> specs
> > > > ?
> > > >
> > > > --
> > > > ------------------------
> > > > Guillaume Nodet
> > > > ------------------------
> > > > Blog: http://gnodet.blogspot.com/
> > > > ------------------------
> > > > Open Source SOA
> > > > http://fusesource.com
> > --
> > Daniel Kulp
> > [hidden email]
> > http://dankulp.com/blog
> > Talend - http://www.talend.com
> >
>



--
------------------------
Guillaume Nodet
------------------------
Blog: http://gnodet.blogspot.com/
------------------------
Open Source SOA
http://fusesource.com
Reply | Threaded
Open this post in threaded view
|

Re: [DISCUSS] JAXWS-API 2.2 spec

Freeman-2
+1

Freeman
On 2011-10-25, at 下午4:53, Guillaume Nodet wrote:

> Sounds like a good plan, especially asa few things have been fixed  
> in the
> spec locator since the last release:
>
> [SMX4-949] use Servicemix OSGi Locator for jaxws 2.2 api bundle
> [SMX4-921] Jaxb api throws jaxb exception when using the jre  
> implementation
> [SMX4-860] Add a configurable timeout
> [SMX4-860] ServiceMix specs should wait for an implementation for a  
> certain
> amount of time when in OSGi
> [SMX4-859] ServiceMix Specs should check class space consistency  
> before
> using an implementation
>
>
> On Tue, Oct 25, 2011 at 10:45, Gert Vanthienen <[hidden email]
> >wrote:
>
>> L.S.,
>>
>>
>> Since Freeman already committed a fix this morning, how about  
>> adding a
>> specs
>> 1.9.0 release to the ServiceMix 4.4.0 release plan?  There's hardly  
>> any
>> work
>> involved in doing the specs release so we might as well include this
>> improvement while we're at it...
>>
>>
>> Regards,
>>
>> Gert Vanthienen
>> ------------------------
>> FuseSource
>> Web: http://fusesource.com
>> Blog: http://gertvanthienen.blogspot.com/
>>
>>
>> On Tue, Oct 25, 2011 at 5:26 AM, Daniel Kulp <[hidden email]>  
>> wrote:
>>
>>> On Monday, October 24, 2011 4:24:36 PM Gert Vanthienen wrote:
>>>> L.S.,
>>>>
>>>>
>>>> Personally, I don't think this was intentional - at least, I don't
>> recall
>>>> any discussion about switching back to the Geronimo  
>>>> implementation so
>> I'm
>>>> guessing we just picked the wrong implementation when we noticed  
>>>> the
>> two
>>>> OSGi activator implementations in the resulting JAR.  Unless  
>>>> Freeman or
>>>> someone else knows about a good reason to stick with the Geronimo
>>>> implementation, it looks like it would be both more consistent and
>> easier
>>>> (i.e. not opt-in required) to switch to the ServiceMix  
>>>> implementation
>> for
>>>> everything.
>>>
>>> I certainly like the SMX locator more.  However, now that I  
>>> understand
>> what
>>> is
>>> going on with the Geronimo version, it's less of an issue to me.    
>>> CXF
>> now
>>> has
>>> the "opt in" header and I've updated the features.xml to pull in the
>>> required
>>> Geronimo registry.   Thus, I now have it working.   So consider  
>>> this a
>> "low
>>> priority, but certainly would be nice to have" kind of thing.
>>>
>>> Dan
>>>
>>>
>>>>
>>>>
>>>> Regards,
>>>>
>>>> Gert Vanthienen
>>>> ------------------------
>>>> FuseSource
>>>> Web: http://fusesource.com
>>>> Blog: http://gertvanthienen.blogspot.com/
>>>>
>>>> On Mon, Oct 24, 2011 at 3:51 PM, Guillaume Nodet <[hidden email]>
>>> wrote:
>>>>> It seems the JAXWS-API 2.2 spec uses a quite different mechanism  
>>>>> than
>>>>> all
>>>>> other specs.
>>>>> It embeds the Geronimo osgi bits for specs, while all other  
>>>>> usually
>> use
>>>>> the ServiceMix one (the osgi locator).
>>>>> In addition to be incoherent, those specs also use a slightly
>> different
>>>>> mechanism, as providers have to opt-in by adding an osgi header.
>>>>>
>>>>> Is there any reason for not using the same mechanism as for other
>> specs
>>>>> ?
>>>>>
>>>>> --
>>>>> ------------------------
>>>>> Guillaume Nodet
>>>>> ------------------------
>>>>> Blog: http://gnodet.blogspot.com/
>>>>> ------------------------
>>>>> Open Source SOA
>>>>> http://fusesource.com
>>> --
>>> Daniel Kulp
>>> [hidden email]
>>> http://dankulp.com/blog
>>> Talend - http://www.talend.com
>>>
>>
>
>
>
> --
> ------------------------
> Guillaume Nodet
> ------------------------
> Blog: http://gnodet.blogspot.com/
> ------------------------
> Open Source SOA
> http://fusesource.com

---------------------------------------------
Freeman Fang

FuseSource
Email:[hidden email]
Web: fusesource.com
Twitter: freemanfang
Blog: http://freemanfang.blogspot.com