Hi all,
The issue is with the Felix plugin that builds OSGi bundles and requires
dependencies of artifact type JAR. I did not made it work to consume a module
of type WAR. That's why classes build as separate library and then consumed by
two successors. One successor is a web application the other is a OSGi bundle.
For the OSGi bundles all class libraries need to be collected and placed in a
lib folder within the bundle. Then the MANIFEST.MF of the bundle has to list
all libraries in the value of Bundle-ClassPath property.
If we get that with having the previous WAR only modules then I will be fine as
well.
Regards,
Stephan
Am 23.09.2011 um 18:32 schrieb Gabriele Columbro:
> Hi Florent,
>
> that was exactly how we implemented it in the build before, with WAR projects
> producing "their" WAR artifact AND a "-classes.jar" JAR artifact attached.
>
> I think is enough for any project (e.g. the OSGI bundles) that needs to
> depend on those JARs. Agreed?
>
> If you guys agree I can have a look at fixing this,
> Gab
>
> On Sep 23, 2011, at 4:48 PM, Florent Guillaume wrote:
>
>> Hi,
>>
>> Whatever you do, please ensure we still have a way of getting at the
>> server bindings as a normal JAR dependency. I have to use some of the
>> server binding classes (servlets, etc) but I need a JAR library, not a
>> completely built WAR.
>> FWIW in the past I used the
>> chemistry-opencmis-server-bindings-0.4.0-classes.jar ("classes"
>> classifier) but now it's been removed.
>>
>> Florent
>>
>>
>> On Fri, Sep 23, 2011 at 11:23 AM, Gabriele Columbro <[email protected]>
>> wrote:
>>> Hi guys,
>>>
>>> both for server bindings and server inmemory I noticed we now have 2
>>> separate projects:
>>>
>>> - <artifact>.jar
>>> - <arfifact>-war.war
>>>
>>> This seems to me overhead and misleading, apart from having caused problems
>>> in the build. What was the reason for this separation?
>>> Can we go back to just having a WAR project which producing an attached JAR
>>> ? (as per [1])
>>>
>>> If you are happy with that I will proceed with creating and issue and
>>> fixing this.
>>>
>>> WDYT?
>>>
>>> Ciao!
>>> Gab
>>>
>>> [1] http://maven.apache.org/plugins/maven-war-plugin/faq.html#attached
>>> Begin forwarded message:
>>>
>>>> From: Gabriele Columbro <[email protected]>
>>>> Date: September 17, 2011 3:19:09 PM GMT+02:00
>>>> To: [email protected]
>>>> Subject: [VOTE][CANCELED] Release Apache Chemistry OpenCMIS 0.5.0 (RC1)
>>>>
>>>> For now I will just update the assembly to include the war, and re-run the
>>>> release.
>>>>
>>>> As a side note:
>>>> Why there are 2 separate projects (war and jar) for inmemory?
>>>>
>>>> Couldn't we just have had the 'war' project and produce the classes jar
>>>> with a separate classifier ? (e..g -"classes.jar" , as per
>>>> http://maven.apache.org/plugins/maven-war-plugin/faq.html#attached). Was
>>>> that the purpose of having two separate projects?
>>>>
>>>> Ciao!
>>>> Gab
>>>>
>>>> On Sep 17, 2011, at 2:54 PM, Stephan Klevenz wrote:
>>>>
>>>>> Ups. That's because of the OSGi stuff. The inmemory war can be found in
>>>>> sub project:
>>>>>
>>>>> chemistry-opencmis-server-inmemory-war
>>>>>
>>>>> Regards,
>>>>> Stephan
>>>>>
>>>>>
>>>>>
>>>>> Am 17.09.2011 um 14:30 schrieb Florian Müller:
>>>>>
>>>>>> -1
>>>>>>
>>>>>> chemistry-opencmis-dist-0.5.0-server-webapps.(zip|tar.gz) do not contain
>>>>>> chemistry-opencmis-server-inmemory-0.5.0.war.
>>>>>> We should provide a ready-to-use InMemory Repository WAR. We have
>>>>>> provided it in previous releases.
>>>>>>
>>>>>> Sorry...
>>>>>>
>>>>>> Florian
>>>>>>
>>>>>>
>>>>>> On 16/09/2011 20:14, Gabriele Columbro wrote:
>>>>>>> Hi Chemists,
>>>>>>>
>>>>>>> finally got time to kick OpenCMIS 0.5.0 release out.
>>>>>>>
>>>>>>> Chemistry OpenCMIS 0.5.0-RC1 is available and ready for your testing
>>>>>>> and voting.
>>>>>>>
>>>>>>> You can find the main release candidate artifacts (for distribution at
>>>>>>> apache.org/dist) at:
>>>>>>> http://people.apache.org/~gabriele/chemistry/opencmis/0.5.0-RC1/dist/ .
>>>>>>> I added OSGI client and server bundles to the /dist packages.
>>>>>>>
>>>>>>> The full set of Maven artifacts (for distribution at
>>>>>>> repository.apache.org and Maven Central) is staged at:
>>>>>>> https://repository.apache.org/content/repositories/orgapachechemistry-068/
>>>>>>>
>>>>>>> Sources tag can be found at:
>>>>>>> http://svn.apache.org/repos/asf/chemistry/opencmis/tags/chemistry-opencmis-0.5.0-RC1/
>>>>>>>
>>>>>>> Maven generated javadoc/test reports are being deployed right now at
>>>>>>> svn.apache.org (and modifiable via CMS) so they will be staged soon at
>>>>>>> http://chemistry.staging.apache.org/java/0.5.0/maven/ (give it a 30
>>>>>>> more minutes at least since SVN deployment is slow)
>>>>>>>
>>>>>>> I also updated (but not published) the site so that you can review the
>>>>>>> correctness of updated index and download page at:
>>>>>>> http://chemistry.staging.apache.org/java/opencmis.html
>>>>>>> (please abstain from pushing the site to prod, I will do that when the
>>>>>>> vote is completed)
>>>>>>>
>>>>>>> For detailed release notes check Jira at:
>>>>>>> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12310936&version=12316542
>>>>>>> (unresolved issues will be pushed to the next release)
>>>>>>>
>>>>>>> You can find a guide to testing staged releases at:
>>>>>>> http://maven.apache.org/guides/development/guide-testing-releases.html.
>>>>>>>
>>>>>>> The vote is open for 72 hours and passes if a majority of at least
>>>>>>> three +1 Chemistry PMC votes are cast.
>>>>>>>
>>>>>>> Please cast your votes!
>>>>>>>
>>>>>>> [ ] +1 Release this package as Apache Chemistry OpenCMIS 0.5.0
>>>>>>> [ ] -1 Do not release this package because...
>>>>>>>
>>>>>>> Thanks everyone for the great effort into making this happen &
>>>>>>> apologies again for the delay!
>>>>>>>
>>>>>>> Ciao,
>>>>>>> Gab
>>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>>
>>
>>
>>
>> --
>> Florent Guillaume, Director of R&D, Nuxeo
>> Open Source, Java EE based, Enterprise Content Management (ECM)
>> http://www.nuxeo.com http://www.nuxeo.org +33 1 40 33 79 87
>
>
----
Stephan Klevenz
Fabrikstr. 45
69126 Heidelberg
Tel.: +49 6221 879625
Fax.: +49 6221 339926
Email: [email protected]
