One solution was given by Dan where he suggested that for RC we create a
branch. That I guess is still in review board.

-abhi

On 22/01/14 10:42 am, "Animesh Chaturvedi" <animesh.chaturv...@citrix.com>
wrote:

>That seems kludgy we need an elegant solution
>
>-----Original Message-----
>From: Abhinandan Prateek
>Sent: Tuesday, January 21, 2014 9:11 PM
>To: dev@cloudstack.apache.org
>Cc: Animesh Chaturvedi
>Subject: Re: git commit: updated refs/heads/4.2 to 2b34dc5
>
>I looked at 4.3-forward and it seems the revert commit is missing. That
>is why the “SNAPSHOT” extension is not restored.
>
>On 22/01/14 10:31 am, "Abhinandan Prateek" <abhinandan.prat...@citrix.com>
>wrote:
>
>>Animesh,
>>
>>  When you run build_asf script it does the following:
>>
>>1. Removes the -SNAPSHOT extensions from all version numbers, including
>>xapi.
>>2. It creates a commit for this updated pom.xml without the extension.
>>Generates a tag too.
>>3. It does all the building stuff etc.
>>4. Reverses the commit created in 2.
>>
>>Now if you push the commits, you should see 2 commits: one commit
>>changing versions and second reverting it, the branch returns to the
>>pre-build state where ³-SNAPSHOT² is restored.
>>
>>-abhi
>>
>>On 22/01/14 10:15 am, "Animesh Chaturvedi"
>><animesh.chaturv...@citrix.com>
>>wrote:
>>
>>>I ran into this issue again with tools/build/build_asf.sh while
>>>building
>>>4.3 RC. I have gotten around it locally with help from Sheng but do
>>>not know what it the correct way to fix it. I have reopened
>>>CLOUDSTACK-4827
>>>
>>>Any help is appreciated
>>>
>>>-----Original Message-----
>>>From: Wei Zhou [mailto:w.z...@leaseweb.com]
>>>Sent: Monday, December 16, 2013 2:25 AM
>>>To: Abhinandan Prateek; dev@cloudstack.apache.org
>>>Cc: aprat...@apache.org
>>>Subject: RE: git commit: updated refs/heads/4.2 to 2b34dc5
>>>
>>>Abhi, Chips
>>>
>>>The issue CLOUDSTACK-4827 happened when one build 4.2 (with
>>>xapi-5.6.100-1) if compiled 4.1 (with xapi-5.6.100-1-SNAPSHOT) before.
>>>
>>>I just tested some times (built 4.1 at first, then 4.2 with
>>>xapi-5.6.100-1), the issue did not appear. It is weird.
>>>Now I vote +1 on removing -SNAPSHOT as I cannot reproduce it.
>>>If anyone meet this issue later, they can fix it by removing the
>>>org/apache/cloudstack/xapi/ folder in local maven repository.
>>>
>>>
>>>Kind Regards,
>>>
>>>Wei ZHOU
>>>Innovation Engineer Cloud, LeaseWeb B.V.
>>>w.z...@leaseweb.com
>>>
>>>-----Original Message-----
>>>From: Abhinandan Prateek [mailto:abhinandan.prat...@citrix.com]
>>>Sent: zaterdag, 14 december, 2013 04:02
>>>To: dev@cloudstack.apache.org
>>>Cc: aprat...@apache.org; Wei Zhou
>>>Subject: Re: git commit: updated refs/heads/4.2 to 2b34dc5
>>>
>>>
>>>
>>>On 13/12/13 9:20 pm, "Chip Childers" <chipchild...@apache.org> wrote:
>>>>>
>>>>> >  <artifactId>xapi</artifactId>
>>>>> >  <version>5.6.100-1-SNAPSHOT</version>
>>>>> 
>>>>> The specific project version ^^
>>>>> 
>>>>> For all previous releases, we have been releasing this specific
>>>>> pom.xml file with the appropriate *non SNAPSHOT* versions for both
>>>>> the parent version number and the XenServerJava project's version
>>>>> number (specifically setting the latter to 5.6.100-1).
>>>>> 
>>>>> Since we are releasing the XenServerJava code as part of ACS, why
>>>>> would we leave the SNAPSHOT in there?
>>>>> 
>>>>> Did something change that requires it to be added back?
>>>>> 
>>>>> -chip
>>>>
>>>>I'll also point out that the reason that this is doing a mv then the
>>>>perl string changes is that there used to be a bug in the mvn
>>>>versions plugin that changed the XenServerJava version to the ACS
>>>>version.  This appears to have been fixed (just tested).  So
>>>>actually, the mv can be removed or not, it doesn't really matter
>>>>because it's basically a noop.
>>>>
>>>>However -1 still stands unless someone convinces me that we should
>>>>release the XenServerJava project with -SNAPSHOT.  IIRC, that
>>>>actually caused problems for us somehow (but I can't find a reference
>>>>to that to back up my sometimes fuzzy memory).
>>>>
>>>
>>>I was pointed to this ticket CLOUDSTACK-4827. The info is not very
>>>clear and it appears that this fixes probably a bad version for the
>>>repo, and not for the build.
>>>
>>>-abhi
>>>
>>>>
>>>
>>
>

Reply via email to