Cool ! Thanks a lot Robert (and Ben).

Cheers
Le 7 août 2013 21:28, "Robert Scholte" <[email protected]> a
écrit :

> https://jira.codehaus.org/**browse/HAUS-2316<https://jira.codehaus.org/browse/HAUS-2316>and
>  fixed
>
> On Tue, 06 Aug 2013 23:27:34 +0200, Baptiste MATHUS <[email protected]>
> wrote:
>
>  Hi all,
>>
>> For the record, I just played a bit with the extra-enforcer-rules bamboo
>> configuration. I manually scheduled a build but this should still take
>> some
>> time before it begins since there's only two agents and one seems stuck.
>> See
>> https://bamboo-ci.codehaus.**org/build/admin/edit/**
>> editBuildTasks.action?**buildKey=MOJO-**MEXTRAENFORCERRULES-JOB1<https://bamboo-ci.codehaus.org/build/admin/edit/editBuildTasks.action?buildKey=MOJO-MEXTRAENFORCERRULES-JOB1>
>> for
>> details.
>>
>> For those interested, I added two executions of "mvn clean verify
>> -DenforcerPluginVersion=1.3.1" with m221 and m3. I might have added more
>> combinations, but it's actually a bit cumbersome to configure many "axis"
>> (IIUC, you actually have to add more executions and just configure each
>> one
>> separately. Matrix builds, I miss you here ;-)).
>>
>> Btw, Robert, as you talk about it, I think that we should actually ask for
>> m3.1 installation isn't it? Or do you think we better wait for m3.1.1?
>>
>> Cheers
>>
>>
>> 2013/7/31 Robert Scholte <[email protected]>
>>
>>  Hi Baptiste,
>>>
>>> not sure why you think that we only build with M2.2.1, see
>>> https://bamboo-ci.codehaus.****org/build/admin/edit/****
>>> editBuildTasks.action?
>>> **buildKey=MOJO-MAPPASM-JOB1<h**ttps://bamboo-ci.codehaus.org/**
>>> build/admin/edit/**editBuildTasks.action?**buildKey=MOJO-MAPPASM-JOB1<https://bamboo-ci.codehaus.org/build/admin/edit/editBuildTasks.action?buildKey=MOJO-MAPPASM-JOB1>
>>> >
>>> This is probably a good example of a lot of combinations of JDKs and
>>> Maven
>>> versions, both M2 and M3.
>>> Ben has installed specific versions of Maven for me in he past, so it
>>> shouldn't be too hard to get M3.1 here as well.
>>>
>>> I recently noticed that the aggregator has been disabled because it
>>> claims
>>> a lot of diskspace. This is probably caused by local repo's per plugin
>>> for
>>> the ITs, so we need think of a way to clean it up afterwards. With
>>> Jenkins
>>> we will have the same problem.
>>>
>>> Yes, there's one important downside with bamboo infrastructure: IIRC due
>>> to the agents-setup you can't access the workspace. So you have to do it
>>> with the logging or ask support@ to send you a zip.
>>> I'm only interested in Jenkins if it stays within the Codehaus
>>> infrastructure, i.e. controlled through xircles.
>>>
>>> If you need help with some specific plugins to run with Bamboo, just PM
>>> me
>>> with your concrete ideas/questions.
>>>
>>> Robert
>>>
>>>
>>>
>>> On Wed, 31 Jul 2013 14:10:25 +0200, Baptiste MATHUS <[email protected]>
>>> wrote:
>>>
>>>  Hi all,
>>>
>>>>
>>>> I was thinking about the fact that IIUC we currently build in Bamboo
>>>> only
>>>> with one version of Maven (2.2.1). Sure, YMMV, but I remember digging a
>>>> bit
>>>> to see how to change it and I found it was a bit complicated and/or slow
>>>> (granted, I don't know bamboo).
>>>>
>>>> I wonder if using Jenkins it wouln't be a little simpler to handle
>>>> multi-aspects testing (*matrix builds*, I'm looking at you).
>>>>
>>>>
>>>> We could in fact come up with a quite standard matrix build template for
>>>> any plugin @MOJO to test for example different maven versions, different
>>>> JDK, or even different parameters values (thinking of
>>>> extra-enforcer-rules
>>>> for examples, where we could test against the differen m-e-p versions
>>>> quite
>>>> easily using a matrix build).
>>>>
>>>> Maybe we could ask CloudBees to see if they can provide the
>>>> corresponding
>>>> required infra? Or ask for Jenkins@MOJO?
>>>>
>>>> WDYT?
>>>>
>>>> Thanks
>>>>
>>>> 2013/3/22 Stephen Connolly <stephen.alan.connolly@gmail.****com<
>>>> stephen.alan.connolly@**gmail.com <[email protected]>>
>>>> >
>>>>
>>>>  There are things we can do with the RBAC plugin and folders that might
>>>>
>>>>> solve those issues... but I need to chat with Ben as to how to delegate
>>>>> auth for a mojo project jenkins to xircles...
>>>>>
>>>>>
>>>>> On 22 March 2013 12:17, Mirko Friedenhagen <[email protected]>**
>>>>> wrote:
>>>>>
>>>>>  Hello,
>>>>>
>>>>>>
>>>>>> I would like Jenkins better for codehaus as well (I hope I may reuse
>>>>>> the
>>>>>> account data from codehaus, jenkins-on-cloudbees e.g. would like to
>>>>>> get
>>>>>> insight into public *and* private repositories on github, which makes
>>>>>> it a
>>>>>> no-go for me).
>>>>>>
>>>>>> Regards Mirko
>>>>>> --
>>>>>> Sent from my mobile
>>>>>> On Mar 21, 2013 10:08 AM, "Stephen Connolly" <
>>>>>> stephen.alan.connolly@gmail.****com <stephen.alan.connolly@gmail.**
>>>>>> com <[email protected]>>>
>>>>>> wrote:
>>>>>>
>>>>>>  Well not suggesting *codehaus* abandon it. Just that we set up a JaaS
>>>>>>
>>>>>>> for the mojo project
>>>>>>>
>>>>>>> On Thursday, 21 March 2013, Baptiste MATHUS wrote:
>>>>>>>
>>>>>>>  Sure, I personnally know Jenkins a lot better and sure would be more
>>>>>>>
>>>>>>>> comfortable with it.
>>>>>>>> But I also like JIRA and don't know if atlassian would be happy with
>>>>>>>> us
>>>>>>>> ditching bamboo in favor of Jenkins.
>>>>>>>>
>>>>>>>> Cheers
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/20 Stephen Connolly <stephen.alan.connolly@gmail.****com<
>>>>>>>> stephen.alan.connolly@**gmail.com <[email protected]>
>>>>>>>> >
>>>>>>>> >
>>>>>>>>
>>>>>>>> Do we want a Jenkins at CloudBees? Might be easier for people to
>>>>>>>> grok?
>>>>>>>>
>>>>>>>>
>>>>>>>> On Tuesday, 19 March 2013, Robert Scholte wrote:
>>>>>>>>
>>>>>>>> Let me go through all the jobs.
>>>>>>>> It looks to me something has changed, because I'm pretty sure I
>>>>>>>> configured notifications for failed builds and first successful to
>>>>>>>> all
>>>>>>>> committers (users who have committed to the build) for every mojo
>>>>>>>> job.
>>>>>>>>
>>>>>>>> No real need to keep checking Bamboo as long as its green ;)
>>>>>>>>
>>>>>>>> Robert
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, 18 Mar 2013 22:57:52 +0100, Baptiste MATHUS <
>>>>>>>> [email protected]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>  FYI, I just committed the corresponding fix/evolution. CI is now
>>>>>>>> back
>>>>>>>> to
>>>>>>>> green (thanks Mirko for noticing it, I'll watch it better next
>>>>>>>> time).
>>>>>>>>
>>>>>>>> I'm gonna launch the 3rd try :-).
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/17 Arnaud Héritier <[email protected]>
>>>>>>>>
>>>>>>>>  I'm in favor to have just a warning.
>>>>>>>> It's enough
>>>>>>>>
>>>>>>>>
>>>>>>>> On Sun, Mar 17, 2013 at 9:25 PM, Baptiste MATHUS <
>>>>>>>> [email protected]
>>>>>>>> >wrote:
>>>>>>>>
>>>>>>>>  Yup, I was just copying the code ;-). Thanks.
>>>>>>>>
>>>>>>>> Btw, I'm gonna just display a warning. But if you feel we should
>>>>>>>> fail
>>>>>>>> the
>>>>>>>> build, just let me know.
>>>>>>>> I feel this might be counter-productive to fail the build for that.
>>>>>>>> People might in fact allow running 2.x and 3.x versions of maven on
>>>>>>>> the
>>>>>>>> same build, and failing here even if there's actually no cycle might
>>>>>>>> make
>>>>>>>> them just remove that rule (banCircularDependencies).
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/17 Arnaud Héritier <[email protected]>
>>>>>>>>
>>>>>>>>  The RequireMavenVersion Rule may help ?
>>>>>>>>
>>>>>>>> https://svn.apache.org/repos/******asf/maven/enforcer/trunk/**<https://svn.apache.org/repos/****asf/maven/enforcer/trunk/**>
>>>>>>>> **<https://svn.apache.org/repos/****asf/maven/enforcer/trunk/**<https://svn.apache.org/repos/**asf/maven/enforcer/trunk/**>
>>>>>>>> >
>>>>>>>> enforcer-rules/src/main/java/******org/apache/maven/plugins/**
>>>>>>>> enforcer/RequireMavenVersion.******java<https://svn.apache.**org/**<https://svn.apache.org/**>
>>>>>>>> repos/asf/maven/enforcer/****trunk/enforcer-rules/src/main/****
>>>>>>>> java/org/apache/maven/plugins/****enforcer/**
>>>>>>>> RequireMavenVersion.**java<htt**ps://svn.apache.org/repos/asf/**
>>>>>>>> maven/enforcer/trunk/enforcer-**rules/src/main/java/org/**
>>>>>>>> apache/maven/plugins/enforcer/**RequireMavenVersion.java<https://svn.apache.org/repos/asf/maven/enforcer/trunk/enforcer-rules/src/main/java/org/apache/maven/plugins/enforcer/RequireMavenVersion.java>
>>>>>>>> >
>>>>>>>> >
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Sun, Mar 17, 2013 at 9:12 PM, Baptiste MATHUS <
>>>>>>>> [email protected]
>>>>>>>> >wrote:
>>>>>>>>
>>>>>>>>  OK, I'll go that way.
>>>>>>>> If anyone sees this message just now and knows *the right way to
>>>>>>>> check
>>>>>>>> the maven version in use in an enforcer rule, I'm interested.*
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/17 Mirko Friedenhagen <[email protected]>
>>>>>>>>
>>>>>>>>  I would prefer solution 1 and just document this behaviour (and set
>>>>>>>> the Bamboo job to Maven 3). Otherwise you will have this switch in
>>>>>>>> code,
>>>>>>>> you should introduce two ITs, one with a failure status on Maven 3
>>>>>>>> and
>>>>>>>> one
>>>>>>>> with a skipped message on Maven 2.
>>>>>>>>
>>>>>>>> Honestly, new feature, new Maven should be no problem.
>>>>>>>>
>>>>>>>> Regards Mirko
>>>>>>>> --
>>>>>>>> Sent from my mobile
>>>>>>>> On Mar 16, 2013 10:42 PM, "Baptiste MATHUS" <[email protected]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>  In the circular IT, there's a circular dependency (now that's some
>>>>>>>> news!).
>>>>>>>>
>>>>>>>> But M2 just ignores it silently, not M3. I think we hit
>>>>>>>> http://jira.codehaus.org/******browse/MNG-1944<http://jira.codehaus.org/****browse/MNG-1944>
>>>>>>>> <http://jira.**codehaus.org/**browse/MNG-1944<http://jira.codehaus.org/**browse/MNG-1944>
>>>>>>>> **>
>>>>>>>> <http://jira.**codehaus.org/**browse/MNG-1944<http://codehaus.org/browse/MNG-1944>
>>>>>>>> <http://jira.**codehaus.org/browse/MNG-1944<http://jira.codehaus.org/browse/MNG-1944>
>>>>>>>> >
>>>>>>>> >
>>>>>>>>
>>>>>>>>
>>>>>>>> So, here we have an choice: either we
>>>>>>>> 1) explicitly state (and check during execution?) that this rule can
>>>>>>>> only be useful with M3 or
>>>>>>>> 2) we simply remove it.
>>>>>>>>
>>>>>>>> I'd be for solution 1. I feel it's a valuable addition and as it
>>>>>>>> works for the latest and greatest Maven, this makes sense to keep
>>>>>>>> it.
>>>>>>>>
>>>>>>>> WDYT?
>>>>>>>>
>>>>>>>> Cheers
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/3/15 Baptiste MATHUS <[email protected]>
>>>>>>>>
>>>>>>>>  No problem, and true for the IT.
>>>>>>>> I've copied the demo from the website and forgot to update that
>>>>>>>> part.
>>>>>>>> I can re-roll a release. Now
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>  --
>>>>>>> Sent from my phone
>>>>>>>
>>>>>>> --
>>>>>>> Baptiste <Batmat> MATHUS - http://batmat.net
>>>>>>> Sauvez un arbre,
>>>>>>> Mangez un castor ! nbsp;!
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>> --
>>> Using Opera's revolutionary email client: http://www.opera.com/mail/
>>>
>>> ------------------------------****----------------------------**
>>> --**---------
>>> To unsubscribe from this list, please visit:
>>>
>>>    
>>> http://xircles.codehaus.org/****manage_email<http://xircles.codehaus.org/**manage_email>
>>> <http://xircles.**codehaus.org/manage_email<http://xircles.codehaus.org/manage_email>
>>> >
>>>
>>>
>>>
>>>
>>
>>
>
> --
> Using Opera's revolutionary email client: http://www.opera.com/mail/
>
> ------------------------------**------------------------------**---------
> To unsubscribe from this list, please visit:
>
>    
> http://xircles.codehaus.org/**manage_email<http://xircles.codehaus.org/manage_email>
>
>
>

Reply via email to