After this change, I'm still seeing the same failures on Jenkins, but
not seeing them locally :(

I may try the DependencyGraphBuilder as Hervé suggested, but it will
probably take another day, because I have some day job work to finish today.

On 10/22/2012 12:09 PM, Arnaud Héritier wrote:
> As discussed on IRC for now I don't see a better solution.
> Someone else has a feedback or some ideas ?
> 
> Arnaud
> 
> On Mon, Oct 22, 2012 at 6:21 PM, Paul Gier <[email protected]> wrote:
> 
>> I have now split out the manual purge option into a separate mojo, and
>> added the requiresDependencyResolution annotation to the
>> purge-local-repository mojo.  Hopefully, this will allow the dependency
>> resolution to happen more predictably.
>>
>> The advantage of this is that the code is much simpler, and most of the
>> ITs seem to work with Maven 2 and 3.  Hopefully, all the active ITs will
>> pass on Jenkins now.
>>
>> However, there are a couple of disadvantages to this approach.  One is
>> that the full dependency tree is always resolved before the purge which
>> makes the plugin inefficient for some use-cases.  But hopefully, those
>> use-cases can be handled by the manual purge goal instead.  The other
>> disadvantage is that the IT which uses a pom with an unresolvable
>> dependency (bad pom) does not work because Maven fails before the plugin
>> has a chance to re-resolve it.  This seems like a fairly important use
>> case to me, because one of the reasons for purging might be that you
>> have a bad pom in the local repo.  Unfortunately, I'm not able to figure
>> out how to get around this in a reliable way.
>>
>> Eventually, we should probably just break Maven 2 compatibility, and use
>> the Aether API for dependency resolution.  Maybe this can be done for
>> version 3.0 of the dependency plugin.
>>
>>
>> On 10/22/2012 08:16 AM, Arnaud Héritier wrote:
>>> I was looking at the issue and don't really understand what's wrong with
>>> what you did.
>>> I think you are facing the *fucking bug/feature* in Maven 3 (from aether)
>>> that rejects some dependencies that are available in the local repository
>>> but not in the remote.
>>> This is because you install manually the artifact in the local repo and
>>> thus ather doesn't want it and we cannot bypass this control.
>>> It may explain why ITs are ok with Maven 2 but not Maven 3 on Jenkins.
>>>
>>> Arnaud
>>>
>>> On Mon, Oct 22, 2012 at 2:58 PM, Paul Gier <[email protected]> wrote:
>>>
>>>> Sorry about that, looks like two of them are failing on Jenkins also.
>>>> I'm not sure why I don't see the failures locally, I tested with Maven
>>>> 2.2.1 and 3 and java 5 and 6.  I'll see if I can figure out the cause.
>>>>
>>>> On 10/22/2012 05:25 AM, Arnaud Héritier wrote:
>>>>> FYI, I will need to release also maven-dependency-analyzer
>>>>> I need to check its status
>>>>> Also 3 ITs added by Paul this WE are failing on my laptop. I will study
>>>> if
>>>>> I have an environment issue of if this is a problem in ITs or the code
>>>>>
>>>>> cheers
>>>>>
>>>>> On Mon, Oct 22, 2012 at 9:45 AM, Arnaud Héritier <[email protected]
>>>>> wrote:
>>>>>
>>>>>> no problem, you're doing the harder work :-)
>>>>>> I'll launch the release process today.
>>>>>> I hope to not have too many glitches as I didn't do that for a (too)
>>>> long
>>>>>> time..
>>>>>>
>>>>>> Arnaud
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Mon, Oct 22, 2012 at 6:25 AM, Paul Gier <[email protected]> wrote:
>>>>>>
>>>>>>> Ok, I'm done working on the purge-local-repositories goal for now,
>>>>>>> thanks for waiting for me.
>>>>>>>
>>>>>>> On 10/18/2012 02:31 AM, Arnaud Héritier wrote:
>>>>>>>> Hi Paul,
>>>>>>>>
>>>>>>>>   Yes no problem I can do it tomorrow or this week end.
>>>>>>>>   I just would like to have it out before the end of next week
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>>
>>>>>>>> Arnaud
>>>>>>>>
>>>>>>>> On Thu, Oct 18, 2012 at 5:06 AM, Paul Gier <[email protected]>
>> wrote:
>>>>>>>>
>>>>>>>>> Hi Arnaud,
>>>>>>>>>
>>>>>>>>> Can you hold off one more day?  I have a couple of fixes I want to
>>>> get
>>>>>>>>> in, and I want to do a bit more testing.
>>>>>>>>>
>>>>>>>>> Thanks!
>>>>>>>>>
>>>>>>>>> On 10/17/2012 01:34 AM, Arnaud Héritier wrote:
>>>>>>>>>> ok thx.
>>>>>>>>>> I'll do the release tomorrow
>>>>>>>>>>
>>>>>>>>>> Arnaud
>>>>>>>>>>
>>>>>>>>>> On Tue, Oct 16, 2012 at 11:31 PM, Robert Scholte <
>>>>>>> [email protected]
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Ok, I'm ready
>>>>>>>>>>>
>>>>>>>>>>> Op Tue, 16 Oct 2012 19:07:29 +0200 schreef Arnaud Héritier <
>>>>>>>>>>> [email protected]>:
>>>>>>>>>>>
>>>>>>>>>>>  On Tue, Oct 16, 2012 at 6:55 PM, Robert Scholte <
>>>>>>> [email protected]
>>>>>>>>>>>>> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>  Hi,
>>>>>>>>>>>>>
>>>>>>>>>>>>> I noticed that my MDEP-300 should be easy to fix, but I would
>>>> need
>>>>>>> to
>>>>>>>>>>>>> write an IT with a custom packaging type.
>>>>>>>>>>>>> That would probably take some time, so I'll pick that up for
>> the
>>>>>>> 2.7
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>> ok
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>  I do like to add MDEP-380, a typical Maven3 + SNAPSHOT issue.
>> It
>>>>>>> has a
>>>>>>>>>>>>> good patch, although I think that the useBaseVersion should be
>>>>>>> 'true'
>>>>>>>>> by
>>>>>>>>>>>>> default. I can fix that today or tomorrow.
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>> ok let me know when it will be good on your side.
>>>>>>>>>>>>
>>>>>>>>>>>> thx
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>> thanks,
>>>>>>>>>>>>>
>>>>>>>>>>>>> Robert
>>>>>>>>>>>>>
>>>>>>>>>>>>> Op Tue, 16 Oct 2012 15:28:10 +0200 schreef Stephen Connolly <
>>>>>>>>>>>>> stephen.alan.connolly@gmail.****com
>>>> <stephen.alan.connolly@gmail.
>>>>>>>>> **com<[email protected]>
>>>>>>>>>>>>>>> :
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>  We should probably close MDEP-124 as WONT FIX given that there
>>>>>>> seems
>>>>>>>>> no
>>>>>>>>>>>>>
>>>>>>>>>>>>>> way
>>>>>>>>>>>>>> to fix it... or at least the addition of a configuration
>> element
>>>>>>> to
>>>>>>>>>>>>>> allow
>>>>>>>>>>>>>> flagging specific dependencies as used even if they seem not.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> MDPE-166 also seems like it should just display a different
>>>> output
>>>>>>>>> for
>>>>>>>>>>>>>> <scope>runtime</scope> namely something like "Unable to
>>>> determine
>>>>>>>>>>>>>> whether
>>>>>>>>>>>>>> used or unused as scope is runtime"
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> I think the above two "solutions" would be quick to implement
>>>> and
>>>>>>>>> would
>>>>>>>>>>>>>> close the issues out.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> On 16 October 2012 14:18, Arnaud HERITIER <
>> [email protected]
>>>>>
>>>>>>>>> wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>  Hi all,
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>   i would like to release the dependency plugin soon.
>>>>>>>>>>>>>>>   There only too old issues not closed attached to the
>> current
>>>>>>>>> version
>>>>>>>>>>>>>>> (
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> http://jira.codehaus.org/****browse/MDEP#selectedTab=com.**<
>>>>>>>>> http://jira.codehaus.org/**browse/MDEP#selectedTab=com.**>
>>>>>>>>>>>>>>>
>>>> atlassian.jira.plugin.system.****project%3Aroadmap-panel<http:**
>>>>>>>>>>>>>>> //jira.codehaus.org/browse/**MDEP#selectedTab=com.**
>>>>>>>>>>>>>>> atlassian.jira.plugin.system.**project%3Aroadmap-panel<
>>>>>>>>>
>>>>>>>
>>>>
>> http://jira.codehaus.org/browse/MDEP#selectedTab=com.atlassian.jira.plugin.system.project%3Aroadmap-panel
>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> )
>>>>>>>>>>>>>>> but I suppose they are migrated release after release without
>>>> any
>>>>>>>>> real
>>>>>>>>>>>>>>> work
>>>>>>>>>>>>>>> on them.
>>>>>>>>>>>>>>>   I'm interested especially by this fix to use it as report (
>>>>>>>>>>>>>>> http://jira.codehaus.org/****browse/MDEP-366<
>>>>>>>>> http://jira.codehaus.org/**browse/MDEP-366>
>>>>>>>>>>>>>>> <http://jira.**codehaus.org/browse/MDEP-366<
>>>>>>>>> http://jira.codehaus.org/browse/MDEP-366>
>>>>>>>>>>>>>>>> )
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> and it has the perf improvement
>>>>>>>>>>>>>>> with the plexus-utils upgrade (http://jira.codehaus.org/**
>>>>>>>>>>>>>>> browse/MDEP-377 <http://jira.codehaus.org/**browse/MDEP-377<
>>>>>>>>> http://jira.codehaus.org/browse/MDEP-377>
>>>>>>>>>>>>>>>> )
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>   Nobody against it ?
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Cheers
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> -----
>>>>>>>>>>>>>>> Arnaud Héritier
>>>>>>>>>>>>>>> http://aheritier.net
>>>>>>>>>>>>>>> Mail/GTalk: [email protected]
>>>>>>>>>>>>>>> Twitter/Skype : aheritier
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>  ------------------------------****----------------------------**
>>>>>>>>>>>>> --**---------
>>>>>>>>>>>>> To unsubscribe, e-mail: [email protected].****org<
>>>>>>>>>>>>> dev-unsubscribe@maven.**apache.org <
>>>>>>> [email protected]
>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> For additional commands, e-mail: [email protected]
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>
>>>>>>>
>>>>
>> ------------------------------**------------------------------**---------
>>>>>>>>>>> To unsubscribe, e-mail: [email protected].**org<
>>>>>>>>> [email protected]>
>>>>>>>>>>> For additional commands, e-mail: [email protected]
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>> ---------------------------------------------------------------------
>>>>>>>>> To unsubscribe, e-mail: [email protected]
>>>>>>>>> For additional commands, e-mail: [email protected]
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> ---------------------------------------------------------------------
>>>>>>> To unsubscribe, e-mail: [email protected]
>>>>>>> For additional commands, e-mail: [email protected]
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> -----
>>>>>> Arnaud Héritier
>>>>>> 06-89-76-64-24
>>>>>> http://aheritier.net
>>>>>> Mail/GTalk: [email protected]
>>>>>> Twitter/Skype : aheritier
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: [email protected]
>>>> For additional commands, e-mail: [email protected]
>>>>
>>>>
>>>
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>>
>>
> 
> 

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to