I'll start with questions you should be asking yourself when suggesting merging in a non-trivial change. There is no change to users, the change is non-trivial from a development perspective.
Do you know what it means to merge in that code? Did you have anything to do with the production of the code you are suggesting to merge in? Do you actually understand the code you are suggesting to merge in? Can you actually support the code you are suggesting to merge in? I know for certain that neither you or Niclas can answer any of those questions in the affirmative, as such I have absolutely no idea why you are suggesting the merge. It's not within your purview to suggest it because you simply don't understand it, nor the consequences of doing so. We expressly put Nexus on Guice first to make that system entirely stable first before attempting anything else. It works well, but we still find issues so I don't think it's wise to move anything else there yet. And what we have is not strict Guice which will only result in confusion if anyone attempted to use it as standard Guice. Stuart is working on documentation which is just as important as the code. I refuse to continue the tradition of foisting out half-baked shit that doesn't work or isn't documented. I don't know why you are in a rush when the people who created the work are not. On Jun 9, 2010, at 6:32 AM, Olivier Lamy wrote: > maybe more details ? > > 2010/6/9 Jason van Zyl <[email protected]>: >> >> On Jun 9, 2010, at 12:54 AM, Olivier Lamy wrote: >> >>> Hi, >>> I have tested >>> (http://code.google.com/p/maven-scm-provider-svnjava/wiki/UsingWithReleasePlugin) >>> and it works nice ! >>> I have only changed a company plugin to made it works : so it was a >>> bad maven usage !. >>> regarding the cnfde there is the issue : >>> https://issues.sonatype.org/browse/SPICE-26 (do you need a patch ?). >>> >>> For all : So now what is the next step ? Integrating this in the maven 3 >>> trunk ? >>> >> >> -1 >> >>> >>> 2010/6/7 Stuart McCulloch <[email protected]>: >>>> On 7 June 2010 07:25, Olivier Lamy <[email protected]> wrote: >>>> >>>>> Hi, >>>>> I have tested some builds. >>>>> Some notes. >>>>> I have this issues currently : >>>>> >>>>> java.lang.UnsupportedOperationException >>>>> at java.util.AbstractMap.put(AbstractMap.java:186) >>>>> at >>>>> org.apache.maven.scm.manager.AbstractScmManager.setScmProvider(AbstractScmManager.java:93) >>>>> >>>>> Now the Map is not any more writable ? >>>>> >>>> >>>> correct - the new container is much more strict about >>>> components monkeying around with internals, such as >>>> directly modifying injected dynamic collections >>>> >>>> if you look at the old Plexus collections code it does >>>> log a warning if you add components directly into the >>>> injected maps / lists, because it makes the internal >>>> book-keeping and synchronization very complicated >>>> >>>> there is a simple workaround which is to push the >>>> contents of the injected map into your own mutable >>>> map, for example by using a setter method like so: >>>> >>>> private Map<Foo> fooMap; >>>> >>>> private void setFooMap( Map<Foo> fooMap ) >>>> { >>>> this.fooMap = new HashMap<Foo>( fooMap ); >>>> } >>>> >>>> the new container would inject the setter method >>>> instead of the field (ie. the setter hides the field) >>>> and your code would still work with Plexus >>>> >>>> I'm not sure something like [1] will works now. (Not tested as I have >>>>> to cut a release :-) ) >>>>> >>>>> AN other issue using the old plugin >>>>> org.codehaus.plexus:plexus-maven-plugin give me : >>>>> -----------------------------------------------------: >>>>> NoClassDefFoundError: >>>>> org.codehaus.plexus.personality.plexus.lifecycle.phase.Suspendable >>>>> >>>>> so upgrading to org.codehaus.plexus:plexus-component-metadata:1.5.4 >>>>> fix the issue. >>>>> >>>> >>>> the new container may be missing some of the more >>>> obscure parts of Plexus, because we took a minimal >>>> approach to keep it lean - any missing pieces can be >>>> reported at https://issues.sonatype.org/browse/SPICE >>>> >>>> I will tests it with some other build (@work). >>>>> >>>> >>>> thanks >>>> >>>> >>>>> [1] >>>>> http://code.google.com/p/maven-scm-provider-svnjava/wiki/UsingWithReleasePlugin >>>>> >>>>> >>>>> 2010/6/7 Jemos Infra <[email protected]>: >>>>>> Hi all, >>>>>> >>>>>> I'm working on the Maven 3 branch created today by Olivier Lamy >>>>>> >>>>>> [email quote] >>>>>> stuff is here : >>>>>> http://svn.apache.org/repos/asf/maven/maven-3/branches/guice-support/ >>>>>> >>>>>> Thanks ! >>>>>> [/email quote] >>>>>> >>>>>> This branch is supposed to have the Maven 3 version which uses Guice >>>>>> instead of Plexus (actually still uses Plexus to startup but the wiring >>>>>> is done by Guice). >>>>>> >>>>>> I noticed that from this branch the tests are still in Junit 3. Would it >>>>>> be ok to move those to TestNG (preferred) or Junit 4? I could do some >>>>>> work on these if you like. >>>>>> >>>>>> M. >>>>>> >>>>>> >>>>>> >>>>>> --------------------------------------------------------------------- >>>>>> To unsubscribe, e-mail: [email protected] >>>>>> For additional commands, e-mail: [email protected] >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Olivier >>>>> http://twitter.com/olamy >>>>> http://fr.linkedin.com/in/olamy >>>>> http://www.viadeo.com/fr/profile/olivier.lamy7 >>>>> >>>>> --------------------------------------------------------------------- >>>>> To unsubscribe, e-mail: [email protected] >>>>> For additional commands, e-mail: [email protected] >>>>> >>>> >>>> -- >>>> Cheers, Stuart >>>> >>> >>> >>> >>> -- >>> Olivier >>> http://twitter.com/olamy >>> http://fr.linkedin.com/in/olamy >>> http://www.viadeo.com/fr/profile/olivier.lamy7 >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: [email protected] >>> For additional commands, e-mail: [email protected] >>> >> >> Thanks, >> >> Jason >> >> ---------------------------------------------------------- >> Jason van Zyl >> Founder, Apache Maven >> http://twitter.com/jvanzyl >> --------------------------------------------------------- >> >> A language that doesn’t affect the way you think about programming is not >> worth knowing. >> >> -— Alan Perlis >> >> >> >> > > > > -- > Olivier > http://twitter.com/olamy > http://fr.linkedin.com/in/olamy > http://www.viadeo.com/fr/profile/olivier.lamy7 > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > Thanks, Jason ---------------------------------------------------------- Jason van Zyl Founder, Apache Maven http://twitter.com/jvanzyl --------------------------------------------------------- believe nothing, no matter where you read it, or who has said it, not even if i have said it, unless it agrees with your own reason and your own common sense. -- Buddha
