[jbosstools-issues] [JBoss JIRA] (JBIDE-21012) Why do we deploy JBT components to Nexus snapshots repo?

[Nick Boldt (JIRA)]

    [ https://issues.jboss.org/browse/JBIDE-21012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13132799#comment-13132799 ] 

Nick Boldt commented on JBIDE-21012:
------------------------------------

No, there was nothing in here which found an issue about bad versioning, only problems with Nexus supporting the idea of versioning things like 1.5.1.Beta1-SNAPSHOT. Workaround was to use 1.5.151-SNAPSHOT.

Conclusions were:

* To support pushing update site zips to Nexus, we need to replace alphabetic BUILD_ALIAS = Beta1 with numeric BUILD_ALIAS_NUM = 51. (More examples/details above - https://issues.jboss.org/browse/JBIDE-21012?focusedCommentId=13125499&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13125499 )

* BUT, to support composite sites containing the Nth and N-1th builds of a given component, we can not use Nexus since it only supports a "latest" or Nth build in /unzip/unzip/ URLs. This means downstream builds will fail should an upstream component republish DURING that downstream component's build. These failures are not "CI builds fail faster" but rather "nuissance failures", since a simple respin would fix them.

---

Re: features that detect bad versioning, we have two:

a) compare with baselines - JBIDE-21052
b) versionwatch - JBIDE-20976


> Why do we deploy JBT components to Nexus snapshots repo?
> --------------------------------------------------------
>
>                 Key: JBIDE-21012
>                 URL: https://issues.jboss.org/browse/JBIDE-21012
>             Project: Tools (JBoss Tools)
>          Issue Type: Bug
>          Components: build, updatesite
>    Affects Versions: 4.3.0.Final, 4.4.0.Alpha1
>            Reporter: Nick Boldt
>            Assignee: Nick Boldt
>             Fix For: 4.3.1.Beta1, 4.4.0.Alpha1
>
>
> Currently, we deploy our x.y.z-SNAPSHOT update sites to the JBoss Nexus snapshots repo.
> But other than Locus and the Browsersim Standalone zip, we don't ever use these artifacts, so they just:
> * eat disk space in Nexus
> * consume time/resources in Jenkins
> Since we have the custom profile deploy-to-jboss.org in place for all our artifacts, why don't we set *maven.deploy.skip=true* in the parent pom for all the projects (except of course Locus and Browsersim Standalone, which would use maven.deploy.skip=false) ?



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)