[ 
https://issues.apache.org/jira/browse/HBASE-16351?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Mike Drob updated HBASE-16351:
------------------------------
    Attachment: HBASE-16351.v2.patch

[~busbey] - interesting. There are two failure modes apparently - missing 
license and incomplete license - and they would fail at different parts of the 
aggregation. I tested with incomplete license (removing jruby from 
supplemental-models) whereas you tested with missing license (removing 
common-beanutils).

To address your other concern - I don't think we need to check in both assembly 
and shaded-server. For releases we're definitely building both. For precommit, 
I think dependency changes will trigger assembly module (we don't need actual 
assembly:single goal, process-resources phase is enough) so that should be fine 
too. Overall I think it will be rare that somebody is building a shaded client 
without also building an assembly. 

If you feel strongly about it, then I can copy the beanshell over into shaded 
modules, but I don't believe it is worth the repetition at this point.

> do dependency license check via enforcer plugin
> -----------------------------------------------
>
>                 Key: HBASE-16351
>                 URL: https://issues.apache.org/jira/browse/HBASE-16351
>             Project: HBase
>          Issue Type: Improvement
>          Components: build, dependencies
>            Reporter: Sean Busbey
>            Assignee: Mike Drob
>             Fix For: 1.4.0, 2.0.0-alpha-2
>
>         Attachments: HBASE-16351.patch, HBASE-16351.v2.patch
>
>
> As a stop-gap measure we've made our velocity template fail things when we 
> attempt to bundle a cat-x dependency (see HBASE-16318). Unfortunately, the 
> error messages in this case are non-obvious and digging to find the culprit 
> in a partially rendered LICENSE file leaves a lot to be desired.
> The maven enforcer plugin should allow us to fail more gracefully, with 
> output given on the maven console.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to