You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@continuum.apache.org by "Harper, Brad" <br...@epsiia.com> on 2008/04/15 18:10:32 UTC

CI Use of Repositories

Hello:

I'm instituting continuous integration builds for a collection of maven
projects. 

Under interactive builds, developers may deploy artifacts to a SNAPSHOT
repo. We may also run deploy:deploy to an internal remote repo [to share
non-SNAPSNOT baselines] and release:perform-release to share artifacts
with QA [from that same internal remote repo] for formal evaluations
resulting in a final release.

Artifacts under formal QA evaluation are submitted as a series of
release candidates [via the internal remote repo], with specific maven
<version> values, e.g. 1.2.3-RC-1, 1.2.3-RC-2, etc., until approved.

The question is ...

Should ci-built artifacts be deployed into a separate repository? Apart
from simply knowing that the build and unit tests succeeded, the working
assumption is that QA should be able to pick-up and test any ci-built
artifact produced by a nightly build -- but that doing so is not the
same as receiving a release candidate and initiating a formal testing
cycle.

Any general thoughts? Thanks.

Brad

Re: CI Use of Repositories

Posted by Brett Porter <br...@gmail.com>.
It sounds like a good approach to me.

On 16/04/2008, Harper, Brad <br...@epsiia.com> wrote:
> Hello:
>
>  I'm instituting continuous integration builds for a collection of maven
>  projects.
>
>  Under interactive builds, developers may deploy artifacts to a SNAPSHOT
>  repo. We may also run deploy:deploy to an internal remote repo [to share
>  non-SNAPSNOT baselines] and release:perform-release to share artifacts
>  with QA [from that same internal remote repo] for formal evaluations
>  resulting in a final release.
>
>  Artifacts under formal QA evaluation are submitted as a series of
>  release candidates [via the internal remote repo], with specific maven
>  <version> values, e.g. 1.2.3-RC-1, 1.2.3-RC-2, etc., until approved.
>
>  The question is ...
>
>  Should ci-built artifacts be deployed into a separate repository? Apart
>  from simply knowing that the build and unit tests succeeded, the working
>  assumption is that QA should be able to pick-up and test any ci-built
>  artifact produced by a nightly build -- but that doing so is not the
>  same as receiving a release candidate and initiating a formal testing
>  cycle.
>
>  Any general thoughts? Thanks.
>
>
>  Brad
>


-- 
Brett Porter
Blog: http://blogs.exist.com/bporter/