RFR: jmod remove
jonathan.gibbons at oracle.com
Thu Apr 19 07:11:51 PDT 2012
On 04/19/2012 01:47 AM, Chris Hegarty wrote:
> I heard the following expression a few times, 'rm -rf has reached the
> limit of its usefulness'! So it's about time we had a 'jmod remove'.
> The webrev is based upon Mandy's alias support patch . I figured it
> would most likely be going in after the alias support.
> Essentially, you can remove any module not required by a
> configuration. Fails if you try to remove a module which will break a
> configuration, unless you specify --force, or remove both the root
> module along with the dependent module.
> I added a comment about the parent/child relationship. That is, a
> module library has no notion of its children so we could remove a
> module in a parent library that is not part of any configuration in
> that library, but may be part of a configuration in a child library. I
> think this is fine for now, but we may have to revisit the
> parent/child relationship in the future ( beyond the scope of jmod
> remove ).
Does it work to leave behind a small zombie that can be detected by the
child when it is next used, so that the child can invalidate any modules
it contains which would otherwise depend on the now-deleted module.
More information about the jigsaw-dev