On Wed, Jul 30, 2008 at 01:49:50PM -0500, Charles Duffy wrote:
Daniel Veillard wrote:
> Are you disagreeing with the message (which your patch doesn't fix)
>or with the semantic of the check (and then why allow to create a domain
>reusing the UUID of another defined but not running domain, I can only
>see confusion or security problems in doing so)
The act of creating a domain reusing the UUID of another defined but not
running domain is presumably an action taken with intent to rename that
other domain (and thus should undefine the prior domain holding that
UUID in the process... which, admittedly, the given patch didn't
explicitly do).
Anyhow, two patches attached; one revises the error messages, while the
other improves behavior with the semantics I assumed, to explicitly
undefine any duplicates found during a create; I'm happy with either.
We shouldn't have 'create' automatically undefine the duplicate - we have
a explicit undefine method apps can already use if they want that to be
done, without needing to encode this policy in the create method. I'm
fine with us applying the 2nd patch to fix the error message.
Daniel
--
|: Red Hat, Engineering, London -o-
http://people.redhat.com/berrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org -o-
http://ovirt.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|