Why not be consistent with http://libvirt.org/goals.html. I'm really agreed with use the concept hypervisor. On Tue, Jun 22, 2010 at 6:09 PM, Eric Blake <eblake@redhat.com> wrote:
On 06/22/2010 03:48 PM, Justin Clift wrote:
On 06/23/2010 04:16 AM, Matthias Bolte wrote:
case VIR_ERR_NO_SUPPORT:
if (info == NULL)
Would it be practical to expand this with category messages like?
case VIR_ERR_NO_SUPPORT_STORAGE: if (info == NULL) errmsg = _("this function is not supported by the storage driver"); ...
case VIR_ERR_NO_SUPPORT_NETWORK: if (info == NULL) errmsg = _("this function is not supported by the network driver"); ...
Or maybe figure out a "%s" replaceable keyword for network/storage/hypervisor/etc above?
The whole error reporting scheme needs some loving care, and I've got a partial patch series posted to the list earlier that started that process. But it was deemed less important than getting 0.8.2 finalized (as well as helping with RHEL-6 backports), so it's been on the back burner. At this point, a minimally-invasive patch, like Daniel's proposal, is better than making things better when the code is all going to be refactored down the road anyways. But thanks for the thought; it may affect my refactoring efforts!
-- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list