Comment 17 for bug 1847105

Thanks Daniel,
yes that is the patch I was referring to.

If you really think reverting it in libvirt is the right path I'm fine following you on that and keep virt-manager untouched in that regard.
But as you say it is in the field for quite a while and other than this report I haven't seen any.
Therefore I wasn't going to suggest a revert, after that much time it almost is a semantic change "again".

Furthermore the reasoning to add it back in [1] was with virt-inst / virt -manager in mind and the exact definition in [2] IMHO is a bit weak for this particular case.
That was the reason I asked for guidance from virt-managers POV first.

If we end up reverting the change we might consider modifying the text in [2] to be more clear what is (expected) to happen if allocation==capacity.

[1]: https://www.redhat.com/archives/libvir-list/2018-April/msg00130.html
[2]: https://libvirt.org/formatstorage.html#StorageVol