On 10/03/2013 01:56 PM, Bill Nottingham wrote:
Adam Williamson (awilliam(a)redhat.com) said:
> On Mon, 2013-08-26 at 15:04 -0400, Bill Nottingham wrote:
>> Adam Williamson (awilliam(a)redhat.com) said:
>>>> I should remember to check common bugs, but since I follow this list
I'm
>>>> normally familiar with the existing issues. But after looking at the
>>>> entry, it wouldn't have been helpful to my case since it says there
>>>> shouldn't be any issues.
>>> Yeah. We need to fix that. The problem is that now I know it *does*
>>> cause some issues, but I still don't know *what* issues. Bill was
>>> supposed to be looking into it and updating the bug, but unless I missed
>>> a comment, he didn't get around to it yet.
>> I belive the patches posted in that bug will make the behavior better, and
>> more predictable. However, I don't maintain either component in question,
>> and it involves changing their behavior, so while I *could* just
>> provenpackager the changes in, I figure I'd wait for a go-ahead from the
>> actual maintainers.
> That's kind of orthogonal, though. We *definitely* need to fix it going
> forward, and that should take priority, but I was also hoping - and
> thought you had kindly volunteered - that we could determine with as
> much detail as possible what the exact status and consequences of the
> F19 situation were, and update the bug report and commonbugs page with
> that information.
Apologies, I haven't had time to do investigation. The note in CommonBugs is
certainly correct as far as anaconda using biosdevname by default still - if
there are other consequences where different names are written to the
interface configuration by anaconda than what are assigned with
udev/systemd, that would exacerbate things further.
Should not biosdevice names be dropped from the release now that udev
[1] has replaced it's behaviour ?
JBG
1.
http://fedoraproject.org/wiki/Features/SystemdPredictableNetworkInterface...