Bug#708430: Increasing severity

[Svante Signell]

On Fri, 2013-12-06 at 10:42 +0000, Colin Watson wrote:
> Control: severity -1 important
> 
> On Fri, Dec 06, 2013 at 10:58:59AM +0100, Svante Signell wrote:
> > severity 708430 serious
> 
> I'm sorry that I haven't dealt with this yet - I do intend to, I hope
> before Christmas - but please stop fiddling with the severity on this
> bug.  

It was filed a long time ago, 15 May 2013, and importance increased to
grave by me in August and lowered to important recently by Philip. I
increased the severity now to get any kind of response, in frustration,
sorry for that. There has been many releases since it was filed the
first time. Why couldn't you have given this reply to the bug report
several months ago? (and maybe added a README to the releases, how to
workaround the problem until a proper solution is found?) 
 
> Trying to force the issue with the severity doesn't impress me,
> and this bug requires a good deal of care in order to avoid ending up in
> an arms race where different OSes all try to install to the fallback
> path in order to avoid firmware bugs and we end up right back in the bad
> old days where installing one OS breaks another.

Understood!

> Note that the requested behaviour explicitly contravenes the UEFI
> specification, and for good reason.  I agreed with Steve at the
> mini-conference in Cambridge recently that we probably have to do it
> anyway to cope with broken firmware, but that does not make an RC bug.

We have at least two manufacturers having a broken UEFI implementation:
Toshiba and Acer, maybe there are more. So this problem has to be
solved, unfortunately not by requesting users to upgrade their firmware
(if available).