On Mon, 2016-07-18 at 20:57 +0200, Ondřej Súkup wrote:
The kernel is special: multiversion(kernel) exists for a reason, and multiversion(kernel) appears to be non-functional in this specific case. That was my point. Maybe I missed something.
yes , you missed multiversin IS between version number not release, so on correctly set system are always installed two kernels .. for Tumbleweed now: i | kernel-default | balíček | 4.6.2-1.6 | x86_64 | (system packages) i | kernel-default | balíček | 4.6.3-1.3 | x86_64 | openSUSE- Tumbleweed-Oss
You're right, I assumed that the "build number only" update scenario could cause the user to end up with only one (in the worst case, broken) kernel installed. Looking at the docs, I don't know if that would be possible. On my system, as you already said, when installing 4.6.3-1.3, "zypper dup" would remove 4.6.3-1.1 but keep 4.6.2-1.6, so I'd be fine anyway. IOW, my conclusions were premature. Thanks for the explanation. Martin
risk of regression is order lower than problems between keyboard and chair
On 18 July 2016 at 20:37, Martin Wilck
wrote: On Mon, 2016-07-18 at 20:23 +0200, Martin Pluskal wrote:
On 18.07.2016 16:33, Martin Wilck wrote:
On Mon, 2016-07-18 at 13:47 +0200, Ondřej Súkup wrote:
My conclusion would be that users would be well-advised to skip these
updates.
your conclusion is wrong
would you mind to give a reason?
Martin
Well I can't speak for Ondrej, but your conclusion is based on avoiding hypothetical regressions, which do not seem to be bound to kernel rebuilds, most of your arguments against installation of such kernel updates could be said for any package that is important for booting, yet they are rebuilt and published.
The kernel is special: multiversion(kernel) exists for a reason, and multiversion(kernel) appears to be non-functional in this specific case. That was my point. Maybe I missed something.
Perhaps you might focus more on arguments supporting you advice or refrain from giving such advices in case of lack of said arguments.
I thought I had presented arguments. You don't have to be convinced. It's true that the risk of a regression is so low that TW users (who live on the edge anyway) don't need to worry.
Martin
-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org