"linux" apport package hook

[Brian Murray]

On Thu, Feb 13, 2020 at 01:25:59AM +0800, You-Sheng Yang wrote:
> How about let each kernel package install their own symbolic link? That
> would probably take many duplicated works for each variant (perhaps just
> generate it from a common script), but it's a more clean and formal way to
> maintain package hooks, and then apport source package can then be freed
> from maintaining an endless list that is really not related to itself.

Many source packages do provide their own package hooks and I don't
recall why apport actually installs the one for the kernel. I think this
would be a better approach overall but in the interim plan on uploading
apport with the additional symlinks. Is there anything wrong with the
ones I've created?

Thanks,

Brian

> You-Sheng
> 
> Seth Forshee <seth.forshee at canonical.com> 於 2020年2月13日 週四 01:06 寫道：
> 
> > On Mon, Feb 10, 2020 at 04:38:33PM -0800, Brian Murray wrote:
> > > I'm working on fixing LP: #1861446 where the apport package hooks
> > > (shipped via apport) are not run for some linux packages. I've currently
> > > added in symlinks for linux-signed-5.4, linux-signed-oem-5.4, linux-5.4,
> > > and linux-oem-5.4[1]. Am I missing anything else?
> >
> > What do we currently do for derivatves, like cloud kernels, linux-kvm,
> > linux-raspi2, and hwe kernels? Those are all likely to have the same
> > problem.
> >
> > This is also going to continue to pop up as we move upstream kernel
> > versions and package names change. I wonder if it might be worth
> > considering some more generalized way of doing these mappings for kernel
> > packages, instead of creating dozens of symlinks.
> >
> > Thanks,
> > Seth
> >
> > --
> > kernel-team mailing list
> > kernel-team at lists.ubuntu.com
> > https://lists.ubuntu.com/mailman/listinfo/kernel-team
> >
--
Brian Murray