[Bug 189400] Review Request: em8300(-kmod) - Hollywood+/DXR3 hardware MPEG decoder drivers and tools

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug report.

Summary: Review Request: em8300(-kmod) - Hollywood+/DXR3 hardware MPEG decoder drivers and tools


https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189400





------- Additional Comments From fedora at leemhuis.info  2006-07-23 10:46 EST -------
(In reply to comment #9)
> If the kernel module has a suitable licence for FC5, then it 
> presumably has a suitable licence for being included directly in the 
> kernel.

Agreed.

> Thus, the correct approach would be to get the support into the 
> upstream kernel and therefore into our base kernel package, rather 
> than having a separate kmod package for it.

Also agreed, and that's why we we plan to try our best to tell upstream
maintainers "get your stuff merged". That why we require a statement
when people plan to get their stuff merged (see comment #4).

> I strongly believe we should veto _all_ kmod packages in Core and 
> Extras.

This was discussed before when we started the development of the
kmod-standard for Extras -- several people from Red Hat were involved in
that discussion and we agreed to work on it nevertheless the political
problems.

Anyway, I can really understand your point. That why I suggested in the
beginning that all kmods are only allowed for X months (X=12 or 18) in
Extras before they get dropped again and that upstream shall work on
getting its suff into the kernel in that timeframe. But that idea was
not accepted by other involved in the discussion.

Anyway, I don't think we should discuss this further here and probably
won't reply again in this bug. fedora-devel-list probably is the better
place for this sort of political discussions.


Site note: It's IMHO more a "kernel" vs. "distributions" problem:

I'm sure all those people currently listed in the CC of this bug agree
that it's important and best to get all drivers into the kernel as fast
as possible (as long as they are "clean" and even if they are far from feature
complete ) and that external drivers have many pitfalls and should be avoided as
much as possible.

But it seems some open-source-driver-developers don't feel that way and
prefer to maintain their drivers outside of the upstream kernel. They
probably fear (for no good reason) getting "Christophed". That's a
pity, but it's life in 2006.

But some distributions include these external drivers. That creates
pressure on those distributions that don't include external drivers
because people run away. E.g.: "I like Fedora. But it doesn't include
the external driver foo -- distribution bar has it already and thus is
far easier to use for me because I need that driver. That's why I
switched over to bar."

So IMHO kernel developers and the kernel-maintainers of the most
important distributions need to work together closely and create
policies like "every distribution may only contain 10 kernel-drivers that are
not yet merged into the upstream kernel". The pressure on driver developers to
get their drivers merged into linus kernel would be a lot bigger then.

Just my 2 cent.

-- 
Configure bugmail: https://bugzilla.redhat.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.