[ please keep bug report on cc, cool thanks :) ] On Wed, Jan 30, 2008 at 11:29:34AM +0100, Alain Kalker wrote: > On Wed, 2008-01-30 at 10:39 +0100, maximilian attems wrote: > > > second the kernel handbug is currently quite out of outdate. > > I beg your pardon? Oh, I understand. But the kernel handbook is what > users (hopefully) will refer to before submitting a bug report. So your > reaction might come as a bit of a surprise when they think they're doing > things "by the book".
yep typo, well i don't know you are the first one to refer to it, it's usage is not soo wide spread.. > > third if the issue affects bootparam we won't make changes > > in debian just for fun, that needs to be upstreamed see > > http://wiki.debian.org/DebianKernelPatchAcceptanceGuidelines > > http://wiki.debian.org/DebianKernelReportingBugs > > The first link deals with kernel patches (which I would gladly make if I > had time) and specifically encourages driver fixes. The second one > mentions the "bts forwarded" command which was entirely new to me. > Thanks for the tip. I would be glad if this was documented more > prominently when reporting kernel bugs. both pages are linked prominently from http://wiki.debian.org/DebianKernel but yes we could add them as preamble to furthure reportbug scripts. also the second page tells to report upstream! ;) > > yeah right and people have it working right now oldstyle. > > irc initramfs-tools has some patches for this i810 either > > merged or in the bts. > > How does that rhyme with your comment above about not changing Debian > because of bootparam handling? IMHO the problem should be fixed as close > to the source (sorry for the pun) as possible, which in this case is the > kernel. Having a patch in there also makes it easier to synchronize with > upstream, whereas having a workaround in a different package > (initramfs-tools) will probably lead to some nasty interactions (and > new, difficult to track down bug reports) when the problem is eventually > fixed in the kernel. > > -Alain no contradiction here, all i'm saying is that current behaviour works for people who know about it. but indeed the fix belongs directly in the kernel. regards -- maks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
