Hamish Moffatt wrote: > On Sat, Jul 16, 2005 at 12:33:12PM +0200, Joerg Rossdeutscher wrote: > >>Am Samstag, den 16.07.2005, 17:22 +1000 schrieb Hamish Moffatt: >> >>>I can't imagine how it's a good idea to upload a package which depends >>>on a kernel not available for Debian yet. >> >>Packages should not depend on any kernel, since many people run their >>own. However, I just don't understand why the package has been published >>at all, since even in experimental there is no 2.6.12. > > > Yes I mean depend as in need, not as in Depends: in the package control > file. The new udev needs a new kernel. Most users don't have that > kernel. It does not seem logical to upload the new udev then. > > Even if you build your own you should use make-kpkg.
I suppose that's true... For whatever reason, I've never been able to get into the habit of make-kpkg on my individual workstation. It's fantastic for groups of production machines and any situation where consistency is warranted. I wonder if it wouldn't be easier for stock kernel users if there were a convenience script that used equivs to satisfy the kernel dependency, Debian-style, as it were. Given something like that was available and easy to use, from a "distribution integrity" perspective an explicit Depends: on a kenrel version doesn't seem like such a bad thing for situations like udev. Of course, it only solves the coarse-grained "this version is required" problem, and not any finer-grained things like "the Debian-patched version of the kernel is required." tony -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
