On Wed, Aug 04, 2004 at 10:51:55PM +0100, Colin Watson wrote: > On Wed, Aug 04, 2004 at 10:10:21PM +0200, Sven Luther wrote: > > On Wed, Aug 04, 2004 at 03:08:54PM -0400, Joey Hess wrote: > > > Why should d-i care about the mouse? Shouldn't you be patching X's > > > configuration program? > > > > Well, this is the modules which will be added to the /etc/modules. I > > understand that it may well be possible that the X installation stuff, > > but i suppose the mouse should be usefull even for gpm or other non-X > > stuff, and having no mouse driver module loaded after the reboot is > > not a nice thing. > > > > Notice that i asked about this (altough addressed to Kamion) on irc > > before making the change while you where there, so you could have > > remarqued on it before i go forward and implemented this. > > If you want to discuss a design decision, this mailing list would be far > better than the transient and overfilled medium of IRC.
Ok. Checking in code is obviously also a good way to get it discussed :). > > So, if this is not the right place for registering this module for > > future use, what is ? discover, which has trouble handling more than > > one module per pci id ? hotplug ? The X configuration stuff ? i have > > some doubt about Branden accepting this kind of stuff though, unless > > there is a generic module loading mechanism for X, and this doesn't > > solve the case of non-X mouse users. > > mdetect seems like clearly the right place. Its package description and > its man page both suggest that gpm could easily be modified to use it > (indeed, mdetect's default behaviour is to produce output suitable for > configuring gpm). And is it part of base, and sure to be run if the user does a clean install and chose the X task in tasksel ? Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]