Frank Fürst wrote: > Junichi Uekawa <[EMAIL PROTECTED]> schrieb: > > >>I recommend filing a bug against such package, but I could not >>remember which package offhand. >> >>It could be somewhere along base-files or debootstrap. >> > > Isn't this the postinst-Skript of makedev? > > | cd /dev > | > | /sbin/MAKEDEV std > | /sbin/MAKEDEV consoleonly > | > | test -e /dev/tty0 || /sbin/MAKEDEV tty0 >
OK, I digged through various files from the package "makedev" and bugs #62222 and #89932 thanks to your hint. I think I finally found out what the problem is: "MAKEDEV input" was never run on my fresh woody installation.
Therefore the corresponding entry in /etc/X11/XF86Config-4 pointed to the non-existant device /dev/input/mice.
After I ran "MAKEDEV input" manually, the correct devices showed up in /dev/input.
So either this has to be run in the case statement "std" (see above code snipplet) which generates standard devices (as far as I understand) or at some other appropriate place in this file. I don't know what other architectures need, but from practice I know that on i386 you absolutely need /dev/input/mice for USB mice in XFree86.
One problem's still there, though. If this "bug" would be corrected, would it make into Woody? (excuse this question but I don't yet fully understand the unstable/testing release procedure nor am I up-to-date with the current Woody freeze status...)
FYI, I didn't file a bug report yet. Beat
