Michel Dänzer wrote:
Right, but macintosh and pc101 caused exactly the same pattern of breakage, just about the same mappings ("`1234" -> "merty") as far as I could tell.On Tue, 2001-09-11 at 00:08, Adam C Powell IV wrote: [ broken keymap with XFree86 4.1.0 ]I tried both "macintosh" and "pc101" in dpkg-reconfigure xserver-xfree86, both of which should give the correct mappings for the new input layer (which I've been using since it first arrived in kernel-image-2.2.18pre21-pmac), right?"macintosh" should for a Mac keyboard, the "pcxxx" XkbModels are for PC keyboards.
Okay, the keyboard works now (see below), I'll find out if the mouse problem is solved soon, by logging out and back in...Second problem: using gdm and gnome-session, the mouse stops working after the first logout, that is, the mouse is frozen at the second gdm login screen. I observed this behavior also on an i386 system freshly installed with X 4.1.0 three weeks ago (but not on previously-installed continuously-upgraded systems), and switched to /dev/gpmdata, which solved it for some time, then it broke again, so I removed gpm. Unfortunately, there is no /dev/gpmdata on PPC,Of course there is, you just have to configure gpm accordingly. It shouldn't be necessary though as /dev/input/mice works with any number of processes using it.
As we say in New York, "Akhaa!" The file exists and contains zero. Replacing it with one solved the problem.Ethan Benson wrote:switch the kernel to use linux keycodes.Done that (long since).Are you sure? /proc/sys/dev/mac_hid/keyboard_sends_linux_keycodes doesn't exist or contains 1?
But wait, didn't we all switch to Linux keycodes way back in December when 2.2.18pre21 hit potato? We had to use Linux keycodes for button emulation, shouldn't that have meant that we were using Linux keycodes?
And is it not a bug that with zero in that /proc entry, the kernel uses two different keymaps for the two functions (button emulation and normal keyboard use)?
Not that this means anything, nor that I'm any expert, but if I was this confused, how about a newbie? Should there be some kind of auto-detection of HID and keyboard_sends_linux_keycodes in the .config and/or .postinst of the xserver, to detect and advise or correct such a situation? Or are we all comfortable with just letting lots of users get frustrated and confused, as with last December's change?
Thanks again for the help, but still somewhat confused, -- -Adam P. GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6Welcome to the best software in the world today cafe! <http://lyre.mit.edu/%7Epowell/The_Best_Stuff_In_The_World_Today_Cafe.ogg>