Further to the OP's report, I'm having the identical result. dpkg-reconfigure keyboard-configuration does nothing for me on two different systems. The common factor is that they both use slightly unusual keyboards.
The function ‘keyboard_present’ in keyboard-configuration.config searches /proc/bus/input/devices for very specific strings. These two keyboards contain no matches, and thus are not considered to be attached. Shouting at the computer has not appeared to help in any way. /proc/bus/input/devices data (from two different machines) for my two problem keyboards are: I: Bus=0005 Vendor=0a5c Product=8502 Version=011b N: Name="Rapoo E6700" P: Phys=b8:27:eb:0e:e8:ee S: Sysfs=/devices/platform/soc/3f201000.uart/tty/ttyAMA0/hci0/hci0:11/0005:0A5C:8502.0002/input/input2 U: Uniq=6c:5d:63:21:fd:ab H: Handlers=sysrq kbd mouse0 event1 B: PROP=0 B: EV=12001f B: KEY=1f0000 10000 2000007 ff9f387a d941d7ff febeffdf ffefffff ffffffff fffffffe B: REL=143 B: ABS=f00 0 B: MSC=10 B: LED=1f I: Bus=0011 Vendor=0002 Product=000e Version=0000 N: Name="ETPS/2 Elantech Touchpad" P: Phys=isa0060/serio1/input0 S: Sysfs=/devices/platform/i8042/serio1/input/input5 U: Uniq= H: Handlers=mouse0 event6 B: PROP=5 B: EV=b B: KEY=e420 10000 0 0 0 0 B: ABS=661800011000003 Suggestions: 1. Can the OP please post the contents of their /proc/bus/input/devices file here to see if it would trigger keyboard_present? 2. Would the maintainer(s) consider revisiting the check in keyboard_present to include more diversity? This is likely hard, and may break something, somewhere for someone who was relying on it. 3. Are there alternative methods (such as adding to a startup rule file) that one could alter these device names so that they pass the keyboard_present test? Yours truly, Stewart C. Russell Toronto, Canada.
signature.asc
Description: OpenPGP digital signature