SHIFT+ALT != ALT+SHIFT
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| X.Org X server |
Invalid
|
Undecided
|
Unassigned | ||
| xkeyboard-config (Ubuntu) |
Triaged
|
Low
|
Unassigned | ||
Bug Description
Binary package hint: xorg
Apparently xorg is not treating these sequences as identical. The problem shows up (loudly) when trying to indent in a Microsoft application via rdesktop, although I'm sure you could see something similar running an X app over rdesktop.
Let me quote David Corbin from http://
With some effort I've figured out that when you type SHIFT+ALT, that doesn't
sent the same keystrokes that get sent when you type ALT-SHIFT. The first
sends "SHIFT, META", while the latter send "ALT,SHIFT".
This is very annoying to me, because the net-effect is that I end up with a
"stuck ALT key" because I do this:
ALT(down) [ sends ALT down]
SHIFT(down) [sends SHIFT down ]
ALT(up) [sends META up]
SHIFT(up) [sends META up]
If I'm really careful, I can remember to release the SHIFT key before the ALT
key and things work OK.
After googling a bit, I'm starting to thing this may be a problem wth the X11
configuration, but it's anything but clear.
Any comments on the "best" way to get this fixed?
As the followup to that message shows, there is a simple workaround:
xmodmap -e "keysym Alt_L = Alt_L"
However, I shouldn't have to rely on Google to get this working, should I?
| Changed in xorg: | |
| importance: | Undecided → Low |
| status: | New → Confirmed |
| Bryce Harrington (bryce) wrote | #1 |
| Changed in xorg-server: | |
| status: | New → Unknown |
| Bryce Harrington (bryce) wrote | #2 |
| Changed in xorg: | |
| status: | Confirmed → Incomplete |
| Changed in xorg-server: | |
| status: | Unknown → Confirmed |
| DaveAbrahams (boostpro) wrote | #3 |
| Changed in xorg: | |
| status: | Incomplete → Triaged |
| Daniel Hollocher (chogydan) wrote | #4 |
| Changed in xorg-server: | |
| importance: | Unknown → Undecided |
| status: | Confirmed → New |
| Changed in xorg-server: | |
| status: | New → Invalid |
| AlexGenaud (alexgenaud) wrote | #5 |
Hi Dave,
I've reported this bug upstream at https:/
If you could, please subscribe to that bug, so if upstream has questions or needs further information or testing, you can provide it directly. (But I suspect this may simply be a case of needing a quirk so hopefully should be easy.)