LGTM!
Seems to work. Exif data read/write functions as expected, too.
Works for me on Carbon X1 Gen 8.
Confirming that this fixes the conflict bug. Cannot speak to the gnome-keyring issue that the original package update here was for; instructions on testing that would be helpful.
Works. Did not validate that the exploit is fixed.
This update seems to work, and in fact was required to get the latest firmware update from Lenovo installed on my Thinkpad X1 Carbon Gen 8.
So, FWIW: I have a F31 system where I just applied the Firefox update and didn't apply a bunch of other updates, and I was getting just white pages. Updating everything else (except the kernel, as I didn't reboot) seems to make it work fine. Maybe some missing dep?
Got bitten by this badly. This fix seems to fix it, as confirmed by not seeing hundreds of zombie processes after upgrading to it. Thanks!
I hit bug #1631970. Verifying that this update resolves it.
This change seems out of line with what I understood to be the plan of record.
@whot Yeah. As soon as you touch it, it jumps to the top left corner, and no manipulation of the pointer causes other movement. However, using the Logitech mouse works fine.
It's reproducable on multiple systems (by plugging in the USB keyboard). On one, I actually had three mice: there's one on the keyboard of the laptop in the docking station plus the external-keyboard mouse plus an additional Logitech mouse I plugged in to debug. Only the USB Keyboard Mouse exhibited the problem. I'll file a bug but it won't be until tomorrow evening as I have a lot of meetings during the day.
Also note that the trackpoint device built into the laptop works just fine. It's only the external device that's affected.
Yeah, 1.10.902-1 has the same failure. Note that evtest
seems to show completely reasonable
type 2 (EV_REL), code 1 (REL_Y), value -12
type 2 (EV_REL), code 0 (REL_X), value 1
events, even as the cursor fails to move.
I have an external keyboard, Lenovo ThinkPad Compact USB Keyboard with TrackPoint, and with libinput-1.10.901-1.fc28, merely touching the pointer makes the cursor jump to the top left and not move. Testing now with 1.10.902-1.fc28
Oh, wow! That's a big change. We've made big changes before, but I don't want to push people into "hero testing". My suggestion at this point is to revert the F40 backgrounds, and make the new one the F41 background -- as rawhide is F41 currently anyway.
But then, start on the F42 backgrounds soonish, with the goal of landing them in rawhide shortly after 2024-08-06 (that is, this August!). And so on for F43 in February 2025 (before the F42 release!). That way, a) Rawhide always (give or take a few days) has different wallpaper from the branch and release and b) we're never at risk of getting close to the release without the backgrounds being ready to go.