Created attachment 469662 [details] Xorg.conf As the title suggests, I'm deploying Fedora on machines with Microsoft Media Desktop 1000 keyboard / mouse wireless combo. On all machines I get uneven mouse acceleration and especially broken mouse-wheel behavior (E.g. uneven scroll speed, sometimes slow, sometimes fast, sometimes backwards, etc). Needless to say, according to the users, Windows 7 (ugh) doesn't produce the same results. I should point out that Xorg was very good at detecting the keyboard, including the media buttons. Xorg.conf and Xorg.log attached. P.S. machines are using the proprietary nVidia driver. I can test w/o it, if it matters. - Gilboa
Created attachment 469664 [details] Xorg.0.log
Created attachment 469665 [details] Xorg.conf (right mime type)
Created attachment 469667 [details] Xorg.0.log (right mime type)
$ xinput list ⎡ Virtual core pointer id=2 [master pointer (3)] ⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)] ⎜ ↳ Microsft Microsoft Wireless Desktop Receiver 3.1 id=9 [slave pointer (2)] ⎣ Virtual core keyboard id=3 [master keyboard (2)] ↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)] ↳ Power Button id=6 [slave keyboard (3)] ↳ Power Button id=7 [slave keyboard (3)] ↳ Microsft Microsoft Wireless Desktop Receiver 3.1 id=8 [slave keyboard (3)]
Looking at https://fedoraproject.org/wiki/Xorg/Input_Triage_Algorithm and discussing this with our developers, I believe that the most important information which we could use is the file generated by the evtest-capture command (from the package evtest) against the devices. Please, generate it and attach to this issue report. Thank you
Created attachment 470249 [details] evtest-capture outpuit I tried to record broken behavior (E.g. even reversed mouse-wheel, weird uneven acceleration, etc). If you want me to record certain evens (to clean up the log) please let me know. - Gilboa
In the case of the mouse-wheel, I should point out that if I -slowly- turn it, I get more-or-less working scroll. If I increase the wheel speed, I get no scroll, or even reverse scroll. - Gilboa
Created attachment 470269 [details] uncompressed evtest-capture.xml
(In reply to comment #6) > If you want me to record certain evens (to clean up the log) please let me > know. yes please, it's hard to know what I'm looking for. right now I see the mouse move and scroll my window, but I don't know what exactly I'm looking for. as for different behaviour: in situations where scrolling causes an application to render large parts, it may seem slower. e.g. scrolling in firefox will feel slower than scrolling in a terminal. this isn't something we can fix in the driver though. as for the reverse scroll: do you see the reverse scroll in evtest and/or in xev?
Created attachment 472443 [details] Slow and fast scroll. OK. I tried to record specific events. I've opened up firefox in a test-only site that includes ~3 pages worth of text. I recorded the mouse events twice, once while slowly turning the mouse-wheel (evtest-capture-slow.xml) from the top of the page to the bottom and than back, and once while turning the mouse-wheel at a what I consider normal rate (evtest-capture-fast.xml), again top to bottom and than back. In both cases, scrolling was pretty slow and uneven, as if large number of events simply disappeared. During the fast scroll test, I noticed at least once reverse scroll event. I should add that I'm conducting this test on a dual Xeon machine with nVidia 9400 GPU w/ latest proprietary drivers. I doubt that this problem is rending related. (Scrolling using PageUp/Dn seems clean of any type of artifacts). - Gilboa
P.S. Can I somehow record pure USB HID events?
This bug was filed against Fedora 14 which is now EOL. Please re-open this bug if you still experience this issue with one of the currently suppported versions of Fedora. Don't forget to update the version field if you do so.