1.- I've made my easystroke scroll movement
2.- CPU usage increase to 100% for a fraction of second
3.- There are 1 or 2 seconds where nothing happens
4.- A extremely fast movement to the bottom of the window
It's not exactly the same problem I suffered with the 2.7.3 driver (now scroll have a bad behaviour occasionally, but 2.7.3 was not usable at all). However 2.6.0 driver performs perfectly, so I've decided to downgrade.
]]>Just confirmed, all works ok. You just have to build easystroke 0.6.0 from abs against the new boost.
Cheers.
The boost problem is solved on easystroke 0.6.0-2.
My 'scroll issue' is quite evident on chromium, but I haven't found it on other programs I use a lot as evince or pcmanfm. I have a 'hypothesis', I used to scroll a webpage before the images are fully loaded; I will try to be 'more patient' and wait until the page is fully loaded before I scroll to see if there is a difference.
]]>Cheers.
]]>Will try later.
]]>There a new easystroke version in the repos, I will test it (with and without mtdev driver) and let you know.
Edit: not a sucessful update https://bbs.archlinux.org/viewtopic.php?id=160642
]]>All working perfectly here.
xf86-input-evdev-2.8.0 (built without MTDEV)
easystroke-0.5.6
co-ordinate matrix script
The scroll where it doesn't stop, I am guessing this is only in Firefox? This is the only place I get this on occassion. There is a Firefox setting in about:config that can stop this, I just haven't gone into the research to find out which.
Cheers.
]]>Downgrading to 2.6.0 without mtdev is still a solution for me.
]]>I can only guess there is something else causing your lag issues.
I am using xf86-input-evdev-2.7.3-2 (built without mtdev support and using the coordinate matrix script), easystroke-0.5.6-1 and xorg-server-1.13.0-2 without any issues at all. Everything is working absolutely perfectly.
Cheers.
]]>- scroll doesn't work at all with the mtdev driver
- scroll is useless (huge lags) with the non-mtdev driver
Downgrading to 2.6.0 without mted driver solve all the issues. The only drawback is the need to rebuilt the 2.6.0 evdev driver forced by a recent update of some package (not sure which package caused, maybe xorg-server)
PD: at least the focus problem with the mtdev driver seems to dissapear
]]>I'm not having any issues at all. Maybe have top running in the background to see if anything is maxing out the cpu when the hang occurs?
Cheers.
]]>There's a new easystroke 0.5.6 version (not in the arch repository yet), I wish it solves all the issues with evdev. I will test it when I have some time, please tell me if you do it.
]]>$ xinput set-prop <DEVICE> "Coordinate Transformation Matrix" 0.5 0 0 0 0.5 0 0 0 1
$ xinput set-prop <DEVICE> "Coordinate Transformation Matrix" 1 0 0 0 1 0 0 0 1
Cheers
]]>I tried all the evdev 2.7 (no mtdev) drivers from ARM, all of them with the cursor-jumping issue. Then I tried older drivers in ARM (2.6.0-4, 2.5.0, 2.3.2) but the touchscreen were frozen. So I thought that maybe these older drivers were built with old xorg versions, so I tried to rebuild them.
I downloaded the last PKGBUILD from 2.7.3-1 from ABS (as you show me) and did the following changes (*)
- Remove the mtdev dependence
- Change pkgver=2.6.0 (and pkgrel=5, but it's not important)
- Change sha256sums=('b0e7f3991a8183a4743196c3e16d7184d439b80bf43653aa2f45b0756a6753ac'), the right value for 2.6.0 sources
I made the package, installed with pacman and... voilà, the cursor-jumping was gone!!! Moreover, I updated to esystroke 0.5.5.1 from repo and everything seems fine and the scroll is smooth. I'm really, really happy, the touchscreen problems were driving me crazy. Now all the issues are gone for me.
Please tell me if you can also solve all the issues
(*) I suppose you'll not have problems to create the pkgbuild, if you want I can post mine.
]]>