I reported this on irc, but it may get lost there.
Touchpads in X and SDL fullscreen apps have a problem with touchpad cursor movement if their mouse movement is not relative (like a shooter) but absolute (like scummvm).
https://answers.launchpad.net/ubuntu/+s ... ion/200823
Here is the fix for exult:
http://bazaar.launchpad.net/~exult-team ... ision/6035
Dosbox, gemrb, pentagram and probably many others are affected, very unwise default by SDL (or api or whatever).
Small bug with touchpad
Moderator: ScummVM Team
-
- Got a warning
- Posts: 173
- Joined: Thu Feb 25, 2010 7:44 am
- dreammaster
- ScummVM Developer
- Posts: 554
- Joined: Fri Nov 04, 2005 2:16 am
- Location: San Jose, California, USA
Re: Small bug with touchpad
Forum rule 3a. Please post bug reports to the bug tracker, not the forum.Serious Callers Only wrote:I reported this on irc, but it may get lost there.
If all the listd software are being affected, is it a problem which each one's use of SDL, or is it a legitimate bug in SDL? If the latter, it would be better to press to have it properly fixed in the ofificial SDL distribution and backported to our SDL copy, rather than trying to hack in a workaround.Serious Callers Only wrote:Dosbox, gemrb, pentagram and probably many others are affected, very unwise default by SDL (or api or whatever).
-
- Got a warning
- Posts: 173
- Joined: Thu Feb 25, 2010 7:44 am
I think it's more of a 'backwards compatibility problem'.
For some reason they choose to turn fullscreen mouse movement with hidden cursor movement relative, probably at the dawn of time. I don't understand why, but it's probably some misguided notion that 'hiding cursor -> fps'. Now they're stuck with it.
To be fair, i think that they had no notion that relative movement updates vs absolute ones would behave so differently. Also, probably the windows drivers are more well behaved.
Also; i lied, i didn't test pentagram and gemrb. Just exult, dosbox and scummvm. But it makes sense they're affected unless they already workaround it.
For some reason they choose to turn fullscreen mouse movement with hidden cursor movement relative, probably at the dawn of time. I don't understand why, but it's probably some misguided notion that 'hiding cursor -> fps'. Now they're stuck with it.
To be fair, i think that they had no notion that relative movement updates vs absolute ones would behave so differently. Also, probably the windows drivers are more well behaved.
Also; i lied, i didn't test pentagram and gemrb. Just exult, dosbox and scummvm. But it makes sense they're affected unless they already workaround it.
-
- Got a warning
- Posts: 173
- Joined: Thu Feb 25, 2010 7:44 am