Noberasco Michele wrote:

On Wed, 23 Jul 2003 12:29:01 +0200

Hi Micha

Micha Nelissen <[EMAIL PROTECTED]> wrote:


Let DirectFB read from /dev/gpmdata. Then you will find that this is a
read-only device so some code has to be disabled in directfb's serial mouse driver. I wrote a patch for that long ago an posted it here,
look in the archives. That patch reads the directfbrc to see if you
want to disables those tests.

I searched the archives and found some discussion about GPM repeater, but no patch. Maybe it is time to bring the discussion about a gpm backend for directfb back to life :-) Currently qingy has to stop gpm (and reactivate it when exiting), otherwise mouse behaviour gets jumpy :-\

To clear one thing up: there is no GPM protocol. GPM can work like a image converter, from xpm to png for example. GPM can convert between mouse protocols (but not all). 'gpm -Rmsc ...' repeats into the mousesystems protocol for example, see 'gpm -t help' IIRC.

One property of this repeating is that it is readonly: you cannot write to the gpmdata pipe.

My patch simply adds option 'mouse-gpm-source' to directfbrc in which case the following things change: (otherwise it's old func)
1) use /dev/gpmdata instead of /dev/mouse
2) open /dev/gpmdata readonly instead of readwrite
3) write tests are disabled because /dev/gpmdata is readonly

Regards,

Micha.

Attachment: directfb-mouse-gpm.diff.gz
Description: application/gzip



Reply via email to