diff options
| author | phintuka <phintuka> | 2012-02-17 07:44:25 +0000 |
|---|---|---|
| committer | phintuka <phintuka> | 2012-02-17 07:44:25 +0000 |
| commit | 693a9595e145d39b3375a39733331924455941c7 (patch) | |
| tree | faa265021b6f3450285230e812f08f0c54a33bb1 /tools | |
| parent | e4c44b89a46f42def6c22d495d7f3ac50af1617b (diff) | |
| download | xineliboutput-693a9595e145d39b3375a39733331924455941c7.tar.gz xineliboutput-693a9595e145d39b3375a39733331924455941c7.tar.bz2 | |
Fixed MWMHints struct in 64-bit systems.
(Thanks to Roland Scheidegger)
there's a problem with the MWMHints definition in vdr-xineliboutput. It
does not work on x86_64 at least with metacity and kwin, as these use a
different definition (hence once borders are toggled off they are gone
for good).
This is a strange issue there seems to be no reference source what the
definition of this struct should be, some quick search finds 3
incompatible versions (on 64bit archs) and a dozen or so more compatible
ones (only differing in signedness).
In any case, metacity, kwin agree, and this is also what mplayer uses -
xine though itself also uses the same version as xineliboutput which
doesn't work with at least metacity/kwin. It's quite likely some obscure
wm out there expects one of the other definitions (I think fvwm is same
as metacity, while enlightenment and afterstep need the all-32bit
definition). What a mess.
Anyway this patch just changes the definition to what's more likely to work.
Diffstat (limited to 'tools')
0 files changed, 0 insertions, 0 deletions
