summaryrefslogtreecommitdiff
path: root/development/README
authorzecke <zecke>2004-05-17 21:15:42 (UTC)
committer zecke <zecke>2004-05-17 21:15:42 (UTC)
commit598c9bc76840120fa3efdb000461bae2c1fef639 (patch) (side-by-side diff)
tree4cf0c66149f7bee38a2bcface9e7fbfd0d28b85a /development/README
parent1827ce23d0719a22c14613dc5859093818da1d0a (diff)
downloadopie-598c9bc76840120fa3efdb000461bae2c1fef639.zip
opie-598c9bc76840120fa3efdb000461bae2c1fef639.tar.gz
opie-598c9bc76840120fa3efdb000461bae2c1fef639.tar.bz2
ich@opiezilla:~/programming/opie/head/opie$ nm lib/libopiecore2.so | grep polish
U _ZN14QPEApplication6polishEP7QWidget ich@opiezilla:~/programming/opie/head/opie$ nm lib/libopiecore2.so | grep polish U _ZN12QApplication6polishEP7QWidget in qt_override we had to overwrite the Palette for some widgets for some styles (setting no background liquid and such) we overwrote the polish method. As we did not inherit from QPEApplication the 'polish' symbol was only internal to libqpe and this way Opie apps worked on Opies and Sharps libqpe, and Sharp apps work on our libqpe. Now with libopiecore the compiler tries to include different symbols as shown above. So for now we could disable the legacy palette polishing which shouldn't hurt anyway. OApplication is the source of all evil, we wouldn't 'pull' in the polish symbol... but having a 'shadow' weak symbol as backup isn't good as well. Chicken you may enable the option in config.in...
Diffstat (limited to 'development/README') (more/less context) (show whitespace changes)
0 files changed, 0 insertions, 0 deletions