http://www.movement.uklinux.net/q2.diff.bz2
Mostly cleanups over the other one.
Some short description of what's been done so far
I think this is a really good opportunity to kill some old cruft
and update stuff to the new way of doing things (boost usage etc.).
PainterBase renamed Painter to fit with source usage. Painter
renamed as XPainter. Both moved into frontends, some cleanup (e.g.
pixmap() now in XPainter only)
LyXView/XFormsView moved, some minor cleanups. Currently createWorkArea
is a hack, I'm not sure where it should go...
WorkArea became a virtual abstract class, and XWorkArea derives from it.
No real meaningful changes except that LyXScreen (now LScreen) became
owned by WorkArea. This is logical and correct IMHO.
The events were changed to use GUII types for the mouse buttons etc (see
event_states.h)
FontLoader got split into a GUII part (containing update() and
available()) and a private part. It became owned by LyXView, so it
can communicate to its minibuffer (this still leaves an implementation
able to share global resources for the real font loader e.g. in xforms/)
lyxlookup has just moved
lyxfont.h got split into GUII and non-GUII part. Currently the GUII part
just takes LyXFonts + string const & and returns an int. xfont_metrics.h
in xforms/ does the XChar2b thing which I understand we want to change.
LScreen now owned by WorkArea but still accessed by BV_pimpl. Also
virtualized for GUIIzation. Common code moved back into frontends/ since
last patch.
ColorHandler, FontInfo, made xforms-private.
TODO
----
1. finish key handling to allow triple of
<LyXKeySym, string const & text, string::size_type count>
and get it working.
LyXKeySym - I think we just borrow the X keysymdefs, and force the
other toolkits to convert to those ones.
2. clean ups
3. unbreak all the stuff I've broken. Keyboard is horked, so is mouse
press for some reason ...
4. the current APIs have lots of strange geometry management and other
stuff going on that isn't really right IMO
5. once keyboard + scrollbar + mouse is reasonably sensible, I can make
a proof of concept Qt frontend. Another issue we have to look at is
what exactly an x,y pair means when passed from wa to bv_pimpl ...
6. Angus' frontends/X11/
thanks
john
--
I am a complete moron for forgetting about endianness. May I be
forever marked as such.
