remake wrlib rule to not compile with -fomit-frame-pointer and stuff if its added to CFLAGS Do ASAP: ======== - fix bestvisual selection code. Broken. - fix RemakeStackList() to account for transient windows - blink border of clients with UrgencyHint set between red and black - finish session stuff - add multiline support for balloons - remove colors_per_channel stuff from wmaker after? Need to do: =========== - allow user to select/restore default root menu from wprefs - support for X11R6.4 extension for getting extra visual info in wrlib's automatic best context guessing - docklet to control AccessX (keyboard accessibility) functions - rewrite all redundant stuff to use WINGs - add function to directly make a thumbnail of an image, using the functionality provided by the image libraries to load a minimal amount of data. - rewrite defaults/wdefaults stuff to use WINGs UD stuff. Search list: ~/G/D/WindowMaker /u/l/s/W/D/WindowMaker built-in-defaults - remake internal string processing to use wchar? unicode? - add new file for stuff like default commands and dnd commands for docked apps, balloons for the dock etc - check whether apps with name.class set to empty strings should be treated like if name.class is NULL.NULL - review the defaults handling code (not to reread after we update a domain if possible, check WINGs apps updating WMWindowAttributes after start making screen to flash on update) Maybe some day: =============== - virtual workspace - optimize for size Never: (so, dont even bother to ask) ====== - different themes for each workspace. Unless you give us a SGI/Power Onyx with 2 CPUs ;). - anything that requires the mouse pointer to be jumped by WindowMaker to somewhere. This is *terrible* behaviour. And it's not just IMO. - rewrite to use Gtk... I wont even bother to explain why... - ICCCM 2.0: ICCCM 2.0 (not 1.0, which is what everybody supports so so) is a relatively new standard and nobody, AFAIK, complies with it (not even twm as people tend to think). It has some neat things, but many of the new stuff is really weird and tricky to implement, not to say unworthy (read the specs and you'll see). This is not bad, since I think it is very unlikely that a client that requires it exists... Anyway, if we get an "official" sample implementation (twm?) it might be supported. Maybe dtwm supports it? I dont know...