On 3 Jun 2007, at 11:39, Ian MacArthur wrote: > In case anyone else is looking at it, I now have a hacked-up > version of OksiD's 1.1.6-utf code that works satisfactorily with > the Xft implementation (backported from 1.1.8). > There are still a few loose ends (mainly around getting xft to work > with OksiD's xim implementation) that I have to tie up, but seems > to be working so far, and looks better on-screen than the non-xft > verison, at least to my eyes. > > If I get a chance (time, as ever being the problem) I *might* try > and merge that into a 1.1.8 snapshot. But don't hold your breath > waiting on me...!
OK, I managed to find a few hours to take a run at this today, so I now have a "working" utf+xft version of 1.1.8 for linux. I'll post a snapshot when I get a chance, but I'm having ISP problems this weekend so can't post stuff on my site... Urgh... Anyway - Known issues are: - OSX not touched at all so far. - Win32, I've edited the code, but have not attempted to build it yet, so I assume it's broken. - The gl font handling is flaky (but it kinda always was with xft anyway. We need a glXUseXftFont method really, but that doesn't seem to exist...I could try and write one, but time...) - The xim stuff is still needing work - I've somehow managed to break the cursor and selection handling in the Fl_Input_ derived widgets - The fractals demo segfaults for some reason I haven't investigated yet, although the other gl demos seem to be OK - I'm pretty sure I've broken the drag'n'drop stuff - OksiD had modified the 1.1.6 code, and the 1.1.8 code had been changed again in an incompatible way. I currently have some sort of Frankenstein hybrid... - I doubt it will build *without* xft now, as I have not tested that yet - To ease i18n of the code, OksiD had changed 1.1.6 so the "fixed strings" in the messages use char arrays held in Fl::ok_txt etc, whereas 1.1.8 had changed it so that the "same" strings are held in global char arrays called fl_ok etc. I have changed *most* of the code to use the strings in the OksiD style, but have probably missed lots... Again hybrid mess. - The code is now really badly formatted from all my random cut'n'paste... - It was tedious! Anyway, I'll try and do a bit more if I get the time, and look at Mariwan's stuff, and maybe try and make sense of the equivalent fltk-2 sections. Time... Cheers, -- Ian _______________________________________________ fltk mailing list [email protected] http://lists.easysw.com/mailman/listinfo/fltk
