And, finally, I'm writing this e-mail from a dfbterm, running on the riva framebuffer. I've simultaneously run two dfbterms, gimp, and the gtk-demo without any color distortion. The problems seems to either be with the radeon framebuffer, or the DirectFB driver for the radeon framebuffer.
Adam On Tue, 12 Nov 2002, Adam K Kirchhoff wrote: > > Me again :-) > > I'm sure everyone is getting tired of my continual posts, but I > thought I should let everyone know that I got the fonts working correctly > at home. I ended up recompiling freetype and then pango. Problem solved. > > The colors when running more than one DirectFB app, however, is > still a big problem. I might try with a different card to see how well it > works. > > Adam > > On Tue, 12 Nov 2002, Adam K Kirchhoff wrote: > > > > > On Tue, 12 Nov 2002, Adam K Kirchhoff wrote: > > > > > > > > First, I'd like to thank Ville, Denis, and Sven for helping me get > > > everything up and running :-) > > > > > > And, for the most part, everything *is* up and running on my machines. I > > > still have font issues at home (though not at work), but I'll have to go > > > home and take a look at the XftConfig file later today and make sure > > > everything is set up there. > > > > > > Unfortunately, there is now one more new bug. Though multiple > > > applications will start and run, as soon as the second application is > > > launched, the colors for the entire display get a little funky :-) In the > > > case of launcing gimp from a dfbterm, I also get an error message: > > > > > > > > > ---------------------- DirectFB v0.9.15 --------------------- > > > (c) 2000-2002 convergence integrated media GmbH > > > (c) 2002 convergence GmbH > > > ----------------------------------------------------------- > > > > > > (*) Multi Application Core. (with MMX support) (2002-11-12 12:40) > > > (*) DirectFB/misc/memcpy: using MMXEXT optimized memcpy() > > > (!) [10840: 0] DirectFB/core/fusion: (reactor_dispatch) >FUSION_REACTOR_DISPATCH > > > --> Message too long > > > (!) [10840: 0] DirectFB/core/fusion: (_reactor_process_message) >FUSION_REACTOR_DETACH > > > --> Invalid argument > > > > > > I don't see that error message if I launch a second dfbterm, but the > > > colors do get all screwed up: the background turns red, one terminal is > > > green/yello, the other is dark green (FYI, these colors are as I see them. > > > I am color blind, so I may not be seeing them correctly, but I'm pretty > > > sure these colors are what's being displayed.) > > > > > > The fusion module is loaded, and I have a /dev/fusion (which symlinks to > > > /dev/misc/fusion since I'm running devfs). > > > > > > Anyway, I'm not sure if this is really an issue with DirectFB or the > > > radeon framebuffer (I am using the framebuffer code for the radeon that's > > > distributed with MPlayer). > > > > > > Adam > > > > Another update: > > > > Using the radeon framebuffer driver that comes with the kernel, all > > DirectFB apps cause my display to go blank. This is what I experienced > > the first time I tried DirectFB (a number of months ago), and why I > > switched to the driver that comes with MPlayer. > > > > Next, I copied the XftConfig file from my computer at work (where gtk apps > > on DirectFB are working properly) to my computer at home (where the fonts > > weren't showing up). Unfortunately, this hasn't solved the problems of > > the empty fonts. Again, I really don't think this is a pango or Xft issue > > since pango-viewer works just fine under X, displaying all the correct > > fonts. I'm wondering i gtk displayed any errors that I didn't pick up > > when running ./configure or make. I might rebuild it just to be sure. > > > > Finally, I am absolutely positive that I was not running DirectFB at 8bit > > depth. df_dok showed that it was benchmarking at 16bit :-) > > > > Adam > > > > > > > > -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe directfb-users" as subject.
