8 December 2000

Have been hacking on pygtk recently, and a small amount on glib HEAD and framebuffer gtk (which is looking really promising). I did up the first cut at allowing arbitrary GtkTreeModels to be defined in python code. It leaks badly, and it will probably be near impossible to fix correctly :(.

The glib patch was to add some convenience functions for the GSignal code, as it is so difficult to use the existing functions people are still creating GtkObjects because of the gtksignal compatibility wrappers. Still waiting on feedback from Tim about it

Yesterday night alex asked me to try compiling pygtk with the framebuffer port of GTK, which he is working on. After adding a single missing function to the framebuffer gdk backend, pygtk compiled with no source modifications, which was good. I was having trouble with the “ms” serial mouse driver in GtkFB and my mouse. I put together a patch to make finding the start of mouse packets a little smarter, and to fix up the mouse button handling for that driver. The level of functionality in GtkFB is quite impressive.

Last Sunday, I went to the reconciliation walk in the city, which went quite well. Lots of people turned up. Also, on the way there I noticed a big banner on the old Swan Brewery (which has been a sore point, because it was an Aboriginal sacred site) saying “sorry”. I don’t know if anything different will happen with the development at that site though.