A weird segfault in reportbug turns into a hunt for some weirdness in libpthread on Linux/m68k. So far, we can duplicate the problem on 040s (1 Amiga, 1 Mac), but not 060s (2 Amigas). Anyone with an 020 or 030 got time to test it?
RoutePlanner is progressing nicely; I'm about 6 hours of hacking away from having RouteEdit working (the main things left are designing the extend/break dialog and figuring out how to get the city selector from the RoutePlanner widget tree to import into RouteEdit's). I've come to the conclusion that glade (with libglade) is something of a double-edged sword, and that Gtk really needs builtin double-click handling. I hate to say it, but MUI got the whole GUI toolkit thing right 5 years ago (i.e. back when I wrote the original version of RoutePlanner), and we're still catching up.