xemacs-gtk now displaying unicode

I have the internal char to utf-8 decoding function done, and now xemacs-gtk can use pango to display well rendered unicode characters. There are still some problems with descenders, see Khmer, but most other combining marks are displayed correctly. Some of them are showing up as extra spaces. I have a screenshot at http://i.imgur.com/PfcSP.png On the left is xemacs, on the right is gedit, which also uses pango. Some of the characters require a taller line, like Khmer. Down near the bottom you can see Shavian, which isn't on the BMP, which xemacs can't currently handle. Here's a second screenshot, which shows right to left languages. Anywhere there's a box with hexadecimal digits is character that's not in the default Monospace font under Gnome, DejaVu Sans Mono. I haven't dealt with alternate fonts yet. The file is one I saved from the gmane.emacs.devel list, which I've attached. Our etc/HELLO also displays well.