There must be a simple explanation to this problem, but I haven't had time to investigate. Perhaps it is a simple pilot error, and not a bug at all. Nevertheless, here goes: My Rawhide 20001222 systems, including gnome-print-0.20-8.i386.rpm, have no problem running the following command, which is included in the postinstall script of gnome-print-0.20-8.i386.rpm: /usr/bin/gnome-font-install --system --scan --no-copy \ --afm-path=/usr/share/fonts/afms \ --pfb-path=/usr/share/fonts/pfbs \ --pfb-assignment=ghostscript,/usr/share/ghostscript/fonts \ --afm-assignment=ghostscript,/usr/share/ghostscript/fonts \ --pfb-assignment=ghostscript,/usr/share/fonts/default/Type1 \ --afm-assignment=ghostscript,/usr/share/fonts/default/Type1 \ /usr/share/fonts Running the same program on my Rawhide 20010103 systems, with either gnome-print-0.25-1.i386.rpm or gnome-print-0.20-8.i386.rpm installed, seems to yield an infinite loop, or at least a command execution that soaks the CPU for longer than I am willing to wait. These systems all have 352 RPMs installed, almost all from Rawhide. If further details would help, I'd be happy to provide them. These systems have no /usr/share/ghostscript/fonts directories, and they have 771 files under /usr/share/fonts.
gnome-print-install was considerably rewritten in gnome-print-0.29, so I'm guessing this probably is fixed now. gnome-print-0.29-2 should show up in rawhide sometime soon, and if it still happens with that, please reopen.