APPROVE 21.5 Nice catch, please commit. 2005-12-20 Aidan Kehoe <kehoea(a)parhasard.net&gt; * objects-x.c (x_find_charset_font): Return Qnil on no match, not Qunbound. Fixes performance issues based on the use of this return value in objects.c:font_instantiate Aidan> Stephen, was this part of a more wide-ranging Xft change Aidan> you didn’t get round to integrating? I don't recall making this change, offhand. It's possible that it was part of this 2005-04-14 Stephen J. Turnbull <stephen(a)xemacs.org&gt; [ omitted ] (x_find_charset_font): Clarify block structure by eliminating early returns. Remove bogus "must free" comments. Free FcPattern objects that were leaking. Use block structure to indicate lifetimes of FcPattern objects. Use "font" rather than "pattern" in names of FcPatterns that name fonts rather than search criteria. Use shortnames more consistently in debug messages. #ifdef shortname vs longname return values. Correct comment to note that FcCharSets are Unparsed to ASCII. Don't ever fall through to X core fonts if we had an Xft font (it should be NULL anyway). Use Qxft_font_name_encoding; this means we have to parse FONT separately for Qx_font_name_encoding. with the intent of provoking an error if the return value wasn't explicitly set on some branch. Thanks for catching this. -- School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.