Jan Vroonhof writes: OK, I've tried this, and it still fails. Here is the traceback. _kill() ["kill.s":15] fatal_error_signal(sig = 6) ["emacs.c":278] _sigtramp() ["sigtramp.s":59] _kill() ["kill.s":15] _raise(<stripped>) ["raise.c":25] abort(<stripped>) ["abort.c":38] assert_failed(file = 0x10098e30 = "/devdisk/d54man/gnu/xemacs-beta/xemacs-21.2/src/lisp.h", line = 639, expr = 0x10098e68 = "RECORD_TYPEP (obj, &lrecord_cons)") ["emacs.c":2828] error_check_cons(obj = 269614188) ["lisp.h":639] reset_all_consoles() ["sysdep.c":1993] shut_down_emacs(sig = 11, stuff = 271693828) ["emacs.c":2459] fatal_error_signal(sig = 11) ["emacs.c":259] _sigtramp() ["sigtramp.s":59] Fcons(car = 271693924, cdr = 271693828) ["alloc.c":983] deferror(symbol = 0x100d6bec, name = 0x1001f478 = "error", messuhhj = 0x1001f480 = "error", inherits_from = 271693828) ["symbols.c":3292] init_errors_once_early() ["data.c":2018] xemacs_21_2_b26_mips_sgi_irix6_2(argc = 2, argv = 0x7fff2f14, envp = 0x7fff2f20, restart = 0) ["emacs.c":893] main(argc = 2, argv = 0x7fff2f14, envp = 0x7fff2f20) ["emacs.c":2264] __istart(<stripped>) ["crt1tinit.s":13] I think it has been broken quite recently, because I'm fairly sure I've built and ran successfully with the vanilla source this year. -- Paul Bibilo, Delcam plc, | Tel: +44 121 766 5544 Talbot Way, Small Heath, Birmingham, | Fax: +44 121 766 5511 England, B10 0HJ. _____|_____ ICBM-mail: 52 17N 1 30W #include <disclaimer.h> | BLAT FOOP | ICQ: 32045109