SL Baur writes: On my home system (Linux sirius 2.0.34, libc.so.5.4.43) XEmacs/No-Mule gives a core dump on your test case. I'll check out the behaviour on MS-Win when I get to the office later on. #0 0x829d938 in Qfloat_output_format () #1 0x81040cf in print_internal (obj=139443936, printcharfun=137106928, escapeflag=1) at print.c:1214 #2 0x810586d in Fprin1 (object=139443936, stream=137106928) at print.c:506 #3 0x80a367c in Fbacktrace (stream=137106928, detailed=136998960) at eval.c:4936 #4 0x809de03 in fatal_error_signal (sig=11) at emacs.c:245 #5 0xbfffe614 in ?? () #6 0x80fd41d in Lstream_flush (lstr=0x82a7014) at lstream.c:372 #7 0x80bc8b8 in unix_send_process (proc=139443936, lstream=0x84c76f8) at process-unix.c:1191 #8 0x81068c0 in Fprocess_send_region (process=139443936, start=3, end=752169) at process.c:974 #9 0x80a0cd3 in Feval (form=139299512) at eval.c:3063 #10 0x809edf1 in Flet (args=139299500) at eval.c:758 #11 0x80a0959 in Feval (form=139299392) at eval.c:2997 #12 0x80a50dd in Fprogn (args=139482880) at eval.c:758 #13 0x80996bd in Fsave_current_buffer (args=139482880) at editfns.c:428 #14 0x80a0959 in Feval (form=139482892) at eval.c:2997 #15 0x80a0e3b in Feval (form=139482628) at eval.c:3083 #16 0x80a5969 in Funwind_protect (args=139482736) at eval.c:1497 #17 0x80a0959 in Feval (form=139482748) at eval.c:2997 #18 0x809edf1 in Flet (args=139482772) at eval.c:758 <snip>

Mauro Condarelli wrote: [snip] Mauro's problems are due to the fatally poor implementation of subprocesses on Windows95/98 and are not the fault of XEmacs in general. A radical overhaul of parts of the Windows subprocess support is required. I have done some work on this, but patches will have to wait 'til after 21.0. Jonathan. -- Jonathan Harris | jhar(a)tardis.ed.ac.uk London, England | Jonathan.Harris(a)symbian.com