"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
Conrad> Thanks to Vin Shelton for his patch to get the ilisp
Conrad> package compiled. Now I have a SEGV in compiling
Conrad> tramp-util.el.
This isn't really a SEGV in the usual sense (writing through a wild
pointer), it's an infinite loop.
Agreed: the immediate cause of death was a SEGV but that may well be
cygwin's way of mentioning that the stack has run out.
Conrad> This is running the latest CVS HEAD copy of xemacs
Known to have at least one such in the lstream code.
Good: at least, it's better that a bug be known than not, which was the
basis of my post really.
As Adrian says, for working on the packages you really should use
the
stable (21.1.14) or gamma (21.4.8) XEmacs. The 21.5 code is very
unstable right now, Mule or no Mule.
I wasn't really trying to work on the packages: I've been looking (vaguely
as yet) at a long standing problem in cygwin where XEmacs loses the output
(occassionally) of shell commands (particularly, and most irritatingly, of
compile output). I've got into the habit of using the CVS HEAD for playing
around with, and this is the first time for quite a while it's bitten me
this badly. Anyhow I'll move back off the leading edge a little and keep on
stumbling about in the code until I've got something useful to say.
Thanks for the responses,
// Conrad