Raymond Toy writes:
> > On the other hand, whatever the code is that triggered the assert
OK, there are two asserts that match the C backtrace, but both test
internal data. (For the curious, Vcharset_tag_lists in
src/specifier.c.) So it is an XEmacs bug somehow. "Thank you for
bearing with us during these technical difficulties."
> Yeah, it's in my init file where I was adding support for other languages
> so that the hello demo displays more. I will try to narrow it down a bit,
> but the relevant code is appended.
Your code conforms to the published API, so it's not your problem
(unless you want to make it so! :-)
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
buildbot(a)lidell.nu writes:
> The Buildbot has detected a new failure on builder imac osx cfg0 while building XEmacs.
It appears that the buildbot wants GMP but support is either not
installed or (maybe more likely) the configure script needs a
--with-site-prefixes option.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
>>>>> Stephen J Turnbull <stephen(a)xemacs.org> writes:
> As I wrote, there *is* an explicit --with-bignum=gmp in the configure
> command.
OOPS! So there is! My fault. Sorry for the noise.
Yours
--
%% Mats
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
hi,
there is binary package for xemacs on suse @
https://build.opensuse.org/package/files?package=xemacs&project=M17N
if you look, there are several patches that get applied to XEmacs. I
wonder if these have been brought to any one's attention, and if they
are rubbish, distribution specific, or maybe some that should go up
stream? I see from the file name some date back to version 21.4
List of patches:
Patch0: xemacs.patch
This patch seems to be a mixed bag, adds some "safeopen" function to
mmencode.c, redefined pop to pop3 on linux in pop.c, adds exit 1 to
failing mktemp call in vcdiff, maybe contains some useful terminal key
bindings or maybe not?
and definitely looks like some useless bits and some distro customization
Patch3: xemacs-app-defaults.patch
- this is most likely distro-specific, sets some default fonts for the
menu or so?
Patch18: xemacs-21.4.8-xevent.patch
- uses XLookupString instead of XLookupKeysym in src/event-Xt.c, might
this be a real fix for some problem?
Patch20: xemacs-21.4.13-ppc64.patch
- adds a m/powerpc64.h file
Patch23: xemacs-ptmx.dif
- move some getpt() code in src/process-unix.c, possibly prefering openpty?
Patch27: xemacs-level3.patch
- adds codes for XK_Num_Lock and XK_ISO_Lock to event-Xt.c
Patch28: xemacs-21.5.18-movemail.patch
- makes "fatal" function take 3 arguments in movemail.c
Patch32: do-not-create-backups-in-temp-directories.patch
Patch33: set-locale-to-c-when-not-supported-by-x.patch
Patch39: xemacs-tinfo.dif
- adds some checks for ncursesw to configure.ac
Patch43: set-language-unicode-precedence-list.patch
- messes with chinese and japanese mule charset lists
Patch45: fix-defface-custom-modified-face.patch
- replace :bold with :bold t in cus-edit.el
Patch50: menus-always-utf8.patch
Patch51: bnc502716-fontmenu.patch
Patch52: bnc502716-xft.patch
Patch53: xemacs-21.5.31-array.patch
- changes a loop in mule-wnnfns.c to do len-- first, might this be a real bug?
Patch56: xemacs-libpng15.patch
Patch292811: bugzilla-292811-make-x-make-font-bold-italic-xft-work.patch
Patch301352: bugzilla-301352-fix-wrong-incrementing-in-macros.patch
- move ic++ outside of macro in mule-ccl.el, maybe just a style nitpick?
I don't understand a lot or any of these patches so I would be
interested in the experienced XEmacs hackers' opinions. I have
summarized some of the patch contents where I can approximately tell
what it does. In some other cases, the patch file name gives a pretty
good idea of what it is doing. Yet in some cases I have no clue what
it is doing. I would especially like to know if any of the patches are
actually breaking XEmacs in some subtle way (remember the Debian
random number generator ;-)).
Thanks,
Nei
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
Hi Stephen
(Moving back to list. I hope you don't mind. My mistake to put it off
list in the first place when replying from my phone. Sorry.)
You wrote:
> Mats Lidell writes:
>> I don't think there is any explicit request for Gmp in that configuration.
>> Check the stdio log from the configure step where you can see the options
>> to configure.
> I did, and there was. GMP is off by default.
So if it is off why does configure try to find it and complains when
it doesn't? bug?
Yours
--
%% Mats
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
XEmacs 21.5 (beta32) "habanero" dede3f658f8e [Lucid] (i386-apple-darwin12.0.0, Mule)
Cannot be built on OS X 10.8 using /usr/bin/makeinfo (v. 4.8) which ships with the system and the newest Xcode. Macports provides version 4.13 with which the build succeeds.
Version 4.8 appears not to understand the command 'frenchspacing'.
--
Dr. Robert Delius Royar PhD <webeditor(a)kcte.org>
Kentucky Council of Teachers of English / Language Arts
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta
Raymond Toy writes:
> GMP is not installed on that machine AFAIK. (Don't think I've ever used
> bignums in xemacs before.)
OK, so I guess the sensible thing to do is to remove that from the
spec for this buildbot. I guess that's for Mats to do?
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta