February 1999 - XEmacs-Beta - XEmacs Mailing List Archives

Ben Wing

I've compiled a short list of bugs that I consider to be showstoppers for the 21.0 release. It's located at: http://devel.xemacs.org/xemacs/showstoppers.html Ben (typed by Martin)

6 participants
6 comments

Re: gud.el broken or is it just me? 23 years, 9 months

SL Baur

Aki Vehtari <Aki.Vehtari(a)hut.fi> writes in xemacs-beta(a)xemacs.org: ... > Btw. gud.el in XEmacs is dated 1993 and gud.el in emacs-20.3 is > dated 1998. Unfortunately gud.el in emacs-20.3 does not work in > XEmacs and I don't know much work it would need to make it work and I've done the initial port of gud.el to XEmacs. It mostly works[1], but it sorely lacks the functionality in gdbsrc.el. Making a gudsrc.el mode is on my list of things to do. It was in ftp://altair.xemacs.org/pub/xemacs-futures/ when I was on the 'net. It was delayed integration into the XEmacs packages pending 21.0 release. I also have a ported cua-mode.el to integrate. NOTE: The gud.el and cua-mode.el XEmacs portings are courtesy of Altrasoft, Inc. Thank you Altrasoft. Footnotes: [1] I have a report that my port does not work on Solaris with dbx.

2 participants
1 comments

RoboCup 24 years

Akram Khodadadi

Hello Please answer my question. We maked a player in Linux with Lisp languaege ,but we dont know how to make a UDP connection between our player and the server with Lisp languaege. We looked in many sources ,but we didn't attain any affects. Please say to us for making a UDP connection and geting and sending messages with that Lisp languaege which orders . Please explan with an example. Thanks Mrs Khodadadi

3 participants
2 comments

[21.0b62] mouse copy and paste on NT4.0 25 years, 9 months

Roger Wimmer

Hi I have built Xemacs 21.0b62 on NT 4.0. All works fine except the mouse binding. I can highlight a region with the left button of the mouse. But if i try to paste the selection with the middle button, nothing appears in the buffer. But I have the following error message in the minibuffer: "Kill ring is empty" When I use C-Insert after highlighting the region, paste with the middle button works fine. How can I change that the highlighted region is in the kill ring? Roger -- Lieber wenige Programme die laufen als viele die abstuerzen! (Ein gluecklicher SUN-user ;-) Aber zuhause hat man ja Zeit zum Basteln.

2 participants
1 comments

Re: caller GCPROs? 26 years

Hrvoje Niksic

Gunnar Evermann <ge204(a)eng.cam.ac.uk> writes: > Is there any particular reason for this rather obvious lie ('This > function can GC') at insdel.c:2644? > > --------------- > Charcount > buffer_insert_from_buffer_1 (struct buffer *buf, Bufpos pos, > struct buffer *buf2, Bufpos pos2, > Charcount length, int flags) > { > /* This function can GC */ > Lisp_Object str = make_string_from_buffer (buf2, pos2, length); > return buffer_insert_string_1 (buf, pos, 0, str, 0, > XSTRING_LENGTH (str), flags); > } That's no lie. buffer_insert_string_1() can not only GC, but invoke arbitrary Lisp code! Remember {before,after}-change-functions.

1 participants
0 comments

Re: Completing users crash XEmacs 21.2.7 26 years

Simon Josefsson

Gunnar Evermann <ge204(a)eng.cam.ac.uk> writes: > Maybe somebody who understands MULE can make a patch for > this. Otherwise I will read up on MULE and send a patch next week. Thank you! > P.S.: Simon, the line numbers in your backtraces are always > completely messed up. I hadn't noticed. Perhaps my gdb isn't working properly, I'll try using dbx or something else instead. -s

1 participants
0 comments

caller GCPROs? 26 years

Gunnar Evermann

Is there any particular reason for this rather obvious lie ('This function can GC') at insdel.c:2644? --------------- Charcount buffer_insert_from_buffer_1 (struct buffer *buf, Bufpos pos, struct buffer *buf2, Bufpos pos2, Charcount length, int flags) { /* This function can GC */ Lisp_Object str = make_string_from_buffer (buf2, pos2, length); return buffer_insert_string_1 (buf, pos, 0, str, 0, XSTRING_LENGTH (str), flags); } ----------------- one consequence is the following in buffer_insert_string_1 at insdel.c:2457 ----------------- /* theoretically not necessary -- caller should GCPRO. #### buffer_insert_from_buffer_1() doesn't! */ GCPRO1 (reloc); ----------------- -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

1 participants
0 comments

Re: Completing users crash XEmacs 21.2.7 26 years

Gunnar Evermann

Simon Josefsson <jas(a)pdc.kth.se> writes: > C-x C-f C-a C-k ~ ? > > Fatal error: assertion failed, file /afs/pdc.kth.se/src/packages/gnu/emacsen/xemacs-21.2.7/xemacs-21.2.7/src/insdel.c, line 364, VALID_CHARPTR_P (ptr) > #7 0x5d1cf8 in bytecount_to_charcount () > at /afs/pdc.kth.se/src/packages/gnu/emacsen/xemacs-21.2.7/xemacs-21.2.7/src/insdel.c:364 > #8 0x49ee24 in user_name_completion () > at /afs/pdc.kth.se/src/packages/gnu/emacsen/xemacs-21.2.7/xemacs-21.2.7/src/dired.c:43 I can reproduce that in 21.2.10 --with-mule after adding a user to /etc/passwd with an non-ASCII char in the username. The problem is that the result of getpwent is just copied to the username cache (dired.c:676): cache[clen++] = strdup (pw->pw_name); later on (dired.c:693) these entries are cast to Bufbyte *: Bufbyte *d_name = (Bufbyte *) user_cache[i]; obviously these need to use one of those funky GET_CHARPTR_* macros. Maybe somebody who understands MULE can make a patch for this. Otherwise I will read up on MULE and send a patch next week. Gunnar P.S.: Simon, the line numbers in your backtraces are always completely messed up. This is a real pity, especially since you apparently get a lot of interesting crashes :-) It would be cool, if you could somehow fix the backtraces. -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

1 participants
0 comments

Re: Assertion failure in redisplay code whilst displaying MIME 26 years

Gunnar Evermann

Philip Aston <philipa(a)parallax.co.uk> writes: > In XEmacs 21.0 "Pyrenean63" [Lucid] (i686-pc-cygwin32) of Thu Feb 18 > In VM, if I middle-click a MIME message or message digest object I get > an assertion failure in the scrolling/redisplay caching code. is that reproducable for you? If yes, please describe exactly what you did. I do not normally use VM and by just playing around with a couple of MIME message I couldn't trigger this. Could you also send Lisp/C backtraces? > I don't grok the redisplay code, so its not immediately obvious what's > going on here. In my humble opinion very few things are obvious when redisplay is concerned. Gunnar -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

3 participants
2 comments

Re: help with bit stealing from address 26 years

Hrvoje Niksic

kifer(a)cs.sunysb.edu (Michael Kifer) writes: > Could somebody point me to the part of code where XEmacs decides, > depending on the OS, which bits can be stolen from address space to > tagging? There's the old code and there's the new code. In the old code, it was done in the s&m files -- look at "DATA_SEG_BITS" defines. It's explained in the "Internals" manual: [...] Also, note that there is an implicit assumption here that all pointers are low enough that the top bits are all zero and can just be chopped off. On standard machines that allocate memory from the bottom up (and give each process its own address space), this works fine. Some machines, however, put the data space somewhere else in memory (e.g. beginning at 0x80000000). Those machines cope by defining `DATA_SEG_BITS' in the corresponding `m/' or `s/' file to the proper mask. Then, pointers retrieved from Lisp objects are automatically OR'ed with this value prior to being used. In the new code, which you get using `--with-minimal-tagbits', we silently assume that two least significant bits are always available for tagging.

1 participants
0 comments

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

XEmacs-Beta February 1999