January 1999 - XEmacs-Beta - XEmacs Mailing List Archives

Re: (current-time) vs (current-time-string) 25 years, 5 months

William M. Perry

"Truls Arnegaard" <arnegaar(a)oslo.geco-prakla.slb.com> writes: > Hi, > > (current-time-string) > "Tue Jan 12 20:27:05 1999" > > (current-time-string (current-time)) > "Fri Jan 16 13:33:28 1970" > > Any hints? I see this here as well with cygwin b20 on win95. Built with the bundled egcs. Already got some confused mail by people wondering how I sent them mail a year before I was born. :) -bp

1 participants
0 comments

src/ChangeLog weirdness 25 years, 5 months

Gunnar Evermann

is it only me or are there some repetitions of Changelog entries? e.g. 1998-12-06 Martin Buchholz <martin(a)xemacs.org> * frame-msw.c (mswindows_init_frame_1): - use make_lisp_hash_table, not Fmake_hash_table - include elhash.h * lisp.h: * alloc.c (make_vector): remove travesty ... can be found at least three times (lines 240, 311, 792 with slight modifications). These numbers refer to 'Artemis'. Gunnar -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

1 participants
0 comments

Re: (current-time) vs (current-time-string) 25 years, 5 months

Hrvoje Niksic

"Truls Arnegaard" <arnegaar(a)oslo.geco-prakla.slb.com> writes: > Hi, > > (current-time-string) > "Tue Jan 12 20:27:05 1999" > > (current-time-string (current-time)) > "Fri Jan 16 13:33:28 1970" > > Any hints? Doesn't happen to me. What does `(current-time)' return for you?

1 participants
0 comments

Re: 21.2-b8 SEGV during garbage collection 25 years, 5 months

Gunnar Evermann

Didier Verna <verna(a)inf.enst.fr> writes: > Karl Kleinpaste writes: > > Karl> Program received signal SIGSEGV, Segmentation fault. 0x808198c in Hi Karl, as you seem to get an above average number of crashes, could you build _with_ error-checking/debug and see if you get crashes again? Yes, I know everything gets even slower, but maybe it helps in tracking the bugs sooner. > Karl> #3 0x80ac439 in Ffuncall (nargs=3, args=0xbfffe80c) at eval.c:3135 > Karl> #4 0x80882f8 in execute_optimized_program ( ... (gdb) list 4112 { 4113 > > One more indice of the (decidedly very probable) bug introduced in the > latest byte compiler mega-patch. One more because the ones I got recently, > along with the doctor-crash someone reported earlier look very very similar. FWIW I don't think that this one is necessarily related to Martin's byte-code mega-patch, but he also rewrote the hashtable implementation:-) Didier, do you have a backtrace for your crahshes? Gunnar -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

1 participants
0 comments

Re: rssh.el interactions with ange-ftp/efs 25 years, 5 months

William M. Perry

Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: > Alan Shutko <ats(a)acm.org> writes: > > > J> /::xsh:user＠host:path > > > > This is a bad idea. As of Emacs 20.1: > > > > ** You can "quote" a file name to inhibit special significance of > > special syntax, by adding `/:' to the beginning. Thus, if you have a > > directory named `/foo:', you can prevent it from being treated as a > > reference to a remote host named `foo' by writing it as `/:/foo:'. > > :-( Then we need some other convention. However I still think my > original idea has merit. What about > > /:user＠host＠method:path > > for a specific method? > > and then > > /:user＠host:path is handled by the general code to figure out the > appropriate method. > /:＠host＠method:path is handled by the method the find the appropriate > default for user. > /:host:path is is handled by the general code to figure out the > appropriate method and then by the method the find the appropriate > default for user. > > This is just slightly less compatible. Or is the extra ＠ in use too? Why not just start using URLs? Once upon a time I had the stuff for Emacs/W3 to do this for files like http://blah/blah/blah Ange-ftp/efs could grab ftp:// stuff, xsh://user:password＠hostname/ would work, yadda yadda yadda. -Bill P.

3 participants
2 comments

Re: rssh.el interactions with ange-ftp/efs 25 years, 5 months

Jan Vroonhof

Alan Shutko <ats(a)acm.org> writes: > J> /:user＠host＠method:path I mean /user＠host＠method:path of course. Jan

2 participants
1 comments

Re: [bug] Newline forced for stream consoles? 25 years, 5 months

Hrvoje Niksic

Kyle Jones <kyle_jones(a)wonderworks.com> writes: > Hrvoje Niksic writes: > > Kyle Jones <kyle_jones(a)wonderworks.com> writes: > > > How is this hurting dunnet? > > > > `xemacs -batch -l dunnet' displays ugly newlines after all the > > prompts. Ugly, ugly, ugly. Cf. with FSFmacs, which handles it > > correctly. > > I think this misses the point of batch mode. From time immemorial > 'batch' has implied 'non-interactive'. So there should be no > interactivity, and therefore no prompts. Um. Not exactly. "Batch" mode can be used to write XEmacs scripts, if anyone were crazy enough to write one. By saying "no interactivity, no prompts," you exclude the possibility of an XEmacs script. By your logic, the whole "stream console" implementation is mostly useless. Also, FSF Emacs handles (lack of) newlines in batch mode just fine; I don't see a compelling reason for us to sugar-coat the output.

1 participants
0 comments

Re: [ERROR] 21.2 w3 display image -> SIGSEGV 25 years, 5 months

Gunnar Evermann

Norbert Koch <n.koch(a)delta-ii.de> writes: > - XEmacs 21.2 "Artemis" [Lucid] (i686-pc-linux, Mule) of Wed Jan 6 1999 > on apollonius.delta-ii.de > > - WWW xp 1998/12/18 > > both from the respective CVS trees. > > I've tried to display images on <URL:http://www.debis.de/>. Always > within the third frame (out of four), no matter what order, I get a > SIGSEGV. I see the same crash with XEmacs 21.2 "Artemis" [Lucid] (i586-pc-linux) (lic5 if thats relevant) and WWW 4.0pre.23 (that's probably quite old, isn't it?). Alas here the stack gets messed up and I get no useful backtrace. Debugging sofar produced nothing conclusive, maybe this is related to the UNBIND_TO thing I posted about a couple of days ago. BTW is Martin still around or is he already in Japan and out of the loop? > I'm not sure, whether w3 or xemacs messes things up, though ... definitely XEmacs. Gunnar -- Gunnar Evermann Speech, Vision & Robotics Group Engineering Department Cambridge University

1 participants
0 comments

[bug] Newline forced for stream consoles? 25 years, 5 months

Hrvoje Niksic

This annoying bug bytes in `xemacs -batch -l dunnet'. Try this: xemacs -batch -eval '(princ "foo")' Now this: emacs -batch -eval '(princ "foo")' XEmacs prints an unwanted newline, while FSFmacs doesn't. I tried to fix it by removing the \n in stream_delete_console(), but it only fixes the trivial case above, not dunnet.

2 participants
1 comments

Re: customize key bindings? 25 years, 5 months

Hrvoje Niksic

Valdis.Kletnieks(a)vt.edu writes: > 1) You don't know what their keyboard looks like (and we probably > can't do the sort of things Apple did with the Macintosh 'Key Caps' > program). This is probably irrelevant, given that we can and should use Emacs keystroke abstractions like "M-foo" or "C-bar". Even if we did know what their keyboard looked like, it would be worth shit on TTY's. Emacs key sequences at least have a well-defined meaning everywhere. > 2) A user trying to customize keymaps may not *know* or *care* > what the underlying Lisp function is called. There's not much Customize can do about this, nor about your third point. I was actually talking about a totally different (and more earthly, dare I say) range of problems. The situation I'd like Customize to deal with is a way to say `M-x customize-keymap RET c-mode-map RET', and handling this sensibly. My first attempts were to convert the keymap into an alist, and present the user with editing the alist in a Customize-like environment, but that sucked, for three reasons: 1) The a-list was extremely large and cumbersome to edit. This is a general problem with Customize, which cannot (easily) *modify* complex data structures. 2) There were problems with conversions of a-list keys (if you'll excuse the pun) to the canonical format. For example, all of C-i, C-I, and TAB can be assigned to different things in XEmacs, with a list of "inheritance". These things are not easy to represent in a UI. 3) It is hard for such view to represent all the more complex things that keymaps can do, such as setting parents or certain flags. A similar problem exists with editing faces via Customize.

1 participants
0 comments

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 January 1999