June 1998 - XEmacs-Beta - XEmacs Mailing List Archives

XEmacs 21.0 Package Building 26 years, 5 months

Oliver Graf

Hi there! What packages are considered core (i.e. should be inside a 'main' xemacs distribution package), and which are considered auxilliary (i.e. the installers choice)? I'm currently writing a automatic package-get-base to rpm.spec converted (including correct and easy to use dependencies). I just wonder which package should go into the main tarball/rpm. Or should none? At least the mule-stuff for a MULE enabled package... Regards, Oliver.

1 participants
0 comments

Re: Any (brave :-) mew users? 26 years, 5 months

Greg Klanderman

It'd be really nice if C-x C-f etc would do ~user completion. Tcsh and most other shells have had this forever. If you send me the /etc/passwd code I can look into it if there's interest. greg >>>>> "Jan" == Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: Jan> Jan> Darryl Okahata <darrylo(a)sr.hp.com> writes: >> * Automatic alias completion, based upon MH aliases, sendmail aliases, >> and /etc/passwd. Jan> Jan> Yes!... (But please make this a seperate package so use poor Gnus guys Jan> get to get /etc/passwd completion too!) Jan> Jan> Jan Jan>

1 participants
0 comments

Re: Modeline shadow recomputed to late (Toggenburg) 26 years, 5 months

Greg Klanderman

I noticed this a while back... digging around a bit, it looks like the problem is in redisplay-x.c, x_output_display_block(): /* #### This is really conditionalized well for optimized performance. */ if (dl->modeline && !EQ (Qzero, w->modeline_shadow_thickness) && (f->clear || f->windows_structure_changed || w->shadow_thickness_changed)) x_bevel_modeline (w, dl); redisplay-msw.c is similar. I'm not familiar enough with this to fix it. BTW, vertical dividers exhibit the same bug with the gui-element face. greg >>>>> "Jan" == Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: Jan> Jan> [This might have been reported before] Jan> Try doing the following in a vanilla Emacs *scratch* and watch the modeline shadows Jan> Jan> (set-face-background 'modeline "green") C-j Jan> C-x 2 Jan> (set-face-background 'modeline "gray80") C-j Jan> C-x 2 Jan> Jan> Jan Jan>

1 participants
0 comments

Re: More C mode syntactic yuckage 26 years, 5 months

Colin Rafferty

Colin Rafferty writes: > Hrvoje Niksic writes: >> Insert this at the beginning of a fresh C buffer: >> /* Sucks >> (sucks >> ...) something else */ >> Now go to end of buffer and press M-C-b. For me, this takes point to >> the beginning of `else' instead of skipping the buffer. > You will be (un)happy to know that every version since 19.11 has done > this. > And if you move the `(' off the BOL, everything works fine. > The problem is not in C-mode per se, but deep in the depths of > find_start_of_comment(). > Line 526 of syntax.c decides that a character code type of Sopen at > the beginning of the line means we are done looking, and the function > itself will then leave you at the end of the comment. > I guess the thing to do is figure out why the assumption at line 526 > is there. I guess I could remove it, rebuild, and see what breaks. > I will. I did. It works correctly, and everything else seems to be fine. I have a feeling that this is very old and stale code that was cut and pasted from a older function that used to search backwards for the beginning of strings. I do not know whether this patch should wait until 21.1, since it may be thoroughly wrong. However, it looks to my eye like I just got rid of some obsolete code.

1 participants
0 comments

Re: More C mode syntactic yuckage 26 years, 5 months

Colin Rafferty

Hrvoje Niksic writes: > Insert this at the beginning of a fresh C buffer: > /* Sucks > (sucks > ...) something else */ > Now go to end of buffer and press M-C-b. For me, this takes point to > the beginning of `else' instead of skipping the buffer. You will be (un)happy to know that every version since 19.11 has done this. And if you move the `(' off the BOL, everything works fine. The problem is not in C-mode per se, but deep in the depths of find_start_of_comment(). Line 526 of syntax.c decides that a character code type of Sopen at the beginning of the line means we are done looking, and the function itself will then leave you at the end of the comment. I guess the thing to do is figure out why the assumption at line 526 is there. I guess I could remove it, rebuild, and see what breaks. I will. -- Colin

1 participants
0 comments

Re: gnus package is old 26 years, 5 months

Jan Vroonhof

SL Baur <steve(a)vmailer.xemacs.org> writes: > I know about this bug -- it's why I didn't immediately package 5.6.11. I am gratefull for it. You probably saved be a lot of work. Jan

1 participants
0 comments

Colors in toggenburg 26 years, 5 months

Jan Vroonhof

Is the new color code in Toggenburg? I have a few questions: [Note: I want to set all colors using elisp only] Short version of the question, why does putting (set-face-background 'default "white") (set-face-background 'gui-element "deepskyblue") in site-start.el produce an XEmacs with a white background (OK), blue modeline (OK) but gray menus and scrollbars (not OK), whereas xemacs -bg deepskyblue does change everything? Jan

2 participants
1 comments

Re: gnus package is old 26 years, 5 months

SL Baur

Hrvoje Niksic <hniksic(a)srce.hr> writes: >> Jonathan Harris <jhar(a)tardis.ed.ac.uk> writes: >> >> > version 1.16, which is gnus version 5.6.10. There is a later gnus >> > version, 5.6.11, which contains some fixes for mswindows and ought to be >> > packaged. ... > You are correct. My fix for the fix was verified to work, so we > should use 5.6.11 + my procmail patch. I know about this bug -- it's why I didn't immediately package 5.6.11. I know about the fix. I was waiting for Lars to come out with 5.6.12.

1 participants
0 comments

misc_user_event and customize problems 26 years, 5 months

Andy Piper

I'm trying to debug the customize problems under mswindows and need some input from someone who knows. Basically a popup menu encodes its result by creating a misc_user_event with msic.function set to Qeval and misc.object set to the thing to eval. Now as far as I can see this gets evaluated as part of dispatch_event() which seems correct. Unfortunately widget-choose appears to rely on the function *not* getting evaluated but rather returning (eval ("<somthing>")). For example with the menu item; ["Name" ("Name") t] widget-choose relies on (eval ("Name")) being returned. Unfortunately under mswindows the eval actually gets evaluated in dispatch_event and you get the error "Inavlid Function "Name"". Now actually the mswindows operation seems correct to me, but I don't understand why it works under X - the event isn't evaluated somehow and all works ok. Also unfortunately it *sometimes* works under mswindows and seems to be timing related - for instance if I step through the lisp it works perfectly. I really would like to fix this for 21.0 if at all possible so any input appreciated. andy --------------------------------------------------------------------------- " .sigs are like your face - rarely seen by you and uglier than you think" Dr Andy Piper, Technical Architect, Parallax Solutions Ltd mail: andyp(a)parallax.co.uk web: www.parallax.co.uk/~andyp

1 participants
0 comments

Re: Any (brave :-) mew users? 26 years, 5 months

Jan Vroonhof

Darryl Okahata <darrylo(a)sr.hp.com> writes: > * Automatic alias completion, based upon MH aliases, sendmail aliases, > and /etc/passwd. Yes!... (But please make this a seperate package so use poor Gnus guys get to get /etc/passwd completion too!) Jan

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 June 1998