August 1999 - XEmacs-Beta - XEmacs Mailing List Archives

Re: USER-INIT-DIRECTORY and all that jazz 25 years, 2 months

Michael Sperber

>>>>> "Hrvoje" == Hrvoje Niksic <hniksic(a)srce.hr> writes: Hrvoje> sperber(a)informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor]) writes: >> Please, guys, it's not a matter that I don't agree with you. It's >> just that I don't know how to fix it. I really invite you to try. >> I thought it would be easy. It's not. >> >> What's worse, a number of packages use the information provided -u >> in the wrong way (or not at all). I think it's better to have no >> feature than a feature which doesn't work at all like you'd expect. Hrvoje> Micheal, is it possible to change the underlying implementation and Hrvoje> Lisp API, but still have `-u' work? I offered to do that further down in the same mail. Moiself> I can envision *replacing* -u by arguments along the lines of Moiself> "-user-init-file" and "-user-init-directory" for ~/.emacs and Moiself> ~/.xemacs, and providing something like -u through that. How does Moiself> that sound? So ...? -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla

1 participants
0 comments

Re: case-sensitivity: bug or feature in query-replace? 25 years, 2 months

Hrvoje Niksic

Stef Epardaud <stef(a)lunatech.com> writes: > if you can't tell, should it not be an option? There is such an option. It's called `case-fold-search'. I agree with Didier. Even disregarding the principle of not making the feature *too* clever, he has actually come up with an example where your logic would fail (emacs -> EMACS).

1 participants
0 comments

Re: USER-INIT-DIRECTORY and all that jazz 25 years, 2 months

Hrvoje Niksic

sperber(a)informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor]) writes: > Please, guys, it's not a matter that I don't agree with you. It's > just that I don't know how to fix it. I really invite you to try. > I thought it would be easy. It's not. > > What's worse, a number of packages use the information provided -u > in the wrong way (or not at all). I think it's better to have no > feature than a feature which doesn't work at all like you'd expect. Micheal, is it possible to change the underlying implementation and Lisp API, but still have `-u' work? For instance, we could have a `user-startup-directory' that defaults, say, to "~/", and which `xemacs -u joe' would set to "~joe/", etc. Then correct startup code would do things like: (expand-file-name ".gnus" user-startup-directory) A note for Didier and Kyle: please note that Michael is right about the code already not using `-u' correctly. I'm fairly certain that it never worked right with Gnus. Even if we don't lose the feature, it could certainly stand changing.

1 participants
0 comments

Re: case-sensitivity: bug or feature in query-replace? 25 years, 2 months

Didier Verna

Stef Epardaud <stef(a)lunatech.com> wrote: > if you can't tell, should it not be an option? Not necessarily: - A user option is supposed to describe a global behavior. The behavior you want appears to me very likely to change from search to search. That's the reason why I hate case-fold-search. It says more or less "you (don't) want case sensitivity all the time". It's stupid. Case sensitivity is heavily dependant on the precise search you're doing right now. - A function argument would be fine, but these functions have already a prefix. So better write new functions. > > BTW, the real problem is that case-fold-search is an abomination > > though. We have to cope with it. > > do we? It's called backward something... I hate it ;-) -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / ENST, INFRES C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. +33 (1) 45 81 73 46 75013 Paris, France Fax. +33 (1) 45 81 31 19

1 participants
0 comments

Re: case-sensitivity: bug or feature in query-replace? 25 years, 2 months

Didier Verna

Stef Epardaud <stef(a)lunatech.com> wrote: > well, it deduces case-sensitivity from uppercase characters in the FROM, > but what do you think the user expects when he wants to replace 'blabla' > with 'blaBla', really ? The whole point is that you can't tell. Maybe Joe wants lowercase `blabla' to be replaced (lie I do most of the time), or maybe not. You just can't tell. However, there's a big difference in finding an uppercase letter in FROM because there's no reason to type uppercase letters in there, unless you actually want case sensitivity. There, you *can* tell[1]. > i mean, if i don't put any uppercase letters in FROM but specifically > put some in TO, does it not mean i _want_ case-sensitivity ? Well, no. If I want to convert all instances of "emacs" and "Emacs" to "EMACS", sure I don't want case sensitivity. I'm happy to do a `replace "emacs" with "EMACS"' without case sensitivity. BTW, the real problem is that case-fold-search is an abomination though. We have to cope with it. Footnotes: [1] Kyle objected once that this is not always the case, like if you're pasting an X selection. -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / ENST, INFRES C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. +33 (1) 45 81 73 46 75013 Paris, France Fax. +33 (1) 45 81 31 19

2 participants
1 comments

Re: case-sensitivity: bug or feature in query-replace? 25 years, 2 months

Didier Verna

Stef Epardaud <stef(a)lunatech.com> wrote: > now i know the replace is case-sensitive if the FROM argument has case, but > should it not be case sensitive as well if the TO argument has case when the > FROM does not have ? No. You search for a /pattern/, you replace with a /string/. If you have `case-fold-search' to non-nil, it's convenient that XEmacs notice the uppercase character in FROM, and deduce that you actually want case sensitiveness. Moerover it's coherent with what interactive searches do. It doesn't make sense to toggle case sensitiveness according to the TO string. > so it did after all what i wanted, but still it did match 'ObjectName' > which i didn't want it to match. actually i think it would make sense to > make it case-sensitive if there are case in any of the 2 arguments . Again, no, because what you want to do is more complex than what the feature is for. You actually want a different treatment for one character in the string, being uppercase or lowercase. That's the kind of thing query-replace-regexp is for. You can query-replace-regexp "object[n\\|N]ame", in which XEmacs will notice the uppercase letter and deduce you want temporarily case sensitiveness. This won't match "ObjectName". For really complex stuff, you'll have no choice but toggle manually case sensitiveness as you wish, but this is a good thing. Let's not complicate the existing functions. And writing `query-replace-with-case' takes 3 lines. > tell me i am wrong, or better make this an option, or am i so wrong i > don't deserve an option ? :-/ You're so wrong you don't deserve an option :-) -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / ENST, INFRES C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. +33 (1) 45 81 73 46 75013 Paris, France Fax. +33 (1) 45 81 31 19

2 participants
1 comments

Success: 21.2.19, hpux 10.20, minor changes required. 25 years, 2 months

Richard Cognot

uname -a: HP-UX fronsac B.10.20 A 9000/782 2015948519 two-user license ./configure '--site-libraries=/home/cognot/APPS/lib/hp700 /opt/audio/lib' '--site-includes=/home/cognot/APPS/include /opt/audio/include' '--debug=no' '--compiler=cc -Ae' '--cflags=-D_HPUX_SOURCES +DA782 +DS782' '--dynamic' '--x-libraries=/usr/lib/X11R6 -L/usr/lib/Motif1.2_R6' '--x-includes=/usr/include/X11R6 -I/usr/include/Motif1.2_R6' '--prefix=/usr/local/XE21.2' XEmacs 21.2-b19 "Shinjuku" configured for `hppa2.0-hp-hpux10.20'. Where should the build process find the source code? /usr/local/src/xemacs-21.2.19/xemacs What installation prefix should install use? /usr/local/XE21.2 What operating system and machine description files should XEmacs use? `s/hpux10-shr.h' and `m/hp800.h' What compiler should XEmacs be built with? cc -Ae -D_HPUX_SOURCES +DA782 +DS782 Should XEmacs use the GNU version of malloc? yes Should XEmacs use the relocating allocator for buffers? yes What window system should XEmacs use? x11 Where do we find X Windows header files? /usr/dt/include /usr/include/X11R6 -I/usr/include/Motif1.2_R6 /usr/contrib/X11R6/include Where do we find X Windows libraries? /usr/dt/lib /usr/lib/X11R6 -L/usr/lib/Motif1.2_R6 /usr/contrib/X11R6/lib Additional header files: /home/cognot/APPS/include /opt/audio/include Additional libraries: /home/cognot/APPS/lib/hp700 /opt/audio/lib Compiling in support for XPM images. Compiling in support for PNG image handling. Compiling in support for (builtin) GIF image handling. Compiling in support for JPEG image handling. Compiling in support for X-Face message headers. Compiling in native sound support. Compiling in support for DBM. Compiling in support for CDE. Compiling in support for ToolTalk. Compiling in EXPERIMENTAL support for Drag'n'Drop ( CDE ). Compiling in support for proper WM_COMMAND handling. Using Lucid menubars. Using Lucid scrollbars. Using Motif native widgets. Using Motif dialog boxes. Compiling in DSO module support. movemail will use "dot-locking" for locking mail spool files. There are some minor changes needed to compile, due to what I think are shortcomings of the HP ANSI C compiler. Trouble is, I've been away from beta testing for so long that I don't know what the policy for sending patches is currently ;-). Two files need patching: glyphs-x.c: 4 instances where IMAGE_INSTANCE_X_MASK is used like this: IMAGE_INSTANCE_X_MASK (p) = <something>; The HP compiler gives this: Pixmap p ; void *val ; val = (void*) p ; // correct. (Pixmap)val = p ; // Wrong: Modifiable lvalue required for assignment operator (around lines 405, 843, 1375, 1981). redisplay-output.c: 2 instances where a struct is initialized from runtime computed values, which the compiler chokes on. (around lines 1240: struct display_box ,1248: struct display_glyph_area). I've #ifdef'ed the changes I made for HPUX10, but they should be safe for any compiler, although they may lead to additionnal warnings about conversions between pointers/integer in glyphs-x.c. Should I send the diff file with the ifdefs, without? Richard. -- |-------------------------------------------------------------| | Richard Cognot | Proceed, with fingers crossed... | | | /\^^/\ | | <cognot(a)ensg.u-nancy.fr> | `' | |-------------------------------------------------------------| | http://www.ensg.u-nancy.fr/~cognot | |-------------------------------------------------------------|

3 participants
3 comments

Re: latest 21.2 CVS sources: wierd hang under Windows NT 25 years, 2 months

Andy Piper

At 03:46 PM 8/30/99 -0700, Damon Lipparelli wrote: >After building the most recent (as of 3:44 pm PDT on 8/30/99) 21.2 sources, >I get a strange hang in XEmacs when I try to access the menubar. The menu >entry that I'm selecting (say, "Options") highlights, but the menu never >appears. XEmacs starts eating CPU at a furious rate, and then, eventually, >I get the following error in the echo area: > >(malformed-list #<INTERNAL OBJECT (XEmacs bug?) (symbol-value-forward type >13) 0x1167680) > >Is anyone else seeing this? Not me. What's your configuration? andy -------------------------------------------------------------- Dr Andy Piper Senior Consultant Architect, BEA Systems Ltd

1 participants
0 comments

case-sensitivity: bug or feature in query-replace? 25 years, 2 months

Stef Epardaud

following the recent patch on isearch about case-sensitive searches, i've been using query-replace to fix a file where i typed 'objectname' instead of 'objectName' (the 'N' is important), but i already had about a thousand instances of 'objectName' around and several 'ObjectName' as well. so i wanted to replace all instances of 'objectname' with 'objectName', and the replace matched all three forms. now i know the replace is case-sensitive if the FROM argument has case, but should it not be case sensitive as well if the TO argument has case when the FROM does not have ? or would it make sense not to? anyways, the replace worked for all instances, because objectname => objectName ObjectName => ObjectName objectName => objectName so it did after all what i wanted, but still it did match 'ObjectName' which i didn't want it to match. actually i think it would make sense to make it case-sensitive if there are case in any of the 2 arguments . tell me i am wrong, or better make this an option, or am i so wrong i don't deserve an option ? :-/ -- # Stef Epardaud, # There is no limit to the power of computing ... # Java Defeater # ... except men maybe ? # Earth # Lunatech Research, # Solar System # soon we will quit researching and start finding...

1 participants
0 comments

Re: USER-INIT-DIRECTORY and all that jazz 25 years, 2 months

Didier Verna

Kyle Jones <kyle_jones(a)wonderworks.com> wrote: > But if you smoke init-file-user and the support code, I don't see how "emacs > -u" is going to keep working. > And I think it should keep working. Forget about the elaborate > user who loads a bazillion packages from his private stash of > bleeding edge Lisp. Think about the simple frood using a shared > account who just wants her customization variables set and her > ＠#&$*^ key bindings loaded. THAT still works, but it won't if we > smoke "emacs -u". I whole-heartedly agree with Kyle. I'm usually pissed off by backward compatibility, but this time I'm sure we can't afford losing "-u". It simply /is/ convenient for many things included the situations Kyle describes, and many others (debugging somebody else's code, using settings from a project's account etc). The fact that some stuff requires the location of ~/.xemacs before the command line is read sounds broken at a very low level. Isn't it possible to fix this, maybe even by making a special early treatment for -u ? -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / ENST, INFRES C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. +33 (1) 45 81 73 46 75013 Paris, France Fax. +33 (1) 45 81 31 19

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 August 1999