August 1998 - XEmacs-Beta - XEmacs Mailing List Archives

Re: Macros: Scheme vs Common Lisp 26 years, 3 months

Michael Sperber

Yo Craig, your message is really hard to read as its formatting is totally broken. I'll try to reconstruct: >>>>> "Craig" == Craig Lanning <CraigL(a)internetx.net> writes: Craig> One example that shows the "symbol names function" vs "symbol Craig> names variable" problem is Craig> (let ((cons '())) Craig> (push "ghengis" cons) Craig> (push "khubla" cons) Craig> cons) Craig> This is a problem in Scheme because CONS is defined to be a Craig> function for constructing lists. Within this LET it is rebound Craig> to be a variable holding an empty list. In CL, function and Craig> variable bindings are separate. CONS is defined as a function Craig> by the standard, but carries no variable definition [it is Craig> unbound], therefore, LET is free to establish a local variable Craig> binding for CONS and PUSH will not be confused because it will Craig> use the function binding for CONS. This form evaluates Craig> correctly in CL and EL. This indeed shows the CONS example to be poor. The point remains the same, however: What if we bind a variable bound in a macro. Say, *features*? What if a temporary identifier in a macro definition shadows a binding in a use? Craig> One example that can be fixed by use of packages is [Scheme Craig> form]: Craig> If the application really needs a function named CONS with Craig> different behavior [ ... ]that CL's CONS, a new package should be Craig> defined which shadows the CONS symbol [CL form]: You're missing the point: When you're using a macro, you don't know which bindings it refers to. Therefore, you have no handle whatsoever on using the right package magic in order to avoid capture problems. In the examples, I don't *need* to use CONS, it just might happen to be a convenient name. What I would also want is, that, if a macro is defined within a certain package, and I refer to bindings from the macro, that these bindings are relative to the package of definition rather than the package of use, obeying the standard rules of lexical scoping. Not having this (as in CL) introduces other dangers. The fundamental reason for this and many of the other troubles in CL (and ELisp, for that matter), is the persistent confusion between symbols and names. Try explaining this mess to a novice in Lisp, and you'll find out just how confusing it is. Scheme doesn't have this historical baggage. -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla

3 participants
2 comments

faster elisp - progress 26 years, 3 months

Martin Buchholz

I finally got my development XEmacs to stop core dumping, and there is visible signs of progress on the faster elisp front: Here is my generic bytecode benchmark: (progn (defun foo (a b c) (bar a b c) (bar a b c) (bar a b c)) (defun bar (d e f) nil) (byte-compile 'foo) (byte-compile 'bar) (race 200000 (foo 1 2 3))) Martin's development version: "Elapsed: 23.430829 User: 17.750000 System: 0.000000 " Function Name Ticks %/Total Call Count ===================== ===== ======= ========== bar 1164 66.174 600000 foo 518 29.449 200000 measure-compiled-form 77 4.377 1 21.2-b1: "Elapsed: 33.032012 User: 25.020000 System: 0.000000 " Function Name Ticks %/Total Call Count ===================== ===== ======= ========== bar 1586 63.874 600000 foo 799 32.179 200000 measure-compiled-form 98 3.947 1 Benchmarking harness: (when (not (fboundp 'profile-results)) (defun profile-results (&optional info stream) (get-profiling-info) (pretty-print-profiling-info))) (defmacro measure (&rest body) `(progn (let ((oct (current-time)) (opt (current-process-time))) (garbage-collect) (clear-profiling-info) (start-profiling) ,＠body (stop-profiling) (let ((ct (current-time)) (pt (current-process-time))) (print (format "Elapsed: %f User: %f System: %f\n" (+ (* 65536 (- (first ct) (first oct))) (- (second ct) (second oct)) (/ (- (third ct) (third oct)) 1000000.0)) (- (first pt) (first opt)) (- (second pt) (second opt)))))) (profile-results))) (defmacro measure-compiled (&rest body) `(progn (defun measure-compiled-form () ,＠body) (byte-compile 'measure-compiled-form) (measure (measure-compiled-form)) (disassemble 'measure-compiled-form) (fmakunbound 'measure-compiled-form))) (defmacro race (times &rest forms) `(with-output-to-temp-buffer "*Race*" ,＠(mapcar (lambda (form) `(measure-compiled (dotimes (--race-tmp ,times) ,form))) forms)))

1 participants
0 comments

Re: Can't search the archives for cperl-mode! 26 years, 3 months

Jason R Mastaler

David C Worenklein <dcw(a)gcm.com> writes: > I'm trying to search the xemacs-beta archive for "cperl-mode" > and I get "0 matches in 0 files"! It should work now. The indexes for the archives must have gotten dirty somehow.

1 participants
0 comments

Re: Suggested fixes for cperl-mode 26 years, 3 months

Karl M. Hegbloom

>>>>> "Jan" == Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: Jan> Ilya Zakharevich <ilya(a)math.ohio-state.edu> writes: >> Newer cperl-mode does not (starting from version 4.1 or some >> such). Please check whether what newer versions (starting from >> 2.?, I think) do suits you. Jan> [again, sorry for not checking out new versions Jan> first]. However this does point to the fact that the current Jan> XEmacs version is hopelessy out of date. Karl, are you still Jan> maintaining the XEmacs package? I can, unless it gets too complicated. I did that eldoc-like lazy help without understanding most of cperl-mode. It's Ilya's baby, not mine. Jan> In the interest of all involved it might be nice if Ilya Jan> could fold in the minor improvements for XEmacs in his Jan> version so that there is no need for a seperate version. Yes, Please, Ilya. Will you run an ediff once and see what's different? Jan> 2. Lazy help was done using eldoc's techniques. Jan> 3. Use defalias for setting 'perl-mode and give it an Jan> autoload cookie. Jan> 4. Obeys delete-key-deletes-forward Jan> 5. Made Auto help toggle use true XEmacs menu toggle bottons. Jan> Ad 2. I do not know if it is a win. XEmacs 21.x has timer.el Jan> in its fsf-compat package. IMHO maintainability wins over Jan> other concerns. However Karl probably introduced his stuff Jan> because use of timer.el could/can? cause problems with other Jan> package. AFAIK all these have been fixed. I am not sure what Jan> the offical word is on this. All I know is that it didn't work, eldoc-mode did a similar thing and did work, so I plagerized and hacked, and got it running. I noticed that in `cperl-mode' now, you can click the mouse on a keyword, and it shows the help right away, but in `eldoc-mode' when you do that, you have to move the cursor with the arrows to make the help string appear. I did not investigate the reason for this... I've lots else to read. (I'm working slowly through about 6 different books these days... at some point I'll be much more capable of understanding the XEmacs codes and XEmacs Lisp codes.)

1 participants
0 comments

Re: Suggested fixes for cperl-mode 26 years, 3 months

Gary D. Foster

>>>>> "IZ" == Ilya Zakharevich <ilya(a)math.ohio-state.edu> writes: IZ> Jan Vroonhof writes: >> 4. Obeys delete-key-deletes-forward >> Ilya would you accept a short patch for 3, 4 and possibly 5? IZ> I had a long email exchange about 4 with some rude person (do IZ> not remember the name). Since he refused to follow CPerl IZ> policy (that XEmacs-specific customizations should be done via IZ> macros defined at the start of the file - if possible), I IZ> could not incorporate his patches. At the extreme (and undesired) risk of starting a flamewar, that "rude person" was probably me and nowhere in Ilya's mail was any note to a specific policy or guidelines on how the patch "should be done", only comments along the lines of refusing to have anything to do with (quote) "XEmacs brokenness" (unquote). I'm hoping that whomever interacts with Ilya on this has better luck than I did because somehow, somewhere, we got off on the wrong foot and made no progress. Communication on both sides broke down quickly and asbestos underwear was required on both sides. I obviously touched a nerve somewhere and if progress is to be made with this _I_ am obviously NOT the right person to do it. *sigh* I might add that the email I sent Ilya about it (minus the actual patch, of course) was identical in tone to the email I sent Barry about cc-mode and THAT exchange (thankfully) didn't erupt into a shouting match. -- Gary F. (who will go out behind the shed and shoot himself in the foot now, just to maintain the status quo) -- "There may not be a place for humans in the future... if we're successful." -- Rodney Brooks, Roboticist

1 participants
0 comments

Re: Faster lisp engine 26 years, 3 months

Jan Vroonhof

Ilya Zakharevich <ilya(a)math.ohio-state.edu> writes: > Newer cperl-mode does not (starting from version 4.1 or some such). > Please check whether what newer versions (starting from 2.?, I think) > do suits you. [again, sorry for not checking out new versions first]. However this does point to the fact that the current XEmacs version is hopelessy out of date. Karl, are you still maintaining the XEmacs package? In the interest of all involved it might be nice if Ilya could fold in the minor improvements for XEmacs in his version so that there is no need for a seperate version. As far as I can see the XEmacs version contained the following improvements (this is already difficult to judge, which is why it is good to have just one version). 1. Customization (Ilya's 4.5 has that) 2. Lazy help was done using eldoc's techniques. 3. Use defalias for setting 'perl-mode and give it an autoload cookie. 4. Obeys delete-key-deletes-forward 5. Made Auto help toggle use true XEmacs menu toggle bottons. Ad 1. No need to anything (Maybe check of XEmacs customisations introduced incompatabilities). Ad 2. I do not know if it is a win. XEmacs 21.x has timer.el in its fsf-compat package. IMHO maintainability wins over other concerns. However Karl probably introduced his stuff because use of timer.el could/can? cause problems with other package. AFAIK all these have been fixed. I am not sure what the offical word is on this. Ad 3 and 4. These are good things. The changes are minimal. Ad 5. I am not sure this needed that badly (although it looks nice) and I am not sure this considered good taste for easymenu. Ilya would you accept a short patch for 3, 4 and possibly 5? Jan P.S. mode-compile needs to be packaged up for XEmacs. Unfortunately, I cannot contact the authors website at the moment.

2 participants
1 comments

Can't search the archives for cperl-mode! 26 years, 3 months

David C Worenklein

I'm trying to search the xemacs-beta archive for "cperl-mode" and I get "0 matches in 0 files"! I have Max. files returned set to 50, Max matches per file set to 10. I've tried setting Misspellings allowed to 0, then 1. I've tried checking partial match. I've tried just searching for "cperl". I know the messages are in there, because I can see them when I look at the August thread!

1 participants
0 comments

[OK] 21.0 "Erzgeberg-pre1" XEmacs Lucid on i586-pc-cygwin32 26 years, 3 months

Andy Piper

Get the wrong type argument stringp, nil error reported by others. andy > XEmacs Build Report as generated > by build-report-version 1.35 follows: > Contents of /development/build/xemacs-20/Installation: > (Output from most recent run of ./configure) uname -a: CYGWIN32_NT ARMAGEDON 4.0 19.3 i586 unknown ./configure '--prefix=/usr/gnu' '--package-path=/usr/gnu/lib/xemacs/packages' '--site-includes=/usr/local/include' XEmacs 21.0 "Erzgeberg-pre1" configured for `i586-pc-cygwin32'. Where should the build process find the source code? /development/build/xemacs-20 What installation prefix should install use? /usr/gnu What operating system and machine description files should XEmacs use? `s/cygwin32.h' and `m/intel386.h' What compiler should XEmacs be built with? gcc -g -O2 -Wall -Wno-switch Should XEmacs use the GNU version of malloc? yes Should XEmacs use the relocating allocator for buffers? default What window system should XEmacs use? x11 Where do we find X Windows header files? /usr/X11R6.3/include Where do we find X Windows libraries? /usr/X11R6.3/lib Additional header files: /usr/local/include Compiling in support for XAUTH. 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 TIFF image handling. Compiling in support for GNU DBM. Compiling in support for ncurses. Compiling in File coding support. Compiling in support for proper session-management. Using Lucid menubars. Using Lucid scrollbars. Using Athena dialog boxes. movemail will use "dot-locking" for locking mail spool files. Using Lisp_Objects with minimal tagbits. Using indexed lrecord implementation. > /development/build/xemacs-20/beta.err does not exist!

1 participants
0 comments

Re: Macros: Scheme vs Common Lisp 26 years, 3 months

Michael Sperber

>>>>> "Karl" == Karl M Hegbloom <karlheg(a)inetarena.com> writes: >>>>> "Craig" == Craig Lanning <CraigL(a)internetx.net> writes: Craig> Having now read "Macros in Scheme" and the first four parts Craig> of "An Introduction To Macro Expansion Algorithms" [...] Karl> Where may I find those so that I can also read them, please? They're all accessible from the Scheme section of my Web page: http://www-pu.informatik.uni-tuebingen.de/users/sperber/xemacs/next-gener... -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla

1 participants
0 comments

Re: [crash] Compiling w/o .png cures 24bpp laptop crashes. 26 years, 3 months

Karl M. Hegbloom

>>>>> "Kazuo" == Kazuo Oishi <oishi(a)agenv4.agr.yamaguchi-u.ac.jp> writes: Kazuo> I rebuilt XEmacs-21.0-b52 without PNG support. Certainly Kazuo> it stops crashing with PNG, but it still crashes to display Kazuo> other formated images, for example JPEG, GIF. Kazuo> Can your XEmacs display these images correct? I didn't even think to try... I know that it allows me to use `M-x customize-face' now, since it's not trying to display the .png buttons anymore, just [Text] ones. The xpm-button INS and DEL work fine, and so do the check buttons. [ checking to see if it will display a .gif ...] Hey, you're right. Same display bug - there are vertical lines through it, and the picture is stretched widthwise by that amount. They display more black and white than color... like the color isn't being shown right. After showing a few, it crashes. It does that with both .gif and .jpeg, but displays .xpm's just fine.

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 1998