- slime-devel - mailman3.common-lisp.net

[slime-devel] changing slime-keys
by Pascal J.Bourguignon 30 Jun '04

30 Jun '04

What function should I call when I change slime-keys to have them taken into account immediately? -- __Pascal Bourguignon__ http://www.informatimago.com/ There is no worse tyranny than to force a man to pay for what he does not want merely because you think it would be good for him. -- Robert Heinlein

3 5

[slime-devel] Patch: show the current debug level in the REPL
by Thomas F. Burdick 30 Jun '04

30 Jun '04

Below is a patch to show the current debug level in the REPL, when it's > 0 Index: ChangeLog =================================================================== RCS file: /project/slime/cvsroot/slime/ChangeLog,v retrieving revision 1.443 diff -u -F^(def -r1.443 ChangeLog --- ChangeLog 29 Jun 2004 22:07:00 -0000 1.443 +++ ChangeLog 30 Jun 2004 04:23:33 -0000 @@ -1,3 +1,10 @@ +2004-06-29 Thomas Burdick <tfb(a)OCF.Berkeley.EDU> + + * slime.el: Indicate when the REPL is in the debugger's context + (slime-debug-level): new connection var + (slime-dispatch-event): set slime-debug-level to match *sldb-level* + (slime-repl-insert-prompt): show debug-level in prompt when > 0 + 2004-06-30 Luke Gorrie <luke(a)bluetail.com> * NEWS: Wrote preliminary release notes for alpha-1. Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.346 diff -u -F^(def -r1.346 slime.el --- slime.el 29 Jun 2004 20:17:05 -0000 1.346 +++ slime.el 30 Jun 2004 04:23:46 -0000 @@ -1561,6 +1561,9 @@ (defmacro slime-def-connection-var (varn (slime-def-connection-var slime-use-sigint-for-interrupt nil "If non-nil use a SIGINT for interrupting.") +(slime-def-connection-var slime-debug-level 0 + "The current level of recursive debugging.") + ;; XXX pending continuations are not removed if Lisp crashes. @@ -1602,9 +1605,11 @@ (defun slime-dispatch-event (event &opti (sldb-activate thread level)) ((:debug thread level condition restarts frames) (assert thread) + (setf (slime-debug-level) level) (sldb-setup thread level condition restarts frames)) ((:debug-return thread level) (assert thread) + (setf (slime-debug-level) (1- level)) (sldb-exit thread level)) ((:emacs-interrupt thread) (cond ((slime-use-sigint-for-interrupt) (slime-send-sigint)) @@ -2168,7 +2173,9 @@ (defun slime-repl-insert-prompt (result (slime-insert-propertized '(face slime-repl-result-face) result) (unless (bolp) (insert "\n")) (let ((prompt-start (point)) - (prompt (format "%s> " (slime-lisp-package)))) + (prompt (if (> (slime-debug-level) 0) + (format "%s [%d]> " (slime-lisp-package) (slime-debug-level)) + (format "%s> " (slime-lisp-package))))) (slime-propertize-region '(face slime-repl-prompt-face read-only t

5 5

[slime-devel] Fedora Core 2 and doc build
by Dan Pierson 30 Jun '04

30 Jun '04

For some reason 'sort -d' either loops forever or takes a VERY long time on Fedora Core 2. This breaks the docs build. Removing the -d fixes it for me, as in the following patch, but there may have been some good reason for using the switch... Index: Makefile =================================================================== RCS file: /project/slime/cvsroot/slime/doc/Makefile,v retrieving revision 1.4 diff -u -r1.4 Makefile --- Makefile 22 Jun 2004 14:06:56 -0000 1.4 +++ Makefile 30 Jun 2004 14:02:30 -0000 @@ -43,7 +43,7 @@ contributors.texi: ../ChangeLog Makefile texinfo-tabulate.awk cat ../ChangeLog | \ sed -ne '/^[0-9]/{s/^[^ ]* *//; s/ *<.*//; p;}' | \ - sort -d | \ + sort | \ uniq -c | \ sort -nr| \ sed -e 's/^[^A-Z]*//' | \

2 1

[slime-devel] Symbol's value as variable is void: slime-net-processes while starting SLIME
by Paolo Amoroso 30 Jun '04

30 Jun '04

I have updated my working copy of the CVS sources a couple of hours ago, and SLIME does no longer work. It worked fine the previous time I had updated the sources, i.e. yesterday or at most a couple of days ago. More specifically, when I issue the `M-x slime' command from Emacs, I get the error: Error in init file: Symbol's function definition is void: slime-def-connection-var I also get this error in another circumstance: Symbol's value as variable is void: slime-net-processes Things are screwed so bad that when I run Gnus, it correctly accesses news, but fails to get mail from my local spool. This, anyway, is no big deal: Gnus and the rest work fine if I disable SLIME-related forms in my ~/.emacs file. Before starting SLIME, I made sure to delete all existing fasl files. I use GNU Emacs 20.7.2 and CMUCL 18e under Debian Woody. I just thought you might be interested in this... :) Paolo "SLIME CVS head or bust" Amoroso -- Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film

1 0

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 30 Jun '04

30 Jun '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.436 slime/ChangeLog:1.443 --- slime/ChangeLog:1.436 Mon Jun 28 14:05:28 2004 +++ slime/ChangeLog Tue Jun 29 15:07:00 2004 @@ -1,3 +1,37 @@ +2004-06-30 Luke Gorrie <luke(a)bluetail.com> + + * NEWS: Wrote preliminary release notes for alpha-1. + +2004-06-29 Luke Gorrie <luke(a)bluetail.com> + + * mkdist.sh: New shell script for creating a tarball for + distribution. + +2004-06-29 Bill Clementson <Bill_Clementson(a)peoplesoft.com> + + * slime.el (slime-who-map): Add extra bindings for the XREF + commands as with the documentation commands. Now `C-c C-w C-c' is + `slime-who-calls' in addition to `C-c C-w c', etc. + +2004-06-29 Luke Gorrie <luke(a)bluetail.com> + + * slime.el (sldb-prune-initial-frames): Tweaked regexp for + matching SWANK's own stack frames for effectiveness in SBCL. + (slime-keys): Shadow remaining inf-lisp keys (C-c C-a, C-c C-v) + with a null `slime-nop' command until we put them to a real use. + + * swank.lisp (open-streams): Renamed the restart around reads from + the user-input stream from ABORT to ABORT-READ. Invoking this + restart seems kinda dangerous, so better for 'a' in SLDB not to do + so. + +2004-06-28 Thomas F. Burdick <tfb(a)OCF.Berkeley.EDU> + + * swank.lisp (inspector-nth-part): + * slime.el (slime-inspector-copy-down, slime-inspector-mode-map): + Added copy-down command (M-RET) to easily move an object from the + inspector to the repl. + 2004-06-28 Luke Gorrie <luke(a)bluetail.com> * slime.el (slime-doc-map): New keymap for documentation

1 0

[slime-devel] New keybindings for documentation commands
by Luke Gorrie 29 Jun '04

29 Jun '04

Ahoy, As previously discussed the documentation commands have now been grouped under a `C-c C-d' prefix. The commands are: a - apropos p - apropos-package z - apropos-all d - describe-symbol f - describe-function h - hyperspec lookup ~ - hyperspec lookup of a format character The final keystroke is bound both unmodified and with control, so both `C-c C-d a' and `C-c C-d C-a' will make an apropos search. The exception is hyperspec-lookup, because it's nice to leave C-h unbound so that `C-c C-d C-h' will summarise the documentation bindings. The old bindings are removed. This has the nice effect that `C-c C-h' in a slime-mode buffer will now summarise the slime keybindings. It also shows that there are still two inf-lisp commands that we aren't currently clobbering, on `C-c C-a' and `C-c C-v'. Anyone want to fight over these bindings? I think we should have a bloody and decisive final keybinding war now and then make a 1.0-alpha release later this week. Cheers, Luke

4 5

[slime-devel] New keybindings for Who-xref commands
by Bill_Clementson＠peoplesoft.com 29 Jun '04

29 Jun '04

Attached is a patch that provides dual key bindings for the Who-xref commands (e.g. - it now allows either "C-c C-w c" or "C-c C-w C-c" for the "Who calls ..." command). This is in keeping with the Documentation key binding change. - Bill Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.345 diff -u -r1.345 slime.el --- slime.el 29 Jun 2004 17:39:01 -0000 1.345 +++ slime.el 29 Jun 2004 19:41:51 -0000 @@ -483,11 +483,6 @@ ("~" slime-sync-package-and-default-directory :prefixed t :inferior t) ("\M-p" slime-repl-set-package :prefixed t :inferior t) ;; Cross reference - ("\C-wc" slime-who-calls :prefixed t :inferior t :sldb t) - ("\C-wr" slime-who-references :prefixed t :inferior t :sldb t) - ("\C-wb" slime-who-binds :prefixed t :inferior t :sldb t) - ("\C-ws" slime-who-sets :prefixed t :inferior t :sldb t) - ("\C-wm" slime-who-macroexpands :prefixed t :inferior t :sldb t) ("<" slime-list-callers :prefixed t :inferior t :sldb t) (">" slime-list-callees :prefixed t :inferior t :sldb t) ;; "Other" @@ -516,6 +511,16 @@ (?h slime-hyperspec-lookup) (?~ common-lisp-hyperspec-format))) +(defvar slime-who-map (make-sparse-keymap) + "Keymap for who-xref commands. Bound to a prefix key.") + +(defvar slime-who-bindings + '((?c slime-who-calls) + (?r slime-who-references) + (?b slime-who-binds) + (?s slime-who-sets) + (?m slime-who-macroexpands))) + ;; Maybe a good idea, maybe not.. (defvar slime-prefix-key "\C-c" "The prefix key to use in SLIME keybinding sequences.") @@ -551,7 +556,17 @@ (let ((modified (slime-control-modified-char key))) (define-key slime-doc-map (string modified) command))))) ;; C-c C-d is the prefix for the doc map. - (slime-define-key "\C-d" slime-doc-map :prefixed t :inferior t)) + (slime-define-key "\C-d" slime-doc-map :prefixed t :inferior t) + ;; Who-xref + (setq slime-who-map (make-sparse-keymap)) + (loop for (key command) in slime-who-bindings + do (progn + ;; We bind both unmodified and with control. + (define-key slime-who-map (string key) command) + (let ((modified (slime-control-modified-char key))) + (define-key slime-who-map (string modified) command)))) + ;; C-c C-w is the prefix for the who-xref map. + (slime-define-key "\C-w" slime-who-map :prefixed t :inferior t)) (defun slime-control-modified-char (char) "Return the control-modified version of CHAR."

2 1

[slime-devel] Coding system oddities
by Lennart Staflin 29 Jun '04

29 Jun '04

Hello, I'm running slime on Mac OS X, with current (CVS) Emacs and openmcl. I tend to use Latin-1 coding for files, and the keyboard generates latin-1. In slime repl I get this (slightly edited for clairty): CL-USER> (print "Åhej") "\305hej" "Åhej" It seems that the coding system emacs uses for the printed output is another then used for the main slime interaction. Changing coding system for the output stream seems to confirm this: (set-process-coding-system (get-process "*lisp-output-stream*") 'latin-1 'latin-1) After that the above example gives: (print "Åhej") "Åhej" "Åhej" //Lennart Staflin

2 1

Re: [slime-devel] Re: New keybindings for documentation commands
by Bill_Clementson＠peoplesoft.com 29 Jun '04

29 Jun '04

Luke Gorrie <luke(a)bluetail.com> writes: > Peter Seibel <peter(a)javamonkey.com> writes: > > > Luke Gorrie <luke(a)bluetail.com> writes: > > > >> Ahoy, > >> > >> As previously discussed the documentation commands have now been > >> grouped under a `C-c C-d' prefix. The commands are: > >> > >> a - apropos > >> p - apropos-package > >> z - apropos-all > >> d - describe-symbol > >> f - describe-function > >> h - hyperspec lookup > >> ~ - hyperspec lookup of a format character > >> > >> The final keystroke is bound both unmodified and with control, so > >> both `C-c C-d a' and `C-c C-d C-a' will make an apropos search. > > > > Why? It seems silly to eat up twice as many key combos as we need to. > > (I haven't been following the discussion of this closely so if there > > was some reason that everyone already agreed was righteous you can > > just tell me so.) > > Well, at least I like the last key unmodified, and at least Bill > Clementson likes the last key modified. I figured that since they're > already on their own documentation-commands prefix (C-c C-d) that > we're unlikely to run out of keys, so we can just make everyone happy > in this respect. > > There's still meta if we do end up running out of letters. :-) We had a similar discussion on the ilisp-devel list last year when we were changing key bindings so that they conformed to fsf standards. Some people (including me at that time) liked the unmodified last key and some preferred a modified last key for key chords that had to use 3 keys. Andreas Fuchs summed up his rationale for the modified key option in this posting: http://sourceforge.net/mailarchive/forum.php?thread_id=2389721&forum_id=5779 A few other people agreed with Andreas - I tried out both combinations and came around to Andreas' POV as well. However, some people (Luke included) prefer the other approach. Remember, this is only an issue with the 3-key chords as the 2-key chords need to be bound to either a control key (for each key) or a special character (for the second key) in order to conform to fsf major mode standards. Since we only use 3-key chords for certain "groups" of commands (e.g. -- documentation), we are probably not going to run out of 3-key sequences any time soon ;-) So, it probably doesn't hurt to bind to both sets of keys as it will keep everybody happy :-) - Bill

1 0

[slime-devel] pdfTeX failage on Slime manual
by Brian Downing 29 Jun '04

29 Jun '04

I just ran into the same stack overflow that's happening in Slime's manual on some of my own stuff. It turns out what my problem was is that pdfTeX really does not like whatever it is that TeXinfo generates for @macro output living in chapter and section titles. If I replaced my @FOOBAR{} with @acronym{FOOBAR} - only in chapter and section titles - it ran through fine. If I had to make a wild guess why this fails on pdfTeX and not plain TeX, it would be that pdfTeX is stuffing the section headers into the PDF file as the bookmark tree. -bcd -- *** Brian Downing <bdowning at lavos dot net>

2 1