June 2004 - slime-devel - mailman3.common-lisp.net

[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

[slime-devel] Feature Request: Generic Function Browser
by Thomas Schilling 29 Jun '04

29 Jun '04

Is there already a SLIME feature like a "Generic Function Browser"? If so can someone tell how to invoke it? Here's what I'd expect from it: - list all methods of a generic function in a separate buffer - by navigating to a function and pressing 'r' it gets removed (like with REMOVE-METHOD) - 'q' closes the window - there might be some other nice things possible but this is what's be most important, cause every time I modify a method the old method is still there and it's awfully complicated to remove it then. hope someone can help -ts -- ,, \../ / <<< The LISP Effect |_\\ _==__ __ | |bb| | _________________________________________________

2 1

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 29 Jun '04

29 Jun '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.431 slime/ChangeLog:1.436 --- slime/ChangeLog:1.431 Sun Jun 27 08:23:54 2004 +++ slime/ChangeLog Mon Jun 28 14:05:28 2004 @@ -1,3 +1,77 @@ +2004-06-28 Luke Gorrie <luke(a)bluetail.com> + + * slime.el (slime-doc-map): New keymap for documentation + commands. These all use the `C-c C-d' prefix, followed by: + 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. + +2004-06-28 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank-allegro.lisp (nth-frame): Skip frames where + frame-visible-p is false. + + * slime.el (slime-buffer-package): Return the cached package if + can't something more sensible; this reverts a previous change. + The Lisp side will now fall back to an existing package if the one + supplied by Emacs doesn't exist. Using the cached version is also + necessary for some commands in the apropos buffer. + (sldb-insert-frame): Set the default-action property; pressing RET + on frame lines now shows/hides details. + (sldb-toggle-details): Preserve the current column. + (slime-inspector-buffer, slime-saved-window-config) + (slime-inspector-quit): Save and restore the window configuration. + (slime-highlight-suppressed-forms, slime-search-suppressed-forms): + Display expressions with reader conditionals (#+/#-) where the + test is false in font-lock-comment-face. No implemented for + XEmacs. + (repl-return): New test. + +2004-06-28 Luke Gorrie <luke(a)bluetail.com> + + * slime.el: Events in the *slime-events* buffer are now exact + on-the-wire messages, without including e.g. Elisp continuation + functions. This is easier for debugging I think. + + * swank-allegro.lisp (compute-backtrace): Only include frames + satisfying `debugger:frame-visible-p'. I did this as a lame + workaround for a problem where `output-frame' was segfaulting on + certain frames, and those frames happened not to be visible-p. I + don't know if it really fixes anything. + + * hyperspec.el (common-lisp-hyperspec-format): This command now + works at the end of the buffer, fixed `char-after' usage as + suggested by Johan Bockgård. + +2004-06-28 Christophe Rhodes <csr21(a)cam.ac.uk> + + * hyperspec.el: add support for issue cross-reference lookups, + strongly inspired by hyperspec symbol lookup. + (common-lisp-hyperspec-issuex-table, + common-lisp-hyperspec-issuex-symbols): new variables + (common-lisp-issuex): new function + + * slime.el (sldb-format-reference-node, sldb-lookup-reference): + (sldb-reference-properties): use new support for issue lookups + to support :ansi-cl :issue reference types. + + * hyperspec.el: add support for glossary lookups. + (common-lisp-glossary-fun): new variable + (common-lisp-glossary-4.0, common-lisp-glossary-6.0): new functions + + * slime.el (sldb-format-reference-node, sldb-lookup-reference): + (sldb-reference-properties): use new support for glossary lookupts + to support :ansi-cl :glossary reference types. + 2004-06-27 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * doc/slime.texi: Remove macros from chapter and section headings @@ -59,7 +133,7 @@ 2004-06-24 Brian Downing <bdowning(a)lavos.net> - * slime.el (slime-repl-send-input): Fixed a subtle difference in + * slime.el (slime-repl-send-input): Fixed a subtle difference in sending input to the Lisp introduced in 1.316. The newline was not getting sent, resulting in the Lisp constantly asking for more read data. I believe the code has been adjusted to behave the

1 0

[slime-devel] Allegro 6.2 problem
by Peter Seibel 28 Jun '04

28 Jun '04

This may be related to the thing Edi mentioned the other day. Anyway, I start SLIME using Allegro 6.2 on Linux. Initially I'm in the directory where test-debug.lisp lives: ; SLIME: The Superior Lisp Interaction Mode for Emacs CL-USER> CL-USER> (compile-file "test-debug.lisp") ;;; Compiling file test-debug.lisp ;;; Writing fasl file test-debug.fasl ;;; Fasl write complete #p"test-debug.fasl" NIL NIL So far so good. Now I ,change-directory up one directory and try again. Which should fail since the file won't be found. But it doesn't just fail; in the SLIME repl window I get the following: CL-USER> CL-USER> (compile-file "test-debug.lisp") Error: Received signal number 11 (Segmentation violation) [condition type: SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL] Restart actions (select using :continue): 0: Prompt for a different filename to compile 1: Abort handling SLIME request. 2: Abort entirely from this process. [Current process: repl-thread] [1] CL-USER(1): Typing Allegro commands (e.g. :cont 1) doesn't seem to do anything. C-c C-c gets me into sldb where I can use one of the ABORT restarts to get back to the main REPL. But this seems wrong. -Peter -- Peter Seibel peter(a)javamonkey.com Lisp is the red pill. -- John Fraser, comp.lang.lisp

3 5