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

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 20 Dec '06

20 Dec '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1001 slime/ChangeLog:1.1002 --- slime/ChangeLog:1.1001 Sun Nov 12 06:18:57 2006 +++ slime/ChangeLog Sun Nov 19 16:25:15 2006 @@ -1,3 +1,35 @@ +2006-11-19 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-compile-file): Let the Lisp side choose the + coding system. + (slime-coding): Deleted. + + * swank.lisp (compile-file-for-emacs): Use guess-external-format. + + * swank-backend.lisp (find-external-format) + (guess-external-format): New. + (swank-compile-file): The external-format argument is now a + backend specific value returned by find-external-format. + + * swank-cmucl.lisp, swank-sbcl.lisp, swank-clisp, + swank-lispworks.lisp, swank-allegro.lisp, swank-corman.lisp, + swank-ecl.lisp, swank-scl.lisp, swank-abcl.lisp, swank-openmcl: + Update implementations accordingly. + + * swank-source-file-cache.lisp (read-file): Use guess-external-format. + + * swank.lisp (*swank-wire-protocol-version*): Is now initialized + by the loader. + (wire-protocol-version): Removed, because it contained a reference + to swank-loader::*source-directory*. + + * slime.el (slime-set-connection-info): On protocol version + mismatch, ask the user how to proceed. + (slime-protocol-version): New variable. Initialize it at compile + time to detect stale elc files. + + * swank-loader.lisp (load-swank): Set the protocol version. + 2006-11-12 Marco Baringer <mb(a)bese.it> * slime.el (slime-make-tramp-file-name): Added (require 'tramp) @@ -10695,3 +10727,5 @@ ;; Local Variables: ;; coding: latin-1 ;; End: + +This file has been placed in the public domain.

4 4

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 20 Dec '06

20 Dec '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1023 slime/ChangeLog:1.1024 --- slime/ChangeLog:1.1023 Mon Dec 18 11:45:15 2006 +++ slime/ChangeLog Tue Dec 19 06:11:43 2006 @@ -1,3 +1,62 @@ +2006-12-19 Attila Lendvai <attila.lendvai(a)gmail.com> + + * HACKING: + Added useful init.el piece into HACKING about update-change-log + + * swank.lisp: + In all-slots-for-inspector pad slot names to be equal length, so the result is more readable + + * slime.el: + Fix slime-insert-presentation to handle multi-line presentations better (use insert-rectangle) + + * swank.lisp: + Properly bind *sldb-printer-bindings* and turn off right margin while printing stuff in sldb + + * slime.el: Smarten up the sldb heuristic that drops swank frames + + * swank-allegro.lisp, swank-backend.lisp, swank-openmcl.lisp, swank-sbcl.lisp, swank.lisp: + Added hash-table-weakness and use it in hash-table-inspecting + + * swank.lisp: + Hashtable inspecting: added [clear hashtable] and [remove entry] actions + + * slime.el, swank.lisp: + FIX dwim inspecting to handle (setf some-fun) functions, too + + * slime.el: FIX: slime-sexp-at-point for foo::|bar::baz| + + * slime.el: + FIX: Properly keep track of slime-buffer-package in the inspector + + * swank.lisp: Small: get rid of notes and warnings + + * slime.el, swank.lisp: + Added dwim-mode to slime-inspect that tries to be smart unless prefixed + + * slime.el: + Make slime-fuzzy-complete-symbol the default in the belife that it's better for new users + + * swank.lisp: + Add (expt 1.2 length) higher scores for longer matches in fuzzy completion. + + A good example: puts "make-instance" before "make-string-input-stream" while completing "make-ins" + + * slime.el: Set slime-fuzzy-completion-in-place enabled by default + + * slime.el: + Added (cons row col) addressing to slime-open-inspector, use in slime-inspector-operate-on-point + + * slime.el: + FIX: operate the inspector in the debug thread when started from sldb + + * slime.el: + Convert some inspector defuns to defun* and use keywords. Other minor cleanups. + +2006-12-18 Marco Baringer <mb(a)bese.it> + + * slime.el (slime-region-for-defun-at-point): end-of-defun and + beginning-of-defun modify match-data, so we should save it. + 2006-12-18 Marco Baringer <mb(a)bese.it> * slime.el (slime-region-for-defun-at-point): end-of-defun and

1 0

[slime-devel] Fuzzy completion in-place on GNU Emacs
by Ury Marshak 20 Dec '06

20 Dec '06

The in-place completion that was now made the default didn't work for me in GNU emacsen 21 and 22 (cvs head) under Debian and Windows. (Didn't work means that the SPC, RET and other such keys in the target buffer didn't end the completion and only performed their normal function instead) After some investigation it seems that the keymap of slime-fuzzy-target-buffer-completions-mode wasn't overriding the standard slime keymap. There's a pretty easy fix for that - move the minor mode definition for slime-fuzzy-target-buffer-completions-mode after the definition of slime-mode. (And also add <return> to the direct bindings in slime-target-buffer-fuzzy-completions-map.) I tried it on gnu emacs 21 on Debian and 22 on Debian and Windows and it seems to fix the issue for me. Unfortunately I don't have a working xemacs at the moment so I can't test it there. The patch is trivial, but I'll attach it just in case. Cheers, Ury Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.719 diff -u -r1.719 slime.el --- slime.el 19 Dec 2006 10:55:24 -0000 1.719 +++ slime.el 19 Dec 2006 18:19:44 -0000 @@ -576,26 +576,13 @@ (call-interactively 'isearch-forward))) ;; some unconditional direct bindings - (dolist (key (list (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) + (dolist (key (list (kbd "<return>") (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) (define-key map key 'slime-fuzzy-select-and-process-event-in-target-buffer))) map ) "Keymap for slime-target-buffer-fuzzy-completions-mode. This will override the key bindings in the target buffer temporarily during completion.") -(define-minor-mode slime-fuzzy-target-buffer-completions-mode - "This minor mode is intented to override key bindings during fuzzy -completions in the target buffer. Most of the bindings will do an implicit select -in the completion window and let the keypress be processed in the target buffer." - nil - nil - slime-target-buffer-fuzzy-completions-map) - -(add-to-list 'minor-mode-alist - '(slime-fuzzy-target-buffer-completions-mode - " Fuzzy Target Buffer Completions")) - - ;;;;; slime-mode (define-minor-mode slime-mode "\\<slime-mode-map>\ @@ -638,6 +625,20 @@ "The Lisp package to show in the modeline. This is automatically updated based on the buffer/point.")) +(define-minor-mode slime-fuzzy-target-buffer-completions-mode + "This minor mode is intented to override key bindings during fuzzy +completions in the target buffer. Most of the bindings will do an implicit select +in the completion window and let the keypress be processed in the target buffer." + nil + nil + slime-target-buffer-fuzzy-completions-map) + +(add-to-list 'minor-mode-alist + '(slime-fuzzy-target-buffer-completions-mode + " Fuzzy Target Buffer Completions")) + + + (defun slime-update-modeline-package () (ignore-errors (when (and slime-update-modeline-package

1 1

[slime-devel] slime-eval-sexp?
by Samium Gromoff 19 Dec '06

19 Dec '06

(defun slime-eval-sexp () (interactive) (mark-sexp) (slime-eval-region (region-beginning) (region-end))) What do you think about this? regards, Samium Gromoff

2 3

[slime-devel] Is bridge.el part of Emacs?
by Leo 19 Dec '06

19 Dec '06

Bridge.el states: ,---- | ;;; This file is part of GNU Emacs. `---- But I can't find it in Emacs 22. -- Leo <sdl.web AT gmail.com> (GPG Key: 9283AA3F)

1 0

[slime-devel] Fuzzy completion in-place on GNU Emacs
by Ury Marshak 19 Dec '06

19 Dec '06

The in-place completion that was now made the default didn't work for me in GNU emacsen 21 and 22 (cvs head) under Debian and Windows. (Didn't work means that the SPC, RET and other such keys in the target buffer didn't end the completion and only performed their normal function instead) After some investigation it seems that the keymap of slime-fuzzy-target-buffer-completions-mode wasn't overriding the standard slime keymap. There's a pretty easy fix for that - move the minor mode definition for slime-fuzzy-target-buffer-completions-mode after the definition of slime-mode. (And also add <return> to the direct bindings in slime-target-buffer-fuzzy-completions-map.) I tried it on gnu emacs 21 on Debian and 22 on Debian and Windows and it seems to fix the issue for me. Unfortunately I don't have a working xemacs at the moment so I can't test it there. The patch is trivial, but I'll attach it just in case. Cheers, Ury Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.719 diff -u -r1.719 slime.el --- slime.el 19 Dec 2006 10:55:24 -0000 1.719 +++ slime.el 19 Dec 2006 18:19:44 -0000 @@ -576,26 +576,13 @@ (call-interactively 'isearch-forward))) ;; some unconditional direct bindings - (dolist (key (list (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) + (dolist (key (list (kbd "<return>") (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) (define-key map key 'slime-fuzzy-select-and-process-event-in-target-buffer))) map ) "Keymap for slime-target-buffer-fuzzy-completions-mode. This will override the key bindings in the target buffer temporarily during completion.") -(define-minor-mode slime-fuzzy-target-buffer-completions-mode - "This minor mode is intented to override key bindings during fuzzy -completions in the target buffer. Most of the bindings will do an implicit select -in the completion window and let the keypress be processed in the target buffer." - nil - nil - slime-target-buffer-fuzzy-completions-map) - -(add-to-list 'minor-mode-alist - '(slime-fuzzy-target-buffer-completions-mode - " Fuzzy Target Buffer Completions")) - - ;;;;; slime-mode (define-minor-mode slime-mode "\\<slime-mode-map>\ @@ -638,6 +625,20 @@ "The Lisp package to show in the modeline. This is automatically updated based on the buffer/point.")) +(define-minor-mode slime-fuzzy-target-buffer-completions-mode + "This minor mode is intented to override key bindings during fuzzy +completions in the target buffer. Most of the bindings will do an implicit select +in the completion window and let the keypress be processed in the target buffer." + nil + nil + slime-target-buffer-fuzzy-completions-map) + +(add-to-list 'minor-mode-alist + '(slime-fuzzy-target-buffer-completions-mode + " Fuzzy Target Buffer Completions")) + + + (defun slime-update-modeline-package () (ignore-errors (when (and slime-update-modeline-package

1 0

[slime-devel] Fuzzy completion in-place on GNU Emacs
by Ury Marshak 19 Dec '06

19 Dec '06

The in-place completion that was now made the default didn't work for me in GNU emacsen 21 and 22 (cvs head) under Debian and Windows. (Didn't work means that the SPC, RET and other such keys in the target buffer didn't end the completion and only performed their normal function instead) After some investigation it seems that the keymap of slime-fuzzy-target-buffer-completions-mode wasn't overriding the standard slime keymap. There's a pretty easy fix for that - move the minor mode definition for slime-fuzzy-target-buffer-completions-mode after the definition of slime-mode. (And also add <return> to the direct bindings in slime-target-buffer-fuzzy-completions-map.) I tried it on gnu emacs 21 on Debian and 22 on Debian and Windows and it seems to fix the issue for me. Unfortunately I don't have a working xemacs at the moment so I can't test it there. The patch is trivial, but I'll attach it just in case. Cheers, Ury Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.719 diff -u -r1.719 slime.el --- slime.el 19 Dec 2006 10:55:24 -0000 1.719 +++ slime.el 19 Dec 2006 18:19:44 -0000 @@ -576,26 +576,13 @@ (call-interactively 'isearch-forward))) ;; some unconditional direct bindings - (dolist (key (list (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) + (dolist (key (list (kbd "<return>") (kbd "RET") (kbd "<SPC>") "(" ")" "[" "]")) (define-key map key 'slime-fuzzy-select-and-process-event-in-target-buffer))) map ) "Keymap for slime-target-buffer-fuzzy-completions-mode. This will override the key bindings in the target buffer temporarily during completion.") -(define-minor-mode slime-fuzzy-target-buffer-completions-mode - "This minor mode is intented to override key bindings during fuzzy -completions in the target buffer. Most of the bindings will do an implicit select -in the completion window and let the keypress be processed in the target buffer." - nil - nil - slime-target-buffer-fuzzy-completions-map) - -(add-to-list 'minor-mode-alist - '(slime-fuzzy-target-buffer-completions-mode - " Fuzzy Target Buffer Completions")) - - ;;;;; slime-mode (define-minor-mode slime-mode "\\<slime-mode-map>\ @@ -638,6 +625,20 @@ "The Lisp package to show in the modeline. This is automatically updated based on the buffer/point.")) +(define-minor-mode slime-fuzzy-target-buffer-completions-mode + "This minor mode is intented to override key bindings during fuzzy +completions in the target buffer. Most of the bindings will do an implicit select +in the completion window and let the keypress be processed in the target buffer." + nil + nil + slime-target-buffer-fuzzy-completions-map) + +(add-to-list 'minor-mode-alist + '(slime-fuzzy-target-buffer-completions-mode + " Fuzzy Target Buffer Completions")) + + + (defun slime-update-modeline-package () (ignore-errors (when (and slime-update-modeline-package

1 0

[slime-devel] Isolating small behavior changes
by Jonathon McKitrick 19 Dec '06

19 Dec '06

Suppose there was a tiny behavior change in SLIME that I did not see documented in ChangeLog. How would I go about tracking down the cause? Specifically, after compiling a system, a window opens showing the compiler messages. This window used to close with just a 'q' keystroke. This no longer works the same way. I'd like to restore the old behavior. Would someone point me in the right direction? Jonathon McKitrick -- My other computer is your Windows box.

4 9

[slime-devel] presentations and speed
by Attila Lendvai 19 Dec '06

19 Dec '06

dear list, recently i've added a present-in-emacs and using that i've implemented a slime-repl-log-appender in arnesi. it's cool to be able to inspect the objects that were logged, but it made (x)emacs dog slow. the cause is this: (when slime-repl-enable-presentations ;; Respect the syntax text properties of presentations. (set (make-local-variable 'parse-sexp-lookup-properties) t)) which is enable to be able to handle the <>'s in presentations as ()'s but only in presentations. this mail is more of a fyi then a real question, but if there's anyone who has an idea how to implement this feature without enabling 'parse-sexp-lookup-properties then it would be cool to have both. -- - attila "- The truth is that I've been too considerate, and so became unintentionally cruel... - I understand. - No, you don't understand! We don't speak the same language!" (Ingmar Bergman - Smultronstället)

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 19 Dec '06

19 Dec '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1022 slime/ChangeLog:1.1023 --- slime/ChangeLog:1.1022 Fri Dec 15 07:40:47 2006 +++ slime/ChangeLog Mon Dec 18 11:45:15 2006 @@ -1,3 +1,9 @@ +2006-12-18 Marco Baringer <mb(a)bese.it> + + * slime.el (slime-region-for-defun-at-point): end-of-defun and + beginning-of-defun modify match-data, added a save-match-data to + prevent this. + 2006-12-15 Edi Weitz <edi(a)agharta.de> * swank-lispworks.lisp (make-weak-key-hash-table): Weak hash

1 0