slime-devel February 2010

slime-devel@common-lisp.net

21 participants
36 discussions

[slime-devel] indentation problem
by Leo 23 Feb '10

23 Feb '10

Hello, Somehow some comments (line 10 and 22) in the attached file don't indent correctly. Here are steps to reproduce the problem: 1. Emacs -q 2. (slime-setup '(slime-fancy slime-banner slime-asdf)) 3. M-x slime 4. then C-c M-q on the function in the attached file. Tested on: GNU Emacs 23.1.92.2 (x86_64-apple-darwin10.2.0, Carbon Version 1.6.0 AppKit 1038.25) of 2010-02-03 SLIME: 2010-02-20 Best wishes, Leo

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 23 Feb '10

23 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1992 slime/ChangeLog:1.1995 --- slime/ChangeLog:1.1992 Sat Feb 20 14:15:59 2010 +++ slime/ChangeLog Mon Feb 22 16:43:30 2010 @@ -1,3 +1,58 @@ +2010-02-22 Tobias C. Rittweiler <tcr(a)freebits.de> + + Make swank-ecl.lisp work with latest ECL Git HEAD. + + * swank-ecl.lisp (assert-TAGS-file): Simplified. + (assert-source-directory): New helper. + (c-function-p): New helper. + (c-function): Type based on above. + (source-location): Move bits from FIND-DEFINITIONS-FOR-TYPE to + this function. Use CONVERTING-ERRORS-TO-ERROR-LOCATION. + (find-definitions-for-type): Simplified by using it. + +2010-02-22 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-backend.lisp (converting-errors-to-error-location): Moved + here from swank-sbcl.lisp so other backends can make use of it, too. + + * swank-sbcl.lisp: The above macro was called slightly + differently, so update uses accordingly. + +2010-02-22 Tobias C. Rittweiler <tcr(a)freebits.de> + + Make M-. be able to jump right into the C source for ECL. + + Because it's based on TAGS files, M-. and M-* will DTRT once in a + .c file. + + * swank-ecl.lisp (assert-TAGS-file): New helper. + (classify-definition-name): Ditto. + (find-definitions-for-type): Ditto. Understands Lisp and C + functions. + (find-definitions): Use them. + (source-location): New helper. Extracted from FIND-SOURCE-LOCATION. + (find-source-location): Use it. + (swank-compile-string): Only try to delete temporary files if they + exist. + +2010-02-22 Tobias C. Rittweiler <tcr(a)freebits.de> + + Make it possible for SWANK backends to specify locations based on + a TAGS file. + + * slime.el (slime-postprocess-xref, slime-postprocess-xrefs): New + functions. They convert TAGS based locations from SWANK into + file+position based locations because the rest of Slime expects + and works with those. + (slime-find-definitions): Call slime-postprocess-xrefs. + (slime-xref): Ditto. + (slime-etags-to-locations): The function which does the actual + conversion. Extracted from `slime-etags-definitions'. + (slime-etags-definitions): Use it. + + * swank-backend (defimplementation): Add implicit BLOCK. + (:etags-file, :tag): Mentioned for possible values in :LOCATION. + 2010-02-20 Tobias C. Rittweiler <tcr(a)freebits.de> More work on ECL's swank-backend.

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 21 Feb '10

21 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1988 slime/ChangeLog:1.1992 --- slime/ChangeLog:1.1988 Thu Feb 18 13:46:02 2010 +++ slime/ChangeLog Sat Feb 20 14:15:59 2010 @@ -1,3 +1,46 @@ +2010-02-20 Tobias C. Rittweiler <tcr(a)freebits.de> + + More work on ECL's swank-backend. + + * swank-ecl.lisp (accept-connection): Handle :buffering, and + :external-format. + (external-format): New helper. + (find-external-format): Make sure to only return :default in case + ECL was built with --disable-unicode; it'll barf on anything else. + (socket-fd): Add two-way-stream case due to recent changes in ECL. + (make-file-location, make-buffer-location): New helpers. + (condition-location): Use them. + (swank-compile-file): Handle :external-format. + (compile-from-stream): Deleted. Slurped into swank-compile-string. + (swank-compile-string): Call SI:MKSTEMP correctly. Make sure to + also remove fasl file, not just source file. + (grovel-docstring-for-arglist): Do not look at "Syntax:" entry in + docstring because that was a kludge. Upstream ECL should be + modified instead. + (in-swank-package-p, is-swank-source-p, is-ignorable-fun-p): + Commented out. They make debugging ECL's swank-backend harder. + +2010-02-20 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-loader.lisp (*architecture-features*): Add :PENTIUM3 and + :PENTIUM4; they're used by ECL. + (handle-swank-load-error): Renamed from HANDLE-LOADTIME-ERROR. Use + *FASL-DIRECTORY* rather than (DEFAULT-FASL-DIR). Parametrize + context to differentiate b/w compilation/loading. + (compile-files): Adapted accordingly. Also make sure that an error + is signaled in case COMPILE-FILE returns NIL as primary result. + +2010-02-20 Stas Boukarev <stassats(a)gmail.com> + + * swank-ccl.lisp: Remove outdated comment at the beginning + since it may be misleading. + + * slime.el: In minor-mode-alist for slime-popup-buffer-mode, + run slime-modeline-string only if slime-mode isn't active, because + slime-mode runs slime-modeline-string from minor-mode-alist too + resulting in duplicate modeline strings. + Reported by Leo Liu. + 2010-02-18 Mark Harig <idirectscm(a)aim.com> The compiler warns about various stuff. Fix some of it. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.347 slime/contrib/ChangeLog:1.348 --- slime/contrib/ChangeLog:1.347 Fri Feb 19 05:38:07 2010 +++ slime/contrib/ChangeLog Sat Feb 20 13:46:24 2010 @@ -1,3 +1,8 @@ +2010-02-20 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-fancy.el: Call init function for fancy + inspector. Necessary due to 2010-02-15. + 2010-02-19 Stas Boukarev <stassats(a)gmail.com> * slime-fuzzy.el (slime-fuzzy-choices-buffer): Make connection buffer-local, otherwise

1 0

[slime-devel] xref callers fails for LispWorks
by Terje Norderhaug 20 Feb '10

20 Feb '10

There seems to be a bug in the LispWorks backend when it comes to getting the callers of a function. To replicate, evaluate the following in the swank package: ? (xref :callers "LIST-CALLERS") Error in LispWorks: Undefined function T in form (SYMBOL-FUNCTION T). ? (list-callers 'list-callers) Error in LispWorks: Undefined function T in form (SYMBOL-FUNCTION T). This is with LispWorks 6.0.0 on a MacOSX Server 10.6.2 (Intel) running Swank 2010-02-14. -- Terje Norderhaug terje(a)in-progress.com

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 20 Feb '10

20 Feb '10

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.346 slime/contrib/ChangeLog:1.347 --- slime/contrib/ChangeLog:1.346 Thu Feb 18 13:46:07 2010 +++ slime/contrib/ChangeLog Fri Feb 19 05:38:07 2010 @@ -1,3 +1,8 @@ +2010-02-19 Stas Boukarev <stassats(a)gmail.com> + + * slime-fuzzy.el (slime-fuzzy-choices-buffer): Make connection buffer-local, otherwise + `swank:fuzzy-completion-selected' will be sent to the default connection. + 2010-02-17 Helmut Eller <heller(a)common-lisp.net> Fix bugs when *inferior-buffers* doesn't exist.

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 19 Feb '10

19 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1987 slime/ChangeLog:1.1988 --- slime/ChangeLog:1.1987 Wed Feb 17 12:04:59 2010 +++ slime/ChangeLog Thu Feb 18 13:46:02 2010 @@ -1,3 +1,10 @@ +2010-02-18 Mark Harig <idirectscm(a)aim.com> + + The compiler warns about various stuff. Fix some of it. + + * slime.el (slime-add-local-hook): Remove support for Emacs 20. + (sldb-recenter-region): Use forward-line; not next-line. + 2010-02-17 Helmut Eller <heller(a)common-lisp.net> Add a command to eval stuff in the inspector. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.345 slime/contrib/ChangeLog:1.346 --- slime/contrib/ChangeLog:1.345 Wed Feb 17 12:04:59 2010 +++ slime/contrib/ChangeLog Thu Feb 18 13:46:07 2010 @@ -33,13 +33,6 @@ (slime-search-exports-in-defpackage): take #:symbol and :symbol into account too. -2010-02-17 Helmut Eller <heller(a)common-lisp.net> - - Fix bugs when *inferior-buffers* doesn't exist. - - * inferior-slime.el (inferior-slime-start-transcript) - (inferior-slime-stop-transcript): Make sure the buffer exists. - 2010-01-14 Stas Boukarev <stassats(a)gmail.com> * slime-repl.el: Revert the previous change because it

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 18 Feb '10

18 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1984 slime/ChangeLog:1.1987 --- slime/ChangeLog:1.1984 Tue Feb 16 06:28:18 2010 +++ slime/ChangeLog Wed Feb 17 12:04:59 2010 @@ -1,3 +1,58 @@ +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Add a command to eval stuff in the inspector. + + * slime.el (slime-inspector-eval): New command. + (slime-inspector-mode-map): Bind it to 'e'. + * swank.lisp (inspector-eval): New function. + * swank-backend.lisp (eval-context): New function. + * swank-cmucl.lisp (eval-context): Implement it. + +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Point-entered hooks in the xref buffer don't work so well. + Use forward/backward commands instead. + + * slime.el (slime-xref-next-line, slime-xref-prev-line): New + commands. + (slime-xref-show-location): New aux function. + (slime-xref-mode-map): Remap up/down to next/prev location + commands. Make SPC and RET do the same. + +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Select the xref buffer. + I never could get used to the next/prev cycling. + + * slime.el (slime-with-xref-buffer): Select the buffer. + (slime-insert-xrefs): Add point-entered hook to automatically + display the current location. + (slime-xref-entered): New function. + (slime-show-buffer-position): Don't use reposition-window which + seems very slow and doesn't even do a particularly good job. The + new heuristic is much cruder but faster. + (slime-xref-mode-map): Remove apparently redundant bindings for + RET. + +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Get rid or snapshots. + + * slime.el (slime-narrowing-configuration, slime-emacs-snapshot) + (slime-current-narrowing-configuration) + (slime-set-narrowing-configuration, slime-current-emacs-snapshot) + (slime-set-emacs-snapshot) + (slime-current-emacs-snapshot-fingerprint, slime-frame-windows): + Deleted. Update call-sites accordingly. + +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Test suite hangs for CCL and CMUCL. + + * slime.el ([test] compile-defun): Remove those compile-time read + errors for which some implementations invoke the debugger. + (sexp-at-point.1): Remove failing tests + 2010-02-16 Tobias C. Rittweiler <tcr(a)freebits.de> * swank-loader.lisp: Compile files on ECL, too. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.343 slime/contrib/ChangeLog:1.345 --- slime/contrib/ChangeLog:1.343 Mon Feb 15 16:42:37 2010 +++ slime/contrib/ChangeLog Wed Feb 17 12:04:59 2010 @@ -1,3 +1,10 @@ +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Fix bugs when *inferior-buffers* doesn't exist. + + * inferior-slime.el (inferior-slime-start-transcript) + (inferior-slime-stop-transcript): Make sure the buffer exists. + 2010-02-15 Tobias C. Rittweiler <tcr(a)freebits.de> * slime-asdf.el, slime-autodoc.el, slime-c-p-c.el, @@ -26,6 +33,13 @@ (slime-search-exports-in-defpackage): take #:symbol and :symbol into account too. +2010-02-17 Helmut Eller <heller(a)common-lisp.net> + + Fix bugs when *inferior-buffers* doesn't exist. + + * inferior-slime.el (inferior-slime-start-transcript) + (inferior-slime-stop-transcript): Make sure the buffer exists. + 2010-01-14 Stas Boukarev <stassats(a)gmail.com> * slime-repl.el: Revert the previous change because it

2 4

[slime-devel] Daily ChangeLog diff
by Helmut Eller 17 Feb '10

17 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1982 slime/ChangeLog:1.1984 --- slime/ChangeLog:1.1982 Mon Feb 15 16:42:37 2010 +++ slime/ChangeLog Tue Feb 16 06:28:18 2010 @@ -1,3 +1,16 @@ +2010-02-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-loader.lisp: Compile files on ECL, too. + +2010-02-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + Pimp my swank. + + * swank-ecl.lisp: We depend on ECL 10.2.1 which is not released + yet -- you need git/cvs HEAD. Added :spawn, and :fd-handler as + communication-style (Thanks to Ram Krishnan), improve compilation + hooks so highligting of warnings works, + various cleanup. + 2010-02-15 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-load-contribs): Do not call SWANK-REQUIRE

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 16 Feb '10

16 Feb '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1981 slime/ChangeLog:1.1982 --- slime/ChangeLog:1.1981 Sun Feb 14 13:26:01 2010 +++ slime/ChangeLog Mon Feb 15 16:42:37 2010 @@ -1,3 +1,10 @@ +2010-02-15 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-load-contribs): Do not call SWANK-REQUIRE + asynchronously, if host Lisp uses :SPAWN that may result in the + attempt to load in code concurrently -- the host Lisp may not + support that. + 2010-02-14 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-attempt-connection): Fix typo. Thanks to Mark Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.342 slime/contrib/ChangeLog:1.343 --- slime/contrib/ChangeLog:1.342 Sun Jan 31 15:17:27 2010 +++ slime/contrib/ChangeLog Mon Feb 15 16:42:37 2010 @@ -1,3 +1,15 @@ +2010-02-15 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-asdf.el, slime-autodoc.el, slime-c-p-c.el, + slime-clipboard.el, slime-fancy-inspector.el, slime-fuzzy,el, + slime-indentation.el, slime-motd.el, + slime-presentation-streams.el, slime-presentations.el, + slime-sbcl-exts, slime-snapshot.el, slime-sprof.el: + + Make sure that contrib code does not try to load in swank code + asynchronously but use `slime-require' instead; also make sure to + move the `slime-require' into the contribs' init function. + 2010-01-31 Stas Boukarev <stassats(a)gmail.com> * slime-repl.el (slime-repl-mode): Don't do

1 0

[slime-devel] slime/ACL - ending a session?
by Duke Normandin 15 Feb '10

15 Feb '10

Hey list... New to CL (and Allegro) and SLIME -- fair-to-middling with emacs. Just installed Allegro / SLIME on my OS X box. Using it from OS X Terminal, i.e.: emacs/M-x slime Seems to work OK. One problem: How do I gracefully (and correctly) terminate a LISP hacking session from within SLIME? SO far, I've just been doing a ^X^C to quit emacs, but I now know that that can't be correct because I noticed last night ( after my forehead hit the keyboard) that I had 2 PIDs alive for `alisp'. SO I'm thinking that I'm not doing something correctly - duh! Clue me in, please. TIA.. -- duke

4 4

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

slime-devel February 2010