March 2010 - slime-devel - mailman3.common-lisp.net

[slime-devel] Daily ChangeLog diff
by Helmut Eller 12 Mar '10

12 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2033 slime/ChangeLog:1.2035 --- slime/ChangeLog:1.2033 Wed Mar 10 06:30:22 2010 +++ slime/ChangeLog Thu Mar 11 04:05:50 2010 @@ -1,5 +1,20 @@ 2010-03-10 Tobias C. Rittweiler <tcr(a)freebits.de> + * swank.lisp (signal-interrupt): Removed. + (interrupt-worker-thread): Slurp in definition of + signal-interrupt. No need for invoke-or-queue-interrupt twice in + case we do not use threads. Thus micro-prettification of + backtraces. + +2010-03-10 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-ecl.lisp (source-location): Move call to + TRANSLATE-LOGICAL-PATHNAME from here into MAKE-FILE-LOCATION + because locations-via-annotations may now also involve logical + pathnames. + +2010-03-10 Tobias C. Rittweiler <tcr(a)freebits.de> + * swank-ecl.lisp (*original-sigint-handler*) (install-sigint-handler): Deleted; we directly implement call-with-user-break-handler instead.

1 0

[slime-devel] Bug Report: SB-KERNEL:SIMPLE-PACKAGE-ERROR not invoking the debugger
by Russ Tyndall 11 Mar '10

11 Mar '10

Slime Version: 3/10 from git mirror of CVS SBCL 1.0.36 GNU Emacs 22.1.1 and 23.x.x (checked on two different versions). Using the latest version of slime, executing the following code in the slime repl results in a totally locked lisp that can't be quit or interacted with via slime. There is also nothing of interest in the backing SBCL terminal. (make-package :short-name) (make-package :long-name :nicknames (list :short-name)) Executing these same forms in the sbcl command line repl correctly results in a debugger asking for me to abort being stupid. The last slime event we see is: (:emacs-rex (swank:listener-eval "(make-package :long-name :nicknames (list :short-name))\n") "COMMON-LISP-USER" :repl-thread 6) This is my setup: (slime-setup '(slime-fancy slime-asdf slime-tramp slime-indentation)) Thanks for the great software, SLIME is a pleasure to use, Russ Tyndall

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 11 Mar '10

11 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2032 slime/ChangeLog:1.2033 --- slime/ChangeLog:1.2032 Tue Mar 9 19:02:53 2010 +++ slime/ChangeLog Wed Mar 10 06:30:22 2010 @@ -1,3 +1,12 @@ +2010-03-10 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-ecl.lisp (*original-sigint-handler*) + (install-sigint-handler): Deleted; we directly implement + call-with-user-break-handler instead. + (call-with-user-break-handler): New. Correctly interrupt main + thread instead of newly spawned handle-signal thread on SIGINT. + (make-interrupt-handler): New helper. + 2010-03-09 Stas Boukarev <stassats(a)gmail.com> * swank-ccl.lisp (emacs-inspect function): Print closed over variables @@ -62,7 +71,7 @@ it boils down to SIGNAL-INTERRUPT which uses USE-THREADS-P which needs a connection. (install-fd-handler): Adapted accordingly. - (simple-serve-event): Adapted accordingly. Additionally, remove + (simple-serve-requests): Adapted accordingly. Additionally, remove superfluous WITH-SWANK-PROTOCOL-HANDLER as that's established by WITH-CONNECTION already. (simple-repl): Show "abort inferior lisp" restart only if not a

1 0

[slime-devel] sldb-initial-restart-limit
by Robert Goldman 10 Mar '10

10 Mar '10

The current definition of this custom is as follows: (defcustom sldb-initial-restart-limit 6 "Maximum number of restarts to display initially." :group 'slime-debugger :type 'integer) Reading the way this is used, I believe this could be changed to (defcustom sldb-initial-restart-limit 6 "Maximum number of restarts to display initially. Will always display all restarts if NIL." :group 'slime-debugger :type '(choice integer (const :tag "Show all" nil)) I don't venture to offer this as a patch because I'm not an expert with customize. HtH, r

1 0

[slime-devel] slime-compile-file gets error in ACL 8.2
by Takuo Watanabe 10 Mar '10

10 Mar '10

Dear Slime Developers I am using Slime with the following environment: Slime 2010-03-01 (CVS head) Emacs 22.3.1 (i386-apple-darwin9.8.0, Carbon Version 1.6.0) Allegro CL Professional 8.2 (allegro-8.2a-ics-macosx-x86) Mac OS X (10.6.2), MacBook After upgrading Allegro to 8.2, slime-compile-file (also slime-compile-and-load-file) gets errors like this: ------------------------------------------------------------ `(6 6)' is not of the expected type `NUMBER' [Condition of type TYPE-ERROR] ------------------------------------------------------------ The backtrace of the error looks like: ------------------------------------------------------------ Backtrace: 0: (ERROR TYPE-ERROR :DATUM (6 6) :EXPECTED-TYPE NUMBER ...) 1: (SWANK-BACKEND::LOCATION-FOR-WARNING #<SIMPLE-WARNING @ #x20a331c2>) Locals: CONDITION = #<SIMPLE-WARNING @ #x20a331c2> SWANK-BACKEND::POS = (6 6) NIL = (:FILE "/Users/takuo/foo.lisp") NIL = #<STANDARD-CLASS READER-ERROR> SWANK-BACKEND::LOC = (#P"/Users/takuo/foo.lisp" 6 6) SWANK-BACKEND::FILE = #P"/Users/takuo/foo.lisp" 2: (SWANK-BACKEND::HANDLE-COMPILER-WARNING #<SIMPLE-WARNING @ #x20a331c2>) 3: (SIGNAL #<SIMPLE-WARNING @ #x20a331c2>) ------------------------------------------------------------ The source of this error seems to be the change of the format of condition location (i.e., the value of the :loc property provided with the excl::plist slot in a warning object). For example, in the previous ACL (8.1), the value of :loc property looks like this. ------------------------------------------------------------ CL-USER> (handler-case (compile-file "foo.lisp") (warning (w) (getf (slot-value w 'excl::plist) :loc))) (#P"foo.lisp" . 6) ------------------------------------------------------------ The car part of the value is the file name and the cdr is the warning position in the file. After upgrading to 8.2, the value of the above expression becomes (#P"foo.lisp" 6 6). I don't know about the difference among these two numbers. Anyway, the function location-for-warning in swank-allegro.lisp assumes that the cdr part of the value should be a number. The (kludgey) patch I applied is the following. I'm not sure this is really a solution to the problem. ------------------------------------------------------------ --- swank-allegro.lisp 2010-03-02 12:25:46.000000000 +0900 +++ swank-allegro.lisp.orig 2010-03-02 12:25:24.000000000 +0900 @@ -279,7 +279,7 @@ (destructuring-bind (file . pos) loc (make-location (list :file (namestring (truename file))) - (list :position (1+ (if (listp pos) (car pos) pos)))))) + (list :position (1+ pos))))) (t (make-error-location "No error location available."))))) ------------------------------------------------------------ Sorry, I didn't check any other platforms (Linux, Windows, etc.). I only have a license for the OS X edition. Cheers Takuo Watanabe

3 2

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

10 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2029 slime/ChangeLog:1.2032 --- slime/ChangeLog:1.2029 Tue Mar 9 03:15:05 2010 +++ slime/ChangeLog Tue Mar 9 19:02:53 2010 @@ -1,3 +1,12 @@ +2010-03-09 Stas Boukarev <stassats(a)gmail.com> + + * swank-ccl.lisp (emacs-inspect function): Print closed over variables + in case of closure. + +2010-03-09 Stas Boukarev <stassats(a)gmail.com> + + * slime.el (slime-parse-context): Add defclass. + 2010-03-09 Helmut Eller <heller(a)common-lisp.net> Some more fixes for Allegro @@ -10,6 +19,8 @@ vs. buffer issue. (ldb-code-to-src-loc): Don't use *temp-file-map* before it is declared. + (pc-source-location): Be a bit more fuzzy when searching the + code-location for a pc. 2010-03-08 Stas Boukarev <stassats(a)gmail.com> Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.353 slime/contrib/ChangeLog:1.356 --- slime/contrib/ChangeLog:1.353 Mon Mar 8 11:35:06 2010 +++ slime/contrib/ChangeLog Tue Mar 9 18:26:23 2010 @@ -1,3 +1,19 @@ +2010-03-09 Stas Boukarev <stassats(a)gmail.com> + + * swank-arglists.lisp (arglist-ref): Don't error if a &key name isn't + a :keyword symbol. And also handle non-:keyword keyword parameters. + +2010-03-09 Stas Boukarev <stassats(a)gmail.com> + + * swank-presentations.lisp (inspect-presentation): Throw an error when + trying to access unrecorded object. + (lookup-presented-object-or-lose): Rename from `get-repl-result'. + +2010-03-09 Stas Boukarev <stassats(a)gmail.com> + + * slime-repl.el (slime-call-defun): When on defclass insert + (make-instance 'name). + 2010-03-08 Stas Boukarev <stassats(a)gmail.com> * swank-fancy-inspector.lisp (all-slots-for-inspector): Sort class

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 09 Mar '10

09 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2021 slime/ChangeLog:1.2029 --- slime/ChangeLog:1.2021 Sun Mar 7 11:22:17 2010 +++ slime/ChangeLog Tue Mar 9 03:15:05 2010 @@ -1,3 +1,86 @@ +2010-03-09 Helmut Eller <heller(a)common-lisp.net> + + Some more fixes for Allegro + + * swank-allegro.lisp (function-source-location): Use + xref::object-to-function-name which seems to work better for some + cases. + (fspec-definition-locations): For :top-level-forms return a list + of ((fspec loc)) not just (fspec loc). Also deal with the file + vs. buffer issue. + (ldb-code-to-src-loc): Don't use *temp-file-map* before it is + declared. + +2010-03-08 Stas Boukarev <stassats(a)gmail.com> + + * slime.el (slime-extract-context): Add defvar and defparameter. + +2010-03-08 Helmut Eller <heller(a)common-lisp.net> + + Fix some of the brokeness in the last change. + + * swank-allegro.lisp (frame-source-location): Deal with frames for + undefined functions better. + (ldb-code-to-src-loc): Handle temp-files properly. + +2010-03-08 Helmut Eller <heller(a)common-lisp.net> + + Try to use source-level debugging features in Allegro 8.2 + + * swank-allegro.lisp (disassemble-frame): Use undocumented + debugger::dyn-fd-analyze to figure out the PC and display it. + (pc-source-location, ldb-code-to-src-loc, longest-common-prefix) + (source-paths-of): New functions. + (frame-source-location): Use pc-source-location. Still far from + optimal since Allegro rarely records source regions and anonymous + functions don't seem to carry source level debug-info at all. + (*temp-file-map*, buffer-or-file-location, find-fspec-location): + Use a table to map temp-file names back to Emacs buffers instead + of putting an eval-when-compile form in the source. The + eval-when-compile form messed up source positions. + (*temp-file-header-end-position*, find-definition-in-buffer): + Deleted. + (compile-from-temp-file): Bind excl:*load-source-debug-info* and + compiler:save-source-level-debug-info-switch so that Allegro + doesn't try to load debug-info from deleted files. Also put + the filename in *temp-file-map*. + +2010-03-08 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (dispatch-interrupt-event): Take a connection because + it boils down to SIGNAL-INTERRUPT which uses USE-THREADS-P which + needs a connection. + (install-fd-handler): Adapted accordingly. + (simple-serve-event): Adapted accordingly. Additionally, remove + superfluous WITH-SWANK-PROTOCOL-HANDLER as that's established by + WITH-CONNECTION already. + (simple-repl): Show "abort inferior lisp" restart only if not a + more appropriate "abort some REX" restart is available. Also make + sure to return in case of END-OF-FILE, otherwise there's an + infinite loop where we end up in the debugger again and again + until the user eventually selects close-connection restart himself. + (make-repl-input-stream): Use WITH-TOP-LEVEL-RESTART so + `sldb-quit' can be used in SLDB. + +2010-03-08 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (close-connection): Include initially passed + coding-system in debugging output. + +2010-03-08 Tobias C. Rittweiler <tcr(a)freebits.de> + + Make swank:connection-info include information about initially + passed coding-system, and the resulting external-format of the + socket. Debugging aid. + + * swank.lisp (connection.external-format): New function. + (start-server, create-server): Pass down coding-system, not + external-format. + (setup-server): Pass down both, coding-system and external-format. + (serve-connection): Ditto. + (create-connection): Set coding-system slot of CONNECTION. + (connection-info): Include coding-system and external-format. + 2010-03-07 Helmut Eller <heller(a)common-lisp.net> * swank-allegro.lisp (count-cr): Deleted. No longer used. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.351 slime/contrib/ChangeLog:1.353 --- slime/contrib/ChangeLog:1.351 Sun Mar 7 12:04:00 2010 +++ slime/contrib/ChangeLog Mon Mar 8 11:35:06 2010 @@ -1,7 +1,16 @@ +2010-03-08 Stas Boukarev <stassats(a)gmail.com> + + * swank-fancy-inspector.lisp (all-slots-for-inspector): Sort class + names when grouping by inheritance the same way as slots are sorted. + +2010-03-08 Stas Boukarev <stassats(a)gmail.com> + + * slime-repl.el (slime-call-defun): Work also on defvar/defparameter. + 2010-03-07 Stas Boukarev <stassats(a)gmail.com> - * swank-fancy-inspector.lisp (stable-sort-by-inheritance): Remove copy-seq, - unnecessarily put in the previous commit. + * swank-fancy-inspector.lisp (stable-sort-by-inheritance): Remove + copy-seq, unnecessarily put in the previous commit. 2010-03-07 Stas Boukarev <stassats(a)gmail.com>

2 1

[slime-devel] communication-style nil and interrupts
by Tobias C. Rittweiler 08 Mar '10

08 Mar '10

Currently, we use NIL on ECL because on one hand while it has threads, some parts are not yet thread-safe, and on the other hand, I do not fully understand Slime's :fd-handler architecture. :FD-HANDLER seems to be tied to an I/O-implies-serve-event architecture like SBCL (and CMUCL?) have. OTOH, I just took a courtesy look at WAIT-FOR-INPUT in swank-cmucl and see *INTERRUPT-QUEUED-HANDLER*. What's the purpose of that? Why does WAIT-FOR-INPUT write to that pipe? :-) ECL does have a select()-based serve-event, but it's not implicitly called. Where in the backend should it call serve-event? The problem with NIL is that if ECL was built with threads enabled, it executes the SIGINT handler function inside a new thread. But we actually want to interrupt the REPL thread -- however SIGNAL-INTERRUPT just ignores its thread argument in case :SPAWN is not used. So I do not see an obvious way out. That said, the problem won't be solved by implementing :FD-HANDLER for ECL either. Thoughts? -T.

2 1

Re: [slime-devel] slime setup in my .emacs
by Duke Normandin 08 Mar '10

08 Mar '10

On Mon, 8 Mar 2010, william douglas wrote: > > On Mon, Mar 8, 2010 at 9:38 AM, Duke Normandin <dukeofperl(a)ml1.net> wrote: > > Hey.... > > > > I've made such a freaking mess of the portion of my .emacs file > > pertaining to slime, hyperspec, etc, that I don't know which way is > > up. I've been pasting bits-n-pieces from web searches, etc and I've > > consequently screwed up the whole issue. > > > > Would it be appropriate to post the slime portion of my .emacs so you > > fellows can set me (and it) straight. > > > > The lesson I've learned is: > > > > Next time I experiment with my .emacs, I'll comment out the working > > portion, leaving a note that the portion is good; and delimit and > > comment the "test" setup. I'm old enough - I should have known better! > > Maybe creeping dotage is setting in - maybe Lisp will make me feel > > better. ;) Tia... > > -- > > duke > > Version control for config files has always been a good choice for me > that you might want to consider. Good suggestion! I'll give it a shot as soon as I get my .emacs file squared away. -- duke

1 0

[slime-devel] slime setup in my .emacs
by Duke Normandin 08 Mar '10

08 Mar '10

Hey.... I've made such a freaking mess of the portion of my .emacs file pertaining to slime, hyperspec, etc, that I don't know which way is up. I've been pasting bits-n-pieces from web searches, etc and I've consequently screwed up the whole issue. Would it be appropriate to post the slime portion of my .emacs so you fellows can set me (and it) straight. The lesson I've learned is: Next time I experiment with my .emacs, I'll comment out the working portion, leaving a note that the portion is good; and delimit and comment the "test" setup. I'm old enough - I should have known better! Maybe creeping dotage is setting in - maybe Lisp will make me feel better. ;) Tia... -- duke

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 08 Mar '10

08 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2019 slime/ChangeLog:1.2021 --- slime/ChangeLog:1.2019 Sun Mar 7 02:40:46 2010 +++ slime/ChangeLog Sun Mar 7 11:22:17 2010 @@ -1,3 +1,16 @@ +2010-03-07 Helmut Eller <heller(a)common-lisp.net> + + * swank-allegro.lisp (count-cr): Deleted. No longer used. + +2010-03-07 Helmut Eller <heller(a)common-lisp.net> + + Handle src-locs of compiler warnings in Allegro 8.2. + Didn't somebody already fix that? + + * swank-allegro.lisp (location-for-warning) + (handle-undefined-functions-warning): In 8.2 src-locs include not + only start but also and end positions. + 2010-03-07 Stas Boukarev <stassats(a)gmail.com> * swank-ecl.lisp (source-location): Don't do Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.348 slime/contrib/ChangeLog:1.351 --- slime/contrib/ChangeLog:1.348 Sat Feb 20 13:46:24 2010 +++ slime/contrib/ChangeLog Sun Mar 7 12:04:00 2010 @@ -1,12 +1,41 @@ +2010-03-07 Stas Boukarev <stassats(a)gmail.com> + + * swank-fancy-inspector.lisp (stable-sort-by-inheritance): Remove copy-seq, + unnecessarily put in the previous commit. + +2010-03-07 Stas Boukarev <stassats(a)gmail.com> + + * swank-fancy-inspector.lisp: Add buttons + for selecting default sorting order and default grouping method of + slots of a class. + + (all-slots-for-inspector): Implement the above feature. + Move the default method from :method option of the GF to a separate + defmethod, this method is quite large and :method eats space for + indentation. + (*inspector-slots-default-order*): + New variable, accepts :unsorted and :alphabetically + (*inspector-slots-default-grouping*): New variable, + accepts :all and :inheritance. + +2010-03-07 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-arglists.lisp (extract-local-op-arglists): Fix for + `(labels ((foo (x) ...)|'. + + * slime-autodoc.el (autodoc.1): Add test cases. + 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 - `swank:fuzzy-completion-selected' will be sent to the default connection. + * 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>

1 0

Re: [slime-devel] [Ecls-list] 10.3.1 slime issues
by Matthew Mondor 07 Mar '10

07 Mar '10

On Sun, 7 Mar 2010 12:02:13 -0500 Matthew Mondor <mm_lists(a)pulsar-zone.net> wrote: > On Sun, 07 Mar 2010 11:30:59 -0500 > Raffael Cavallaro <raffaelcavallaro(a)mac.com> wrote: > > > ;; Error while loading /Users/raffaelc/.slime/fasl/2010-03-05/ecl-10.3.1-darwin-i686/swank-ecl.fas: > > ;; The slot ASDF:SOURCE-FILE in the object #<ASDF:SYSTEM :PROFILE 4359849904> is unbound. > > ;; Aborting. > > ;; > > I also tried the latest slime today and the same error occurs. More details included (the log of the inferior-lisp buffer). Also cross-posting to slime-devel. This might not be related, but interestingly, it also warns about not detecting the OS in a list of OSs (:NETBSD is in *FEATURES* in my case), but it also warns about not detecting the architecture, but :I386 is in *FEATURES*). -- Matt

1 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 07 Mar '10

07 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2018 slime/ChangeLog:1.2019 --- slime/ChangeLog:1.2018 Fri Mar 5 12:45:34 2010 +++ slime/ChangeLog Sun Mar 7 02:40:46 2010 @@ -1,3 +1,10 @@ +2010-03-07 Stas Boukarev <stassats(a)gmail.com> + + * swank-ecl.lisp (source-location): Don't do + (setq file (tmpfile-to-buffer file)) in a COND condition, + otherwise next cond clause will get null file. + Apply translate-logical-pathname to pathnames. + 2010-03-05 Helmut Eller <heller(a)common-lisp.net> Remove some unused stuff.

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 06 Mar '10

06 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2014 slime/ChangeLog:1.2018 --- slime/ChangeLog:1.2014 Thu Mar 4 08:22:29 2010 +++ slime/ChangeLog Fri Mar 5 12:45:34 2010 @@ -1,7 +1,40 @@ +2010-03-05 Helmut Eller <heller(a)common-lisp.net> + + Remove some unused stuff. + + * swank-ccl.lisp (openmcl-set-debug-switches) + (*interesting-internal-frames*) + (interesting-frame-p): Unused. Deleted. + +2010-03-05 Helmut Eller <heller(a)common-lisp.net> + + * swank-ccl.lisp: Indentation fixes. + +2010-03-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-ecl.lisp: Make backend depend on ECL version 10.3.1 which + just got released. We do not support older versions. Previous + version pretty much didn't work in combination with Slime anyway. + +2010-03-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + Ecl: Make M-. work on function interactively compiled via C-c C-c. + + * swank-ecl.lisp (*tmpfile-map*, note-buffer-tmpfile) + (tmpfile-to-buffer): New helpers. + (swank-compile-string): Use them. Also use new COMPILE-FILE + keywords :SOURCE-TRUENAME and :SOURCE-OFFSET available in ECL + HEAD. + (find-definitions): Slurp in definition of + FIND-DEFINITIONS-BY-NAME. + (find-definitions-by-name): Hence not needed anymore. + (source-location): Use TMPFILE-TO-BUFFER to get buffer source + location of interactively compiled functions. + 2010-03-04 Mark Evenson <evenson(a)panix.com> * swank-abcl.lisp (emacs-inspect): Define default method to use - the result of SYS:INSPECTED-PARTS if non-nil. + the result of SYS:INSPECTED-PARTS if non-nil. 2010-03-03 Stas Boukarev <stassats(a)gmail.com>

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 05 Mar '10

05 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2013 slime/ChangeLog:1.2014 --- slime/ChangeLog:1.2013 Wed Mar 3 14:51:33 2010 +++ slime/ChangeLog Thu Mar 4 08:22:29 2010 @@ -1,3 +1,8 @@ +2010-03-04 Mark Evenson <evenson(a)panix.com> + + * swank-abcl.lisp (emacs-inspect): Define default method to use + the result of SYS:INSPECTED-PARTS if non-nil. + 2010-03-03 Stas Boukarev <stassats(a)gmail.com> * doc/slime.texi (Inspector): document slime-inspector-eval.

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 04 Mar '10

04 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2006 slime/ChangeLog:1.2013 --- slime/ChangeLog:1.2006 Tue Mar 2 09:36:47 2010 +++ slime/ChangeLog Wed Mar 3 14:51:33 2010 @@ -1,15 +1,54 @@ +2010-03-03 Stas Boukarev <stassats(a)gmail.com> + + * doc/slime.texi (Inspector): document slime-inspector-eval. + Patch by Mark Harig. + +2010-03-03 Helmut Eller <heller(a)common-lisp.net> + + Don't special case special operators with M-. for CMUCL. + + * swank-cmucl.lisp (function-definitions): Special operators will + have IR1-translators anyway, so no need to mark them "special". + (gf-definitions): Renamed from generic-function-definitions. + +2010-03-03 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-inspector-history): New command. + +2010-03-03 Helmut Eller <heller(a)common-lisp.net> + + Don't try to bind keywords in inspector-eval. + + * swank.lisp (inspector-eval): For now, don't bind symbols which + are constantp. Maybe something better can be found later. + +2010-03-03 Helmut Eller <heller(a)common-lisp.net> + + Use @ instead of # to mark object in the inspect history. + + * swank.lisp (print-part-to-string): Use @. + +2010-03-03 Helmut Eller <heller(a)common-lisp.net> + + Make eval-in-frame display multiple values; not only the first. + + * swank.lisp (values-to-string): New macro. + (eval-string-in-frame): Use it. + 2010-03-02 Stas Boukarev <stassats(a)gmail.com> - * swank-sbcl.lisp (definition-source-for-emacs): Don't error when source - location doesn't contain form position, point emacs to the beginning of a source file - and if it's a function provide a snippet "(defun <name> ". + * swank-sbcl.lisp (definition-source-for-emacs): Don't error when + source location doesn't contain form position, point emacs to the + beginning of a source file and if it's a function provide a + snippet "(defun <name> ". 2010-03-02 Stas Boukarev <stassats(a)gmail.com> * slime.el (slime-compile-and-load-file): Accept C-u arguments for compilation policy the same way as slime-compile-defun. - * swank.lisp (compile-file-for-emacs): Take an additional policy argument. + * swank.lisp (compile-file-for-emacs): Take an additional policy + argument. * swank-backend.lisp (swank-compile-file): Ditto. * swank-sbcl.lisp (compiler-policy, (setf compiler-policy)):

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 03 Mar '10

03 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2003 slime/ChangeLog:1.2006 --- slime/ChangeLog:1.2003 Mon Mar 1 10:42:07 2010 +++ slime/ChangeLog Tue Mar 2 09:36:47 2010 @@ -1,3 +1,25 @@ +2010-03-02 Stas Boukarev <stassats(a)gmail.com> + + * swank-sbcl.lisp (definition-source-for-emacs): Don't error when source + location doesn't contain form position, point emacs to the beginning of a source file + and if it's a function provide a snippet "(defun <name> ". + +2010-03-02 Stas Boukarev <stassats(a)gmail.com> + + * slime.el (slime-compile-and-load-file): Accept C-u arguments for + compilation policy the same way as slime-compile-defun. + + * swank.lisp (compile-file-for-emacs): Take an additional policy argument. + * swank-backend.lisp (swank-compile-file): Ditto. + + * swank-sbcl.lisp (compiler-policy, (setf compiler-policy)): + rename from get/set-compiler-policy. + (with-compiler-policy): New macro. + (swank-compile-file): Use with-compiler-policy. + (swank-compile-string): Ditto. + + * doc/slime.texi (Compilation): Update. + 2010-03-01 Stas Boukarev <stassats(a)gmail.com> * swank.lisp (documentation-symbol): Show arglists for functions too.

3 2

[slime-devel] (swank:set-default-directory) not working with CLISP
by Teemu Likonen 02 Mar '10

02 Mar '10

It seems that ,cd command from the REPL, which I think runs (swank:set-default-directory), does not work correctly with CLISP. It works _once_ but after that it doesn't change the directory anymore. CLISP's (ext:cd) works as longs as I don't touch REPL's ,cd command After that (ext:cd) stops working. My Slime is up-to-date version from Git mirror and I can see the bug with CLISP version 2.44.1 and 2.48.

2 2

[slime-devel] fuzzy completion
by Ramakrishnan Muthukrishnan 02 Mar '10

02 Mar '10

Hi, I am using the latest slime sources with the swank-clojure. It had been working rather good until I updated the slime yesterday to the latest and greatest from the git repository (from git://git.boinkor.net/slime.git) After a bit of testing with git bisect, I understood that the commit which caused the fuzzy-completions was this one: commit a82bb3be2e5b3dc22909303784cb72b65486ccf2 I do not really understand slime/swank well. Can someone help me on where exactly I need to look in the swank-clojure to fix this problem? The slime-events buffer shows this as I type (filt (:emacs-rex (swank:fuzzy-completions "filt" "user" :limit 300 :time-limit-in-msec 1500) "user" :repl-thread 3) (:return (:abort) 3) In a working version (commit ccab8f8dea8d23ecfd5699cc37a4b98617437ac7), this is what I get: (:emacs-rex (swank:fuzzy-completions "filt" "user" :limit 300 :time-limit-in-msec 1500) "user" :repl-thread 5) (:return (:ok ((("filter" "42.68" (...) "-f------") ("NoSuchFieldException/" "5.64" (... ... ...) "b-------")) nil)) 5) Any help to debug this will be appreciated. Thanks. -- Ramakrishnan

2 2

[slime-devel] Daily ChangeLog diff
by Helmut Eller 02 Mar '10

02 Mar '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2001 slime/ChangeLog:1.2003 --- slime/ChangeLog:1.2001 Fri Feb 26 16:02:59 2010 +++ slime/ChangeLog Mon Mar 1 10:42:07 2010 @@ -1,3 +1,18 @@ +2010-03-01 Stas Boukarev <stassats(a)gmail.com> + + * swank.lisp (documentation-symbol): Show arglists for functions too. + +2010-03-01 Tobias C. Rittweiler <tcr(a)freebits.de> + + Fix that annoying bug that caused being constantly asked to switch + connection in case one had multiple connections running and quit + from one. + + * slime.el (slime-modeline-string): Bail out early because + `slime-current-package' involves `slime-connection' in case the + slime-repl contrib is used, and querying the user in a + periodically called function is rather annoying.. + 2010-02-26 Stas Boukarev <stassats(a)gmail.com> * swank-loader.lisp (ecl-version-string): Check length of

1 0

Re: [slime-devel] CVS slime
by Helmut Eller 01 Mar '10

01 Mar '10

* CVS User trittweiler [2010-03-01 13:26+0100] writes: > + (let ((conn (slime-current-connection))) > + ;; Bail out early in case there's no connection, so we won't > + ;; implicitly invoke `slime-connection' which may query the user. > + (if (not conn) > + (and slime-mode " Slime") > + (let ((local (eq conn slime-buffer-connection)) > + (pkg (slime-current-package))) There was no call to slime-connection so this is no different than before with a misleading comment added. Helmut

2 1