September 2005 - slime-devel - mailman3.common-lisp.net

[slime-devel] openmcl: apropos bug fix
by Dave Baum 22 Sep '05

22 Sep '05

The current implementation of describe-symbol-for-emacs assumes that ccl::setf-function-spec-name will always return a symbol. Although this is true for short-form defsetf forms, when the long form has been used, an anonymous function is returned instead of a symbol, which then causes describe-symbol-for-emacs to fail. For example, doing an apropos on "sub" will cause an error because "subseq" has a long form defsetf. Here is a fix: Index: swank-openmcl.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank-openmcl.lisp,v retrieving revision 1.101 diff -u -r1.101 swank-openmcl.lisp --- swank-openmcl.lisp 21 Sep 2005 11:43:47 -0000 1.101 +++ swank-openmcl.lisp 22 Sep 2005 02:18:52 -0000 @@ -634,7 +634,8 @@ (maybe-push :setf (let ((setf-function-name (ccl::setf-function-spec-name `(setf ,symbol)))) - (when (fboundp setf-function-name) + (when (or (functionp setf-function-name) ;; long form defsetf + (fboundp setf-function-name)) (doc 'function setf-function-name)))) result)))

1 0

[slime-devel] Problem with SBCL 0.9.4
by Pekka Niiranen 22 Sep '05

22 Sep '05

Hi there, slime (22.9.2005) starts with Clisp ok, but sbcl v0.9.4 gives error: This is SBCL 0.9.4, an implementation of ANSI Common Lisp. More information about SBCL is available at <http://www.sbcl.org/>. SBCL is free software, provided as is, with absolutely no warranty. It is mostly in the public domain; some portions are provided under BSD-style licenses. See the CREDITS and COPYING files in the distribution for more information. * ; loading #P"/home/pekka/slime/swank-loader.lisp" ; compiling file "/home/pekka/slime/swank-backend.lisp" (written 28 AUG 2005 05:47:11 PM): ; ; compilation unit aborted ; caught 2 fatal ERROR conditions * * The *Messages* windows shows: (emacs) Loading disp-table...done Loading tool-bar...done Loading image...done Loading tooltip...done Loading 00debian-vars...done Loading 20gnus-init (source)...done Loading 49url (source)...done Loading 50autoconf (source)...done Loading 50dictionaries-common (source)... Loading /var/cache/dictionaries-common/emacsen-ispell-dicts.el (source)...done Loading ispell...done Loading 50dictionaries-common (source)...done Loading 50emacs-goodies-el (source)...done Loading 50w3 (source)...done Loading cl-macs...done Loading cl-seq...done Loading regexp-opt...done Loading easy-mmode...done Loading edmacro...done Loading derived...done Loading byte-opt...done Loading cl-extra...done For information about the GNU Project and its goals, type C-h C-p. Polling "/tmp/slime.5314".. (Abort with `M-x slime-abort-connection'.) [36 times] the number in last line ([36 times]) keeps increasing. I am using Ubuntu 5.04 in AMD64 -pekka-

2 1

[slime-devel] Error in process filter: Wrong type argument: char-or-string-p, :present
by Paolo Amoroso 22 Sep '05

22 Sep '05

I have just updated my CVS copy of SLIME. It works fine with CMUCL Snapshot 2005-09 (19B) and GNU Emacs 21.3.2 under Slackware Linux 10.0, including loading the present.lisp file. But when I try to compile present.lisp from within a SLIME session, I get this error in the Emacs echo area: error in process filter: Wrong type argument: char-or-string-p, :present Compilation of that fine worked fine in earlier versions of SLIME. Paolo -- Lisp Propulsion Laboratory log - http://www.paoloamoroso.it/log

1 1

[slime-devel] slime-after-change-function performance issue
by Russell McManus 22 Sep '05

22 Sep '05

On XEmacs, this function appears to cause serious thrashing for me. When I take it out of after-change-functions, the problem goes away. -russ

3 2

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 21 Sep '05

21 Sep '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.785 slime/ChangeLog:1.787 --- slime/ChangeLog:1.785 Mon Sep 19 10:20:48 2005 +++ slime/ChangeLog Wed Sep 21 22:55:50 2005 @@ -1,3 +1,49 @@ +2005-09-21 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-setup-command-hooks): Make + after-change-functions a buffer-local variable; it's by default + global in XEmacs. + + * swank.lisp (throw-to-toplevel): Invoke the `abort-restart' + request instead of throwing to the `slime-toplevel' catch tag. + (handle-request): Rename the restart from abort to abort-request. + (call-with-connection): Remove the slime-toplevel catch tag + because with-connection is used in far to many places which aren't + at "toplevel". + + * present.lisp (presentation-start, presentation-end): Use + finish-output instead of force-output. + + * swank-gray.lisp, swank-cmucl.lisp: Improve stream efficiency by + buffering more output. stream-force-output simply does nothing, if + the output buffer was flushed less than 200 millisecons before. + stream-finish-output can still be used to really flush the buffer. + (slime-output-stream): New slot last-flush-time. + (stream-finish-output): New function. Do what stream-force-output + did previously. + (stream-force-output): Buffer more output. + + * slime.el (slime-process-available-input): Oops, don't start a + timer for every event. + (slime-write-string): Renamed from slime-output-string. + (slime-dispatch-event): Rename :read-output to :write-string. + (slime-io-speed-test): New command. + (slime-open-stream-to-lisp): Fix parens. The coding system should + also be set if presentations are disabled. + + * swank.lisp (make-output-function): Rename :read-output to + :write-string. + (eval-for-emacs, interactive-eval, eval-region): Use finish-output + not force-output. + + * swank-sbcl.lisp, swank-openmcl.lisp, swank-allegro.lisp, + swank-lispworks: Import `stream-finish-output'. + + * swank-scheme48/io.scm (empty-swank-output-buffer): Rename + :read-output to :write-string. + + * swank-scheme48/load.scm (slime48-start): Fix '() vs. #f bug. + 2005-09-19 Luke Gorrie <luke(a)synap.se> * nregex.lisp: Released into the public domain by Lawrence E. Freil.

1 0

[slime-devel] Patch for stepper support in SBCL and (sort of) CLISP
by Svein Ove Aas 21 Sep '05

21 Sep '05

One patch, one new file. The stepper works in SBCL, and should be fairly self-explanatory. C-s to toggle stepping, c to step inwards and x to quit stepping but keep running. You need to compile with high debug quality. One gotcha: Slime will only display the source correctly if you compile the entire file with C-c C-k. The same works in CLISP, but only for uncompiled code, and you won't see your location. Fix the sldb view-source command, and it'll work. OpenMCL and CMUCL appear to have broken step implementations; if these work for anyone, please contact me so I can write backends. Finally, while working on this I had some ideas about improving other parts of Slime, and I've left commented-out code for this in at least one place. When the debugger pops up, we might want to (a) skip "internal" frames such as break when positioning the cursor, and (b) attempt to view the source for the resulting frame automatically. There is very little work left to do this, but is it a good idea? Also, the way the stepper currently works there are (at least in clisp) a couple of stepper-internal frames *between* those belonging to the user. Would filtering out those be okay?

1 1

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 19 Sep '05

19 Sep '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.784 slime/ChangeLog:1.785 --- slime/ChangeLog:1.784 Sun Sep 18 23:16:51 2005 +++ slime/ChangeLog Mon Sep 19 10:20:48 2005 @@ -1,3 +1,7 @@ +2005-09-19 Luke Gorrie <luke(a)synap.se> + + * nregex.lisp: Released into the public domain by Lawrence E. Freil. + 2005-09-19 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime48): New command.

1 0

[slime-devel] lag in posting to slime-devel
by Matthew D Swank 19 Sep '05

19 Sep '05

I often see a delay of up to several hours when I post to this list. I was wondering if my experience was atypical. Matt

2 1

[slime-devel] patch to load.scm
by Matthew D Swank 19 Sep '05

19 Sep '05

CAUTION: scheme ahead! From (define (slime48-start . port-file) ... Scheme list are never #f, so the test is always true: (eval (if port-file `(start-swank ,@port-file) '(slime48)) diff -ur swank-scheme48.orig/load.scm swank-scheme48/load.scm --- swank-scheme48.orig/load.scm 2005-09-19 11:55:46.000000000 -0500 +++ swank-scheme48/load.scm 2005-09-19 11:50:13.000000000 -0500 @@ -73,9 +73,9 @@ (in 'slime48 (lambda () (call-with-values (lambda () - (eval (if port-file - `(start-swank ,@port-file) - '(slime48)) + (eval (if (null? port-file) + '(slime48) + `(start-swank ,@port-file)) (interaction-environment))) (lambda (world server) (user (lambda ()

2 1

[slime-devel] trouble updating slime from cvs
by Alan Ruttenberg 19 Sep '05

19 Sep '05

cvs server: Updating slime/swank-scheme48 cvs server: failed to create lock directory for `/project/slime/ cvsroot/slime/swank-scheme48' (/project/slime/cvsroot/slime/swank- scheme48/#cvs.lock): Permission denied cvs server: failed to obtain dir lock in repository `/project/slime/ cvsroot/slime/swank-scheme48' cvs [server aborted]: read lock failed - giving up -Alan

2 2