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

[slime-devel] How to stop a lengthy computation?
by Pascal J.Bourguignon 07 Jul '04

07 Jul '04

How do you stop a lengthy computation? Obviously, if I ask, it's because a mere C-g does not work! (I use clisp under slime). -- __Pascal Bourguignon__ http://www.informatimago.com/ There is no worse tyranny than to force a man to pay for what he does not want merely because you think it would be good for him. -- Robert Heinlein

4 4

[slime-devel] no applicable method for the generic function SWANK-BACKEND:SPAWN
by Liam Healy 07 Jul '04

07 Jul '04

I was using an older version in XEmacs, then did a cvs update today. Using sbcl-mt, slime now ends in an error: M-x slime ends with * ;; Swank started at port: 32846. debugger invoked on a SIMPLE-ERROR in thread 2825: There is no applicable method for the generic function #<STANDARD-GENERIC-FUNCTION SWANK-BACKEND:SPAWN (0)> when called with arguments (# :NAME "Swank"). You can type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [ABORT ] Reduce debugger level (leaving debugger, returning to toplevel). 1: [TOPLEVEL] Restart at toplevel READ/EVAL/PRINT loop. ((NO-APPLICABLE-METHOD (T)) #<unavailable argument> #<unavailable argument> #<STANDARD-GENERIC-FUNCTION SWANK-BACKEND:SPAWN (0)> (#<FUNCTION "CLOSURE" {965627D}> :NAME "Swank")) 0] :a Liam

1 0

[slime-devel] Bug report: finish-output does not finish with slime!
by Pascal J.Bourguignon 06 Jul '04

06 Jul '04

CL-USER> (dolist (max '(10 100 1000)) (format t "~2% max = ~D ~%" max) (finish-output) (time (make-grid-dfa 10))) max = 10 Real time: 0.030129 sec. Run time: 0.03 sec. Space: 614340 Bytes max = 100 Real time: 0.080329 sec. Run time: 0.07 sec. Space: 614340 Bytes GC: 1, GC time: 0.05 sec. max = 1000 Real time: 0.027293 sec. Run time: 0.03 sec. Space: 614340 Bytes NIL CL-USER> (dolist (max '(10 100 1000)) (format t "~2% max = ~D ~%" max) (finish-output) (time (make-grid-dfa max))) In the first case, the output shows, but actually it's at the end. In the second case, since we've got O(n^2) processing time, I'm still waiting the first line! -- __Pascal Bourguignon__ http://www.informatimago.com/ There is no worse tyranny than to force a man to pay for what he does not want merely because you think it would be good for him. -- Robert Heinlein

2 1

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

04 Jul '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.458 slime/ChangeLog:1.461 --- slime/ChangeLog:1.458 Fri Jul 2 17:08:16 2004 +++ slime/ChangeLog Sat Jul 3 22:17:26 2004 @@ -1,3 +1,38 @@ +2004-07-04 Luke Gorrie <luke(a)bluetail.com> + + * slime.el, swank-backend.lisp, swank.lisp: Added a new backend + function `buffer-first-change' which is called via Emacs's + `first-change-hook' in slime-mode buffers. This gives Lisp a + chance to do something with source files before you change them on + disk. + + * swank-cmucl.lisp (buffer-first-change): Suck the source file + into the cache unless already present. This is for M-. to increase + the chances of our having a copy of the sources corresponding with + the loaded code. Should help with the case where a user edits and + saves a file (without recompiling it) and then M-.'s for one of + its definitions. + + * swank-allegro.lisp (make-process/inherit): Changed reader + conditionals to use fwrappers for #+(version>= 6). + + * swank-backend.lisp (make-stream-interactive): This backend + function is called with each stream that will be used for + user-interaction, i.e. the redirected stdio streams. Can be used + to setup special output-flushing or similar. + + * swank.lisp (open-streams): Call make-stream-interactive on the + redirected io streams. + + * swank-allegro.lisp (make-stream-interactive): Set + interactive-stream-p slot on the stream to make it auto-flush. + (*swank-thread*, *inherited-bindings*): New variables. + (spawn): Bind *swank-thread* to T. + (make-process/inherit): Fwrapper (advice) for + mp:make-process. When *swank-thread* is T then make the new thread + inherit "sliminess": debugger hook, I/O streams, and also + *swank-thread* so that its children will inherit too. + 2004-07-03 Luke Gorrie <luke(a)bluetail.com> * hyperspec.el (common-lisp-hyperspec-section-4.0): Bugfix from

1 0

[slime-devel] Multi-threaded debugging in Allegro?
by Peter Seibel 04 Jul '04

04 Jul '04

So I was checking out whether multi-threaded debugging in Allegro worked and did this: CL-USER> (mp:process-run-function "foo" #'(lambda () (break))) One time it seemed to have opened up an SLDB buffer to debug the thread which I didn't notice at first and found later. But I just tried it again from a freshly started SLIME and it didn't go into SLDB. So two questions: 1) Am I doing something wrong now that's causing it not to use SLDB and 2) Assuming I am, is it supposed to silently create the SLDB buffer? Is there some way to know that a thread has ended up in the debugger? -Peter -- Peter Seibel peter(a)javamonkey.com Lisp is the red pill. -- John Fraser, comp.lang.lisp

2 7

[slime-devel] debug-in-emacs mod suggestion
by Alan Ruttenberg 03 Jul '04

03 Jul '04

Instead of calling compute-restarts directly call an implementation-define swank-compute-restarts. Default that function to call compute-restarts. In openmcl we would define (defun swank-compute-restarts (condition) (butlast (compute-restarts condition) 2)) or something more robust, if I can figure that out, to get rid of the restarts that blow away the slime/emacs connection. BTW, I would make the change myself but my password to the slime CVS seems to have gone bad and mail to the common-lisp.net maintainers has not yet yielded a response. Perhaps Helmut or Luke could help me out... -Alan

2 2

Re: [slime-devel] return value
by Nicolas Neuss 03 Jul '04

03 Jul '04

Rolf Wester <rolf.wester(a)ilt.fraunhofer.de> writes: > Thank you for the answer. Unfortunately C-x M-e and C-x C-e show the same > behaviour. When evaluating > (let ((v (make-array 10 :initial-element 1.0))) v) > with either C-x M-e or C-x C-e the 1.0s ore displayed in the mini-buffer > but not in the *slime-repl* nor in the *SLIME-Typeout* buffer. > > Regards > > Rolf > > P.S.: I'm using slime-1.0alpha and CMUCL18e Oops. You're right. I have returned to SLIME only recently (I have even made a pause with Lisp). Sorry for the misinformation, Nicolas.

1 0

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 03 Jul '04

03 Jul '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.452 slime/ChangeLog:1.458 --- slime/ChangeLog:1.452 Wed Jun 30 23:35:50 2004 +++ slime/ChangeLog Fri Jul 2 17:08:16 2004 @@ -1,3 +1,29 @@ +2004-07-03 Luke Gorrie <luke(a)bluetail.com> + + * hyperspec.el (common-lisp-hyperspec-section-4.0): Bugfix from + Lennart Staflin. + + * slime.el (slime-repl-clear-output): Avoid clearing the previous + REPL expression too. Patch from Andras Simon. + + * swank-backend.lisp (definterface): Don't use + NO-APPLICABLE-METHOD for default methods. Instead just define them + as regular methods with all argument types being + T. Defimplementation will then replace them by using the same + signature. N-A-M was a stupid idea! + +2004-07-02 Brian Downing <bdowning(a)lavos.net> + + * slime.el (slime-reindent-defun): Added a check for (boundp + 'slime-repl-input-start-mark) before checking the variable, as + XEmacs leaves variables unbound when `make-variable-buffer-local' + is run, while GNU Emacs binds them to NIL. + +2004-07-02 Martin Simmons <martin(a)xanalys.com> + + * swank-lispworks.lisp (dspec-stream-position, + make-dspec-location): Fix typo in features for LW 4.1 and 4.2. + 2004-07-01 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank-lispworks.lisp (frame-actual-args): Bind

1 0

[slime-devel] HyperSpec 4.0, section lookup
by Lennart Staflin 03 Jul '04

03 Jul '04

I would like to suggest the following change to hyperspec.el --- hyperspec.el 30 Jun 2004 19:48:40 +0200 1.5 +++ hyperspec.el 02 Jul 2004 11:51:16 +0200 @@ -1135,10 +1135,9 @@ "") ".htm"))) -(defun common-lisp-hyperpsec-section-4.0 (indices) - (let ((string (format "%sBody/sec%s_" - common-lisp-hyperspec-root - (pop indices)))) +(defun common-lisp-hyperspec-section-4.0 (indices) + (let ((string (format "%sBody/sec_" + common-lisp-hyperspec-root))) (concat string (mapconcat (lambda (n) (format "%d" n)) That section naming works with the hyperspec copy I got. I hope it's the same for all copies. That the function name is misspelled suggests that that function hasn't been tested much. //Lennart Staflin

4 3

[slime-devel] slime-repl-clear-output
by Andras Simon 03 Jul '04

03 Jul '04

clears the last input as well; is this intentional? There's probably a better way with properties or markers, but here's what I mean: --- slime.el.~1.352.~ Wed Jun 30 23:12:30 2004 +++ slime.el Fri Jul 2 12:22:26 2004 @@ -2428,6 +2428,8 @@ (interactive) (let ((start (save-excursion (slime-repl-previous-prompt) + (forward-sexp) + (forward-line) (point))) (end (1- (slime-repl-input-line-beginning-position)))) (when (< start end) Andras

2 1