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

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 05 Oct '04

05 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.544 slime/ChangeLog:1.545 --- slime/ChangeLog:1.544 Sun Oct 3 14:33:33 2004 +++ slime/ChangeLog Tue Oct 5 23:39:58 2004 @@ -1,3 +1,10 @@ +2004-10-05 Luke Gorrie <luke(a)synap.se> + + * swank.lisp (arglist-for-echo-area): Handle errors and return a + message. + (parse-symbol): Recognise an empty package name as the KEYWORD + package. + 2004-10-03 Reini Urban <rurban(a)x-ray.at> * swank-clisp.lisp (getpid)[win32]: Use

1 0

[slime-devel] SWANK:ARGLIST-FOR-ECHO-AREA drops into the debugger!
by Pascal J.Bourguignon 05 Oct '04

05 Oct '04

Whatever the configuration, slime should not drop in the common-lisp debugger when I'm editing a file! FIND-SYMBOL: There is no package with name "ASDF" [Condition of type SYSTEM::SIMPLE-PACKAGE-ERROR] Restarts: 0: [USE-VALUE] You may input a value to be used instead. 1: [ABORT] Abort handling SLIME request. Backtrace: 0: frame binding variables (~ = dynamically): | ~ SYSTEM::*FASOUTPUT-STREAM* <--> NIL 1: EVAL frame for form (SWANK:ARGLIST-FOR-ECHO-AREA '("asdf:oo")) 2: EVAL frame for form (SWANK:START-SERVER "/tmp/slime.28369") 3: EVAL frame for form (SWANK:START-SERVER "/tmp/slime.28369") -- __Pascal Bourguignon__ http://www.informatimago.com/ To vote Democrat or Republican, it's like changing of cabin in the Titanic.

2 1

[slime-devel] implementation type / name of slime buffers
by Pascal J.Bourguignon 04 Oct '04

04 Oct '04

With ilisp we could launch various implementation invoking by a command of same name (M-x sbcl, M-x clisp-hs, etc). Then, it would create a buffer named for the implementation instead of *ilisp[1]*. It was then easy to know in which implementation we were working and to switch between implementations, because they were distinguished by the buffer name, clearly indicated in the mode line and in the buffer list. Now, slime confuses all this, naming all its buffer *slime-repl[N]*. One thing that could be done is to add the implementation type in the prompt, perhaps with something like: (defun package-string-for-prompt (package) "Return the shortest nickname (or canonical name) of PACKAGE." (concatenate 'string (lisp-implementation-type) " " (or (canonical-package-nickname package) (auto-abbreviated-package-name package) (shortest-package-nickname package)))) but I think it would be better if slime named its buffer for the implementation type. *clisp/slime[1]*, *sbcl/slime[2]*, for example. (It may be useful that the inferior lisp buffer includes the implementation type too, or includes always the same "reference number"). -- __Pascal Bourguignon__ http://www.informatimago.com/ To vote Democrat or Republican, it's like changing of cabin in the Titanic.

3 2

[slime-devel] multiple connections
by Andras Simon 04 Oct '04

04 Oct '04

After starting two inferior lisps (the first with M-x slime, the second with C-u M-x slime), slime-list-connections correctly gives Nr Name Port Pid Type -- ---- ---- --- ---- 1 allegro (127.0.0.1 33032) 10312 International Allegro CL Enterprise Edition * 2 cmucl (127.0.0.1 33036) 10314 CMU Common Lisp But hitting RET on the first line doesn't seem to do anything. Hitting 'd' and then RET does the right thing. Also, I get various errors if I do ,restart-inferior-lisp, or even ,sayoonara. Digging a bit deeper: (slime-process) returns nil when in cmucl's repl, and the reason is probably that the inferior-lisp processes (as given by process-list) are process: #<process inferior-lisp<1>> pid: 10313 <-- shouldn't this be 10314? process: #<process inferior-lisp> pid: 10312 I use the latest cvs slime with GNU Emacs 21.2.1. Helmut says multiple connections work for him, so maybe I'm missing something, or it's an emacs version thing. Andras

2 3

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

03 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.543 slime/ChangeLog:1.544 --- slime/ChangeLog:1.543 Fri Oct 1 15:39:10 2004 +++ slime/ChangeLog Sun Oct 3 14:33:33 2004 @@ -1,3 +1,35 @@ +2004-10-03 Reini Urban <rurban(a)x-ray.at> + + * swank-clisp.lisp (getpid)[win32]: Use + win32:|GetCurrentProcessId|. + +2004-10-03 Helmut Eller <heller(a)common-lisp.net> + + * slime.el: Reduce dependency on inf-lisp internals. Make it + possible to start the inferior lisp in a buffer different from + "*inferior-lisp*". + (slime): Parse the command argument explicitly and don't rely on + `inferior-lisp'. Don't close all connections, but only the one + for the inferior lisp buffer we are using. + (slime-maybe-start-lisp): Take the command and buffer as argument. + Decide here whether we should start start a new processwe or just + disconnect and reconnect . + (slime-start-lisp): Load verbosely. + (slime-inferior-lisp): New function. Replaces call to + `inferior-lisp'. + (slime-inferior-connect, slime-start-swank-server): Take the + inferior process as argument + (slime-read-port-and-connect): Set the slime-inferior-process + variable in the new connection. + (slime-inferior-process): New connection local variable. + (slime-process): Use it. + (slime-restart-inferior-lisp): Don't use inferior lisp stuff. + (slime-switch-to-output-buffer): Process interactive arguments + properly. + + * swank-loader.lisp (compile-files-if-needed-serially): Load + verbosely. + 2004-10-01 Helmut Eller <heller(a)common-lisp.net> * swank-allegro.lisp (find-fspec-location): excl:source-file can @@ -5,7 +37,7 @@ * swank-cmucl.lisp (code-component-entry-points): Only include entry points with "valid" functions names. This excludes internal - lambdas, which have usually a string as name, like "defun foo". + lambdas which have usually a string as name, like "defun foo". * swank.lisp (parse-symbol): Don't use the reader to avoid interning unknown symbols. The downside is that we no longer @@ -14,7 +46,6 @@ * slime.el (slime-set-connection-info): Hide the *inferior-lisp* buffer after we know Lisp's pid. Print the words of encouragement here, when all the other asynchronous initialization is completed. - (slime-find-buffer-package): We need to preserve the case for things like (:in-package "foo"), so return "\"foo\"".

1 0

[slime-devel] SLIME starts with two windows
by Edi Weitz 03 Oct '04

03 Oct '04

Since some time last week SLIME from CVS starts up with two windows, one for the inferior Lisp buffer and for the SLIME REPL. It used to start up with only the SLIME REPL, right? Is this new window split on purpose? Sorry, I haven't been following the mailing list very closely in the last weeks. Thanks, Edi.

5 9

[slime-devel] pid
by Andras Simon 03 Oct '04

03 Oct '04

Is there a reason why the pid of the inferior lisp should be asked from the lisp itself? Isn't (process-id (inferior-lisp-proc)) good enough? (Are multiple lisps in the same slime instance or a remotely running lisp the problem?) I'm asking this both out of curiosity and because Java (and hence abcl) doesn't have a clue about its pid(s), at least AFAIK. Andras

3 7

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 01 Oct '04

01 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.541 slime/ChangeLog:1.543 --- slime/ChangeLog:1.541 Tue Sep 28 00:39:40 2004 +++ slime/ChangeLog Fri Oct 1 15:39:10 2004 @@ -1,3 +1,23 @@ +2004-10-01 Helmut Eller <heller(a)common-lisp.net> + + * swank-allegro.lisp (find-fspec-location): excl:source-file can + return stuff like (:operator ...); try to handle it. + + * swank-cmucl.lisp (code-component-entry-points): Only include + entry points with "valid" functions names. This excludes internal + lambdas, which have usually a string as name, like "defun foo". + + * swank.lisp (parse-symbol): Don't use the reader to avoid + interning unknown symbols. The downside is that we no longer + handle escaped |symbols| correctly. + + * slime.el (slime-set-connection-info): Hide the *inferior-lisp* + buffer after we know Lisp's pid. Print the words of encouragement + here, when all the other asynchronous initialization is completed. + + (slime-find-buffer-package): We need to preserve the case for + things like (:in-package "foo"), so return "\"foo\"". + 2004-09-27 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-process): New function intended to replace all

1 0

[slime-devel] reasoning behind slime-maybe-list-compiler-notes
by Alan Ruttenberg 01 Oct '04

01 Oct '04

I read this as Don't show the notes a) if there aren't any b) There are but they all have the location noted a) makes sense to me b) doesn't. I keep getting a note that there is a warning but then have to switch to the repl to see what it is. (defun slime-maybe-list-compiler-notes (notes) "Show the compiler notes if appropriate." (unless (or (null notes) (and (eq last-command 'slime-compile-defun) (every #'slime-note-has-location-p notes))) (slime-list-compiler-notes notes))) -Alan

2 1

[slime-devel] break
by Andras Simon 01 Oct '04

01 Oct '04

The HS says: break binds *debugger-hook* to nil before attempting to enter the debugger. Still, with all slime backends (but abcl) I tried, (break) enters the slime debugger. How is this arranged? Andras

2 2