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

[slime-devel] patchlets on buffering.
by Mario S. Mommer 11 Nov '05

11 Nov '05

Hello, By default, the connection between the lisp and the slime repl is buffered. I would like to propose to have it unbuffered. on the grounds that if the user has a (format t ...), he/she probably wants the message to appear immediately on the screen :-) I've attached two microscopic patches that make the connections unbuffered for sbcl and cmucl. Regards, Mario.

2 2

[slime-devel] SLIME and SBCL
by Nicolas Neuss 11 Nov '05

11 Nov '05

Hello, first let me say thank you for incorporating SLIME into Debian. That version of SLIME also works fine with CMUCL. However, I have problems with making it work with SBCL from Debian. I get the messages: Connecting to Swank on port 33826.. Process bridge is installed Lisp connection closed unexpectedly: exited abnormally with code 256 I also have tried it with the CVS version of SLIME, with the same result. Does it work for others? Do I have to compile SBCL from CVS? Thanks, Nicolas.

3 2

[slime-devel] second micropatch
by Mario S. Mommer 11 Nov '05

11 Nov '05

Oops. Stepped on the "send" button by accident. Here is the second patch, to make swank-sbcl.lisp unbuffered. Regards, Mario.

1 0

[slime-devel] SBCL 0.9.6 and SLIME
by Alan Shields 10 Nov '05

10 Nov '05

Dear list, While trying to solve issues that have cropped up with SBCL 0.9.6 and Araneida, I put some (warn "blah") lines in. With SBCL 0.9.5 this printed to the slime REPL - with 0.9.6 it does not. I'm using a threaded server, so the warnings are happening in another thread. Has this happened to anyone else? Is this intended behavior? Thanks, Alan

1 1

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 07 Nov '05

07 Nov '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.806 slime/ChangeLog:1.807 --- slime/ChangeLog:1.806 Sun Nov 6 10:09:47 2005 +++ slime/ChangeLog Mon Nov 7 09:30:00 2005 @@ -1,3 +1,12 @@ +2005-11-07 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-eval-macroexpand): Use lisp-mode (and + font-lock-mode) when dispaying the expansion. Suggested by Jan + Rychter. + + * swank-source-path-parser.lisp (make-source-recording-readtable): + Suppress the #. reader-macro. + 2005-11-06 Juho Snellman <jsnell(a)iki.fi> * swank-sbcl.lisp (find-definitions, make-source-location-specification

1 0

[slime-devel] source path finding breaks in the presence of #.
by Andreas Fuchs 07 Nov '05

07 Nov '05

Hi all, I encountered a construct like the following in cl-irc. It makes the source path finder in M-. break: ------------- C-c C-k, then M-. A RET ------------ (cl:defpackage :swank-bork (:use :cl)) (cl:in-package :swank-bork) (eval-when (:compile-toplevel :load-toplevel :execute) (defvar *foo* '(a b c))) (defmacro define-foos (args) `(progn ,@(mapcar (lambda (arg) `(defun ,arg (,arg) ,arg)) args))) (define-foos #.*foo*) ------------- snip ------------ trying to M-. A or B or C will fail with "Error: The variable *FOO* is unbound." With Juho Snellman's help, I tracked this down to swank::read-source-form. My current suspicion on why this is happening is that it READs the form and doesn't bind *package* to the package that the stream uses. This causes evaluation of the #. form to yield bad results. Unfortunately, I have no idea how to fix this, or if this is fix-worthy at all. I guess read-source-form would have to try and find a top-level in-package form somewhere, possibly expanding macros along the way and digging into top-levelness preserving forms like eval-when and progn. That would be a lot of work for this one error, in my eyes. Thanks for any advice (or fixes (-8), -- Andreas Fuchs, <asf(a)boinkor.net>, asf(a)jabber.at, antifuchs

3 3

[slime-devel] non-blocking read from REPL
by Daniel Patru 07 Nov '05

07 Nov '05

This is a clarification of my previous question.. My question is: How do I do a non-blocking read in the slime REPL? I tried (and (listen) (read-line)) and (read-char-no-hang) but keep getting the message: ; pipelined request... (swank:listener-eval "<input>") in the minibuffer. E.g. CL-USER> (progn (sleep 3) (and (listen) (read-line))) 2 2 CL-USER> NIL I typed the first 2, the minibuffer flashed the message ; pipelined request... (swank:listener-eval "2 ) then returned the 2, CL-USER>, and NIL. In constrast, typing (read-line) without the (listen) works: CL-USER> (read-line) 2 "2" NIL Similar things happen when I try (read-char-no-hang) in place of (and (listen) (read-line)). Both of these work (function as a non-blocking read) in a REPL run from the shell (no emacs, no slime.) It seems that any new input in the slime REPL during a running command is buffered as a new command instead of as input to the currently running command. I tried a blocking read-line after the non-blocking read-line and typed: 1<newline>2<newline>3<newline>4<newline> and got: CL-USER> (progn (sleep 10) (and (listen) (read-line)) (print (list (read-line)))) 1 2 3 4 3 CL-USER> 4 CL-USER> 2 CL-USER> 1 CL-USER> 5 ("5") ("5") CL-USER> I typed the last 5 after the sleep period. It seems that the 1, 2, 3, and 4 were processed at toplevel while the last 5 was used as input to unblock the previous read-line. Is it even possible to do a non-blocking read in the slime REPL, or is this one of the things I have to give up if I want to use slime? Thanks, Dan

2 1

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 06 Nov '05

06 Nov '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.805 slime/ChangeLog:1.806 --- slime/ChangeLog:1.805 Fri Nov 4 10:07:43 2005 +++ slime/ChangeLog Sun Nov 6 10:09:47 2005 @@ -1,3 +1,11 @@ +2005-11-06 Juho Snellman <jsnell(a)iki.fi> + + * swank-sbcl.lisp (find-definitions, make-source-location-specification + make-definition-source-location, source-hint-snippet): As of + SBCL 0.9.6.25 SB-INTROSPECT has better support for finding + source locations. Use as much of it in swank-sbcl as possible. + (Original version left reader-conditionalized for older SBCLs). + 2005-11-04 Helmut Eller <heller(a)common-lisp.net> * swank.lisp (connection-info): Docfix.

1 0

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 04 Nov '05

04 Nov '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.804 slime/ChangeLog:1.805 --- slime/ChangeLog:1.804 Mon Oct 31 09:26:54 2005 +++ slime/ChangeLog Fri Nov 4 10:07:43 2005 @@ -1,3 +1,11 @@ +2005-11-04 Helmut Eller <heller(a)common-lisp.net> + + * swank.lisp (connection-info): Docfix. + + * slime.el (slime-set-connection-info): Generate a new connection + name only if the implementation-name and the inferior-lisp-name + are different. + 2005-10-31 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-start, slime-lookup-lisp-implementation)

1 0

[slime-devel] CVS slime and Lispworks no longer working
by Luke J Crook 04 Nov '05

04 Nov '05

I receive an error in Lispworks after downloading the latest version of slime from CVS. After entering M-x slime-connect in Emacs, Lispworks returns the following: Undefined function SWANK::INSTALL-DEBUGGER-GLOBALLY called with arguments (#<function SWANK:SWANK-DEBUGGER-HOOK 21593A82>). -Luke

2 2

[slime-devel] SLIME history file location
by Peter BARABAS 04 Nov '05

04 Nov '05

Hello, What do you think about putting .slime-history.eld in ~/.slime? Since we already have that directory, we could just use it for storing the history file. -- '(Yours parenthetically), peter barabas.

1 0

[slime-devel] Endless Loop in Lisp when ,loading twice
by Stefan Kamphausen 04 Nov '05

04 Nov '05

Hi, currently I am experiencing a quite strange behaviour with slime and (I think) clsql: I had two lisp-files with def-view-class'es for two different databases, plus some interface functions. Everything worked fine. However, I had to merge the databases into different tables in one database and merrged the two lisp-files, too. Since then I two strange things happen: 1. First time I load the whole project (,l at the REPL) everything is fine (except for repeated compilations of some files in clsql; I posted that to clsql-help-list). But when I try to do that again, the files get compiled and then lisp goes into an endless loop, growing in memory as long as I say sayoonara to the REPL. 2. When trying to go to a function defined in that file (M-.) I end up at the previous defun in that file. I use CVSHEAD slime, CMU Common Lisp CVS 19b 19b-release-20050628-3 + minimal debian patches (19B), XEmacs 21.4 (patch 17) "Jumbo Shrimp" XEmacs Lucid. Can anyone give my any pointers what may be wrong here? Regards and thanks, Stefan -- 10 GB Mailbox, 100 FreeSMS/Monat http://www.gmx.net/de/go/topmail +++ GMX - die erste Adresse f�r Mail, Message, More +++

2 3

[slime-devel] slime stopped working
by Jeff Cunningham 04 Nov '05

04 Nov '05

I had this problem once before and I can't for the life of me remember what I did to fix it. I just downloaded the latest CVS of slime (todays) and blew away allthe fasls like always. Then I started emacs and when I ran M-x slime I get the following: (load "/home/jcunningham/installed/slime/swank-loader.lisp" :verbose t) (swank:start-server "/tmp/slime.10888" :external-format :iso-latin-1-unix) Error in function UNIX::SIGSEGV-HANDLER: Segmentation Violation at #x100168E8. [Condition of type SIMPLE-ERROR] Restarts: 0: [ABORT] Skip remaining initializations. Debug (type H for help) (UNIX::SIGSEGV-HANDLER #<unused-arg> #<unused-arg> #.(SYSTEM:INT-SAP #x3FFFEC98)) Source: Error finding source: Error in function LISP::%ENUMERATE-SEARCH-LIST: Undefined search list: default 0] I'm running CMUCL 19b and it hasn't changed in awhile. I'm stumped. Does anyone have any ideas? --Jeff

2 2

[slime-devel] Source display not working in SLDB
by Jonathan Lee 04 Nov '05

04 Nov '05

When I am in the SLDB, the manual states that I can display the associated source code by hitting the <v> key. However, when I do this I just get a message in the minibuffer: Cannot find source for frame: #<address #??????> NIL #<address #??????> The center value is sometimes T. I am running CLISP 2.35 as my base LISP. Is this the problem or is the function broken for all Lisp implementations?

2 1

[slime-devel] input not getting through
by Daniel Patru 04 Nov '05

04 Nov '05

The following code: (defun prompt () (do ((line nil (and (listen) (read)))) ((eq line 'q)) (when line (print line)) (print '>) (sleep 5))) when run in directly in the clisp interpreter in a shell, will output a > every 5 seconds, and print out any symbols entered until q is entered. when run under slime, only > is printed every 5 seconds, the input is not printed out, instead a message in the minibuffer says: ; pipelined request... (swank:listener-eval "<input>") Anyway to make this work? (I'm trying to do some processing in between user input.) I'm running clisp 2.34, slime 1.2.1, emacs 21.3.1 and windows xp. Thanks, Dan

3 2

[slime-devel] Current way to use multiple lisp implementations?
by Cody Koeninger 01 Nov '05

01 Nov '05

I'd like to be able to easily choose between several lisp implementations when starting slime (sbcl, cmucl, clisp). The most recent message I was able to find, on http://article.gmane.org/gmane.lisp.slime.devel/3932 suggested using slime-register-lisp-implementation, but I see that this function has been deleted. What's the current 'right' way to do this?

5 8