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

[slime-devel] Internals Wiki?
by Svein Ove Aas 05 Jul '05

05 Jul '05

I'm currently working on a stepper for Slime, and would love to have somewhere I can write down its internal structure; for my own reference, if nothing else. As I'll probably have to work through some of the other Slime code, and having such documentation be public seems like a good idea, I figured I should propose that a Wiki is put somewhere accessible. What do you think?

4 3

[slime-devel] Re: It's not just me! [was: CLISP on Windows?]
by Hoehle, Joerg-Cyril 04 Jul '05

04 Jul '05

Hi, Karsten Poeck wrote: >Summary: >clisp cvs head cygwin works with slime cvs head >clisp cvs head mingw (no need for cygwin) works with slime cvs head >clisp released version for windows from sourceforge does not >work (starts >up, loads slime but does not connect) I just build an MS-VC 6.0 version last friday. Would you like me to send that to you for testing? [Note that I'm not the one who did the official clisp-win32 distributions, built -- I believe -- with MS-VC 6.0(?)] Helmut Eller wrote: >I think I fixed that in CVS by sending an extra newline. I don't know >why, but it seems to work for me. Did you try adding any other presumably superfluous character beside #\Newline? Maybe there was/is some problem with read ahead, not with line-terminated buffering? Regards, Jorg Hohle.

2 1

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 03 Jul '05

03 Jul '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.719 slime/ChangeLog:1.720 --- slime/ChangeLog:1.719 Sat Jul 2 19:49:18 2005 +++ slime/ChangeLog Sun Jul 3 17:55:59 2005 @@ -1,3 +1,37 @@ +2005-07-03 Joerg Hoehle <hoehle(a)users.sourceforge.net> + + * swank-clisp (describe-symbol-for-emacs): Report :setf and :type + where appropriate. + +2005-07-03 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (next-single-char-property-change) + (previous-single-char-property-change) [xemacs]: Only define them + if not present. + (next-char-property-change, previous-char-property-change): Define + if needed. + + * README: Show examples for the filenames instead of the general + "/the/path/to/this/directory". Suggested by Brandon J. Van Every. + + * swank-corman.lisp (default-directory): Return a namestring + instead of the pathname. + (inspect-for-emacs, inspect-structure): Teach the inspector how to + deal with structures. + (spawn, send, receive): Implement rudimentary threading support. + It's now possible to connect with the :spawn communication style + and to bring up a listener. Unfortunately, debugging the + non-primary threads doesn't work at all. Still no support for + interrupt-thread. + + * slime.el (slime-start-swank-server): Send an extra newline + before the "(swank:start-server ...". I don't know why, but this + seems to fix the problem when starting CLISP/Win32. Interrupting + CLISP/W32 is still horribly broken. + + * swank-loader.lisp (compile-files-if-needed-serially) [corman]: + force-output after each file. + 2005-07-02 Marco Baringer <mb(a)bese.it> * slime.el (save-some-lisp-buffers): New Function.

1 0

[slime-devel] Projects
by Nicolas Neuss 03 Jul '05

03 Jul '05

Hello, with help from c.l.l, I have made ASDF putting my FASLs to different directories depending on the implementation. Unfortunately, SLIME's C-c C-k still puts FASLs in the directory where the file lies. This looks like a problem which can only be solved by considering Lisp files as parts of ASDF systems (similar to what IDEs for other languages do with projects). Maybe this is a direction to take for SLIME? Yours, Nicolas. P.S.: Some further ideas: Maybe SLIME could somehow search for a system for the file when it is loaded. Or one could think of assisting SLIME by putting the system name in a comment at the beginning of the file.

2 1

[slime-devel] Indenting and matching square brackets
by Steve Huffman 03 Jul '05

03 Jul '05

I'm having trouble matching and indenting square brackets. All is well in *scratch* and in emacs-lisp mode, but in Lisp Slime I cannot get square brackets to match or auto-indent. It's certainly not mission-critical, but it's definitely a nuisance. Thanks for any help, Steve

2 1

[slime-devel] a question and a bug
by Larry D'Anna 03 Jul '05

03 Jul '05

The question: I'd really like to be able to use slime with a unix domain socket. I saw some stuff about this in the archives but it wasn't clear whether or not there were any plans on actually incorporating this into slime. So my question is: is anyone working on this? and if not would you accept a patch that did it? The bug: For some reason slime sets the common-lisp-indent-function property on symbols instead of the lisp-indent-function property, which is apparently what emacs actually uses. The bug is fixed by throwing the following into slime-handle-indentation-update: (put symbol 'lisp-indent-function indent)

2 1

[slime-devel] CLISP on Windows?
by Peter Seibel 03 Jul '05

03 Jul '05

So I was under the impression that CLISP and Slime Just Worked on Windows. However I'm trying to build the CLISP lispbox and it's not quite working. As far as I can tell the call to comint-send-string in slime-start-swank-server is not doing much. If I manually call (swank:start-server ...) in the *inferior-lisp* buffer, passing it the name of the file that slime is polling, everything starts up fine. Is there some obvious reason why that might not be working. I'm using the vanilla win32 binaries of CLISP (2.33.2) and Emacs (21.3) supplied by their respective developers and the 1.2.1 tar of SLIME. -Peter

6 13

[slime-devel] patch: improve CLISP slime-apropos-* output: setf and type
by Hoehle, Joerg-Cyril 03 Jul '05

03 Jul '05

Hi, with that (any my prior patches), SLIME yields valuable apropos output. Here are some examples: bignum Type: (not documented) bit Function: (not documented) Setf: (not documented) Type: (not documented) bit-vector Type: (not documented) Class: (not documented) load-time-value Special Operator: (not documented) locally Macro: (not documented) Setf: (not documented) ext:local Macro: Return the closure defined locally with LABELS or FLET. Setf: Modify the local definition (LABELS or FLET). Although most stuff in CLISP has no documentation string, I find it already valuable to appear there as type/class/variable/ setf-able/macro/special-operator/generic-function/function. Enjoy, Jorg Hohle.

2 1

[slime-devel] Comment indentation in newer Emacsen
by Edi Weitz 03 Jul '05

03 Jul '05

Dumb question number two for today: I've recently switched to a CVS build (circa end of January) of GNU Emacs on Windows. Now, if I have a function like this in a lisp mode buffer (defun foo () "This is a very very very very very very very very very very very very long comment line." (bar x)) and press M-q within the comment string the result looks like so: (defun foo () "This is a very very very very very very very very very very very very long comment line." (bar x)) However, in 21.3 the result looked like this (defun foo () "This is a very very very very very very very very very very very very long comment line." (bar x)) which is what I prefer. Is this a bug or can the behaviour be changed with some setting that I'm missing? Thanks, Edi.

3 3

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 02 Jul '05

02 Jul '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.718 slime/ChangeLog:1.719 --- slime/ChangeLog:1.718 Fri Jul 1 15:52:55 2005 +++ slime/ChangeLog Sat Jul 2 19:49:18 2005 @@ -1,3 +1,10 @@ +2005-07-02 Marco Baringer <mb(a)bese.it> + + * slime.el (save-some-lisp-buffers): New Function. + (slime-repl-only-save-lisp-buffers): New customizable variable. + (slime-repl-compile-and-load): Use save-some-lisp-buffers. + (slime-oos): Use save-some-lisp-buffers. + 2005-07-01 Gabor Melis <mega(a)hotpop.com> * swank-sbcl.lisp (threaded stuff): make SBCL 0.9.2.9+ work while

1 0