slime-devel

slime-devel@common-lisp.net

4231 discussions

[slime-devel] PATCH: ABCL on Windows
by Vladimir V. Korablin 30 Apr '09

30 Apr '09

By default, ABCL's streams use the platform-native end-of-line conventions. This makes is impossible for Slime to interact with Swank when Swank is running on ABCL/Windows, since Slime expects Unix EOLs. The attached patch fixes that by explicitly setings the EOL convention for Swank's network streams to LF.

4 5

[slime-devel] [Patch] typep and check-type extended arglists
by Stas Boukarev 30 Apr '09

30 Apr '09

The following patch adds typep/check-type and the extended arglists. -- With best regards, Stas.

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 30 Apr '09

30 Apr '09

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1729 slime/ChangeLog:1.1733 --- slime/ChangeLog:1.1729 Tue Apr 28 16:41:32 2009 +++ slime/ChangeLog Wed Apr 29 18:29:18 2009 @@ -1,3 +1,33 @@ +2009-04-30 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-backend.lisp ([default] declaration-arglist): Add + arglist of DECLARATION declaration. + + * swank-openmcl.lisp ([eql 'optimize] declaration-arglist): + Implement it for CCL. + (describe-symbol-for-emacs): Add :TYPE. + (describe-definition): Add :TYPE. + + Adapted from patch by Stas Boukarev. + +2009-04-29 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-extend-region-for-font-lock): (nth 0 + <parser-state>) may return negative numbers. + +2009-04-29 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el: Fix infinite loop during fontification introduced + by yesterday's changeset. + + (slime-region-for-tlf-at-point): New. Like + `slime-region-for-defun-at-point' but tries harder to get the + toplevel form right. + (slime-region-for-extended-tlf-at-point): Previously + `slime-region-for-extended-defun-at-point'. + (slime-extend-region-for-font-lock): Use it. + (slime-mark-defun-for-font-lock): Ditto. + 2009-04-28 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el: Fix fontification of suppressed (by reader Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.200 slime/contrib/ChangeLog:1.201 --- slime/contrib/ChangeLog:1.200 Tue Apr 21 09:03:41 2009 +++ slime/contrib/ChangeLog Wed Apr 29 19:26:36 2009 @@ -1,3 +1,9 @@ +2009-04-30 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-parse.el (slime-parse-extended-operator/proclaim): New. + (slime-extended-oprator-name-parser-alist): Adapt tbe entry for + PROCLAIM. + 2009-04-21 Tobias C. Rittweiler <tcr(a)freebits.de> * slime-indentantion-fu.el (slime-update-local-indentation): Save

1 0

[slime-devel] [Patch] some fixes regarding type declarations
by Stas Boukarev 29 Apr '09

29 Apr '09

The following patch does this, (i already sent optimize on ccl bit, but since it's somewhat related, i included it too): swank-backend.lisp (declaration-arglist): Add declaration `declaration'. swank-openmcl.lisp (describe-symbol-for-emacs): Add information about type specifiers. (declaration-arglist ((decl-identifier (eql 'optimize)))): new method. contrib/slime-parse.el (slime-parse-extended-operator/declare): Fix ability to provide arglist for (proclaim '(...)). Remove unused variables, slightly refactor. -- With best regards, Stas.

2 1

[slime-devel] Slime opening files behind users back
by Nikodemus Siivola 29 Apr '09

29 Apr '09

Currently some part of the compilation note magic opens files where notes occur if they are not already open. This can be highly annoying if those files eg. set buffer local variables that are not considered safe, so Emacs asks what to do about them. It is undesirable in other ways as well to me -- consider eg. a slow NFS partition, or compiling on a remote host. Ideally I would like it if the opening were deferred to the point in time where the user browses the compilation note buffer, and hits enter to be taken to the source location in question. As a stopgap measure, something like this -- used in appropriate places -- would also make life better for me: (defvar slime-try-set-buffer-opens-files :ask "Defines the behaviror or `slime-try-set-buffer': if T opens files without asking; if NIL never opens a file; if anything else permission asks before opening.") (defun slime-try-set-buffer (reason filename &optional nowarn) "Returns the buffer for FILENAME if one exists. Behavior depends on `slime-try-set-buffer-opens-files'." (let ((buffer (or (get-buffer filename) (when (and slime-try-set-buffer-opens-files (or (eq slime-maybe-set-buffer-opens-files t) (y-or-n-p "Open %s for %s? " filename reason))) (find-file-noselect filename nowarn))))) (when buffer (set-buffer buffer)))) Cheers, -- Nikodemus

1 0

[slime-devel] Problem with slime-compile-and-load-file
by Nikodemus Siivola 29 Apr '09

29 Apr '09

If compile-file returns a tertiary value of T, the file is not loaded. This is not a huge issue per se, but lack of clear notification of the fact is IMO a problem: time can be wasted debugging, trying to figure out what went wrong, without ever realizing that the .fasl was never loaded. A message is not visible enough: it's too easy to do something else and not notice it before it disappears -- I'd prefer something like: Compilation of foo.lisp failed. Load the resulting fasl anyways? (y or n) Cheers, -- Nikodemus

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 29 Apr '09

29 Apr '09

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1728 slime/ChangeLog:1.1729 --- slime/ChangeLog:1.1728 Sat Apr 25 04:53:16 2009 +++ slime/ChangeLog Tue Apr 28 16:41:32 2009 @@ -1,3 +1,22 @@ +2009-04-28 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el: Fix fontification of suppressed (by reader + conditionals) forms. That is make it reliably and totally work. + + (slime-region-for-extended-defun-at-point): New. Like + `slime-region-for-defun-at-point' but takes preceding reader + conditionals into account. + (slime-extend-region-for-font-lock): New. Make sure that + fontification operates on regions spanning a whole toplevel form + only. So we never operate within the context of a reader + conditional and we never miss any of those. + (slime-search-suppressed-forms): Remove ignore-errors; not needed + anymore now as we extend the region for fontification. + (slime-mark-defun-for-font-lock): New. + (slime-activate-font-lock-magic): Push + `slime-extend-region-for-font-lock' onto + `font-lock-extend-region-functions'. + 2009-04-25 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-show-description): Put the connection name into

2 2

[slime-devel] REPL buffer is not created in Vista
by Adam Dodek Michalik 29 Apr '09

29 Apr '09

Hey, Due to set of coincidents, I need to do some Lisp development on Windows Vista. I've got working SBCL 1.0.22, GNU Emacs 22.3 and slime from cvs. The problem is, after setting everything up as described in README, doing M-x slime won't cause slime repl buffer to pop up - instead, after inferior-lisp buffer is created and sbcl is run, the prompt suprisingly shows up in inferior-lisp buffer. Needless to say, it doesn't work at all. Here's how it looks like: http://wstaw.org/images/free/2009/04/28/26b7795f19bd27997b1eed6879338f.jpeg I will provide additional info if needed. -- Adam Michalik vel Dodek Dodecki <dodecki[at]gmail.com>

3 2

[slime-devel] Split windows resizing
by Patrick May 28 Apr '09

28 Apr '09

I'm sure this is something simple, but I can't find the answer via Googling. I'm running Slime in Aquamacs and I use split windows to see both my code and the repl. I initially set the repl window to be just 10 lines or so with c-u-55-x-2, but when any scrolling output appears, it gets resized to half the available frame. Is there a way I can tell Slime to leave my window sizes alone? Thanks, Patrick ---- pjm(a)spe.com S P Engineering, Inc. Large scale, mission-critical, distributed OO systems design and implementation. (C++, Java, Common Lisp, Jini, middleware, SOA)

2 1

[slime-devel] Wrong type argument: arrayp, nil for C-c C-s
by Tamas K Papp 28 Apr '09

28 Apr '09

Hi, When trying to complete a symbol with C-c C-s, I get the error message Wrong type argument: arrayp, nil in the *Messages* buffer, I see char-syntax: Wrong type argument: arrayp, nil in *slime-events*: (:emacs-rex (swank:arglist-for-echo-area '(("make-instance") ("bind") ("defun")) :arg-indices '(0 2 5) :print-right-margin 1000 :print-lines 1) ":istc-sunk" t 29) (:return (:ok "(make-instance class &allow-other-keys &rest initargs)") 29) (:emacs-rex (swank:arglist-for-echo-area '(("make-instance" "'model") ("bind") ("defun")) :arg-indices '(2 2 5) :print-right-margin 1000 :print-lines 1) ":istc-sunk" t 30) (:return (:ok "(make-instance 'model &key r sigma delta q h b beta m alpha theta &allow-other-keys &rest ===> initargs <===)") 30) The cursor was standing at the end of this form: (make-instance 'model I am using the latest SLIME from CVS. Thanks, Tamas

2 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

slime-devel