slime-devel July 2006

slime-devel@common-lisp.net

26 participants
34 discussions

[slime-devel] patch: Allegro function profiling
by Willem Broekema 28 Jul '06

28 Jul '06

Hi all, Here's a patch for profiling functions on Allegro (except for profile-package). Please let me know if there is an alternative for the ugly y-or-n-p overruling. - Willem

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 25 Jul '06

25 Jul '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.921 slime/ChangeLog:1.922 --- slime/ChangeLog:1.921 Sun Jul 16 11:25:41 2006 +++ slime/ChangeLog Mon Jul 24 10:01:22 2006 @@ -1,3 +1,40 @@ +2006-07-24 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + Add support for destructuring macro arglists in arglist display, + form completion, and keyword completion; in particular for + with-open-file. + + * swank.lisp (find-valid-operator-name): New, factored out from + arglist-for-echo-area. + (arglist-for-echo-area): Use it here. + (print-arglist): New, factored out from decoded-arglist-to-string. + Handle recursive arglist structures that arise in destructuring + macro arglists. + (decode-required-arg, encode-required-arg): New, handle + destructuring patterns. + (decode-keyword-arg, encode-keyword-arg, decode-optional-arg) + (encode-optional-arg, decode-arglist, encode-arglist): Use them + here to handle destructuring patterns. + (print-decoded-arglist-as-template): Change interface, handle + destructuring patterns. + (decoded-arglist-to-template-string): Use it here. + (enrich-decoded-arglist-with-keywords): New, factored out from + enrich-decoded-arglist-with-extra-keywords. + (enrich-decoded-arglist-with-extra-keywords): Use it here. + (compute-enriched-decoded-arglist): New generic function, factored + out from arglist-for-insertion, form-completion. Add specialized + method for with-open-file. + (arglist-for-insertion, form-completion): Use it here. + (arglist-ref): New. + (completions-for-keyword): Change interface, handle destructuring + macro arglists. + + * slime.el (slime-enclosing-operator-names): For nesting levels + without operator, record nil. + (slime-completions-for-keyword): New argument arg-indices. + (slime-contextual-completions): Pass all enclosing operators and + arg-indices to slime-completions-for-keyword. + 2006-07-16 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> * slime.el (slime-edit-definition): Invoke the @@ -7,6 +44,7 @@ rather than find-tags). 2006-07-15 Juho Snellman <jsnell(a)iki.fi> + * swank-sbcl.lisp (preferred-communication-style): Remove use of linux_no_threads_p alien variable (the value has been hardcoded to false for about a year), so that we can also remove it from from SBCL

1 0

[slime-devel] [SBCL/FSF Emacs/Fedora] SLIME Doesn't Start
by Robert Uhl 22 Jul '06

22 Jul '06

So I've been running SLIME at home for many months now--it's a great bit of work: thanks guys! I figured that it's time to start sing CL in the office, maybe port some of my work over from Python. So I downloaded the latest version of SLIME, invoked the appropriate .emacs incantations & then ran M-x slime. It opens *inferior-lisp*, compiles itself, prints ';; Swank started at port: XXX' (where XXX can be whatever), then SBCL prints its '*' prompt and...nothing more. The SLIME REPL never appears. SBCL is still running; in fact, code typed into the SBCL REPL in *inferior-lisp* runs & prints as usual. My home machine is running Fedora Core 4, kernel 2.6.16-1.2111_FC4, emacs 21.4-5, SBCL 0.9.14 and an old SLIME from CVS that was current once-upon-a-time. My work machine is running Fedora Core 5, kernel 2.6.16-1.2107_FC5, emacs 21.4-14, SBCL 0.9.14 (built the same day, by the vendor, as the FC4 SBCL) and I've tried: current CVS SLIME; downloadable packaged SLIME 2.0; the same SLIME as used at home. I've also created an entirely new user with a clean .emacs and profile, then installed SLIME for him; he had the identical issue. Anyone have any ideas? I'm afraid that my lisp-foo is pretty weak, so I'm not even certain where to start looking:-( -- Robert Uhl <http://public.xdi.org/=ruhl> Whenever you walk by a computer and see someone using pico, be kind. Pause for a second and remind yourself that: `There, but for the grace of God, go I.' --Harley Hahn

3 3

[slime-devel] Re: [Sbcl-devel] Swank/SBCL Problem
by Bill Atkins 20 Jul '06

20 Jul '06

With that change, I'm no longer getting the backtrace. However, Swank starts up at port 37470 and then the echo area shows "Connecting to Swank on port 37470..". It simply hangs here. This is the latest CVS SLIME and SBCL (as of ten minutes ago). On 7/19/06, Christophe Rhodes <csr21(a)cam.ac.uk> wrote: > > Juho Snellman <jsnell(a)iki.fi> writes: > > > <atkinw(a)rpi.edu> wrote: > >> Using the latest SBCL and the latest SLIME, I'm having trouble > load-op'ing > >> swank (or systems that depend on it). I either evaluate (asdf:oos > >> 'asdf:load-op :swank) or do ,load-system swank RET. In either case, > ASDF > >> goes to work and some output is printed ot the REPL buffer. Then the > output > >> abruptly stops. the *inferior-lisp* buffer shows a backtrace > (attached). > >> This is an x86 Linux machine. I tried clearing ~/.slime/fasl as well. > > > > http://common-lisp.net/pipermail/slime-devel/2006-May/004955.html > > > > (I think Christophe has a patch to make Swank do the right thing here, > > which was something like replacing the default DEFMETHOD in > > DEFINTERFACE with a specialization on NO-APPLICABLE-METHOD. But it > > looks like it was never sent to the list.) > > Actually, I did, but in June :-) > > <http://common-lisp.net/pipermail/slime-devel/2006-June/005028.html> > > Cheers, > > Christophe > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share > your > opinions on IT & business topics through brief surveys -- and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Sbcl-devel mailing list > Sbcl-devel(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/sbcl-devel > -- Bill Atkins

1 2

[slime-devel] Swank/SBCL Problem
by Bill Atkins 19 Jul '06

19 Jul '06

Using the latest SBCL and the latest SLIME, I'm having trouble load-op'ing swank (or systems that depend on it). I either evaluate (asdf:oos 'asdf:load-op :swank) or do ,load-system swank RET. In either case, ASDF goes to work and some output is printed ot the REPL buffer. Then the output abruptly stops. the *inferior-lisp* buffer shows a backtrace (attached). This is an x86 Linux machine. I tried clearing ~/.slime/fasl as well. -- Bill Atkins

2 1

[slime-devel] Built CLISP 2.39 with cygwin....
by Richard Levenberg 19 Jul '06

19 Jul '06

And now SLIME wont load: I have (add-to-list 'load-path "C:\\DVP\\slime") (require 'slime) (slime-setup :autodoc t) (add-hook 'lisp-mode-hook (lambda () (slime-mode t))) in my .emacs file and when I do M-x slime it says (load "c:/DVP/slime/swank-loader.lisp" :verbose t) (swank:start-server "c:/slime.2860" :external-format :iso-latin-1-unix) [1]> *** - LOAD: A file with name /c/DVP/slime/swank-loader.lisp does not exist The following restarts are available: ABORT :R1 ABORT Break 1 [2]> *** - READ from #<INPUT CONCATENATED-STREAM #<INPUT STRING-INPUT-STREAM> #<IO TERMINAL-STREAM>>: there is no package with name "SWANK" I know that this is because CLISP is compiled with cygwin which wants pathnames like /cygdrive/c/DVP and emacs is not compiled with cygwin. My question is what to do about it. Im not a Lisp expert yet. r

2 1

[slime-devel] Output from Lisp to slime-repl
by Lynn Quam 17 Jul '06

17 Jul '06

I am attempting to better integrate the use of the SLIME REPL with my application. I have an event loop that listens for tcl/tk events and dispatches to Lisp event handlers. Some of the event handlers would like to insert "results" into the *slime-repl* buffer and bind *, + and / in the same manner as is done in swank::listener-eval. What is the "proper" way to do this with swank? I have tried to use swank:send-to-emacs, but Emacs is not expecting to see the (:present ...) forms that I send. The following seems to work, but there might be a better way: (swank:eval-in-emacs '(slime-repl-eval-string "(swank::get-freedius-repl-output)") where SWANK::GET-FREEDIUS-REPL-OUTPUT returns the multiple-values that I want to appear in the *slime-repl* buffer.

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 17 Jul '06

17 Jul '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.920 slime/ChangeLog:1.921 --- slime/ChangeLog:1.920 Sat Jul 15 07:03:29 2006 +++ slime/ChangeLog Sun Jul 16 11:25:41 2006 @@ -1,3 +1,11 @@ +2006-07-16 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + * slime.el (slime-edit-definition): Invoke the + slime-edit-definition-fall-back-function also in the case where + find-definitions-for-emacs returns an error message. + (slime-edit-definition-fallback-function): Fix typo (find-tag + rather than find-tags). + 2006-07-15 Juho Snellman <jsnell(a)iki.fi> * swank-sbcl.lisp (preferred-communication-style): Remove use of linux_no_threads_p alien variable (the value has been hardcoded to

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 16 Jul '06

16 Jul '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.918 slime/ChangeLog:1.920 --- slime/ChangeLog:1.918 Thu Jul 13 16:09:19 2006 +++ slime/ChangeLog Sat Jul 15 07:03:29 2006 @@ -1,3 +1,22 @@ +2006-07-15 Juho Snellman <jsnell(a)iki.fi> + * swank-sbcl.lisp (preferred-communication-style): Remove use of + linux_no_threads_p alien variable (the value has been hardcoded to + false for about a year), so that we can also remove it from from SBCL + in the future. + (*definition-types*): defcondition -> define-condition, + to make slime-show-definitions display condition FOO as + (DEFINE-CONDITION FOO) instead of (SWANK-BACKEND::DEFCONDITION FOO). + +2006-07-15 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + * slime.el (slime-shared-lisp-mode-hook): New function, factored + out from slime-lisp-mode-hook. + (slime-lisp-mode-hook): Use it here. + (slime-scheme-mode-hook): New function, use + slime-shared-lisp-mode-hook. + (slime-setup): If scheme-mode is one of the slime-lisp-modes, + install our hook. + 2006-07-13 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> * swank.lisp (keywords-of-operator): New support function for

1 0

[slime-devel] REPL directory and package stacks
by Taylor R Campbell 14 Jul '06

14 Jul '06

In a fresh SLIME instance, the following sequence of commands will leave the REPL in the CL-USER package, rather than the FOO package: ,change-package foo ,push-package bar ,pop-package This is because the bottom of the package stack is always the initial package in which SLIME starts up; the first ,PUSH-PACKAGE doesn't save the current package, but rather pushes the *new* package onto the stack, so FOO is forgotten. I'd expect that the package stack would be a stack of the packages that were current before ,PUSH-PACKAGE commands; is there a reason that it is not so? (The same goes for the directory stack.) If not, I've put a reimplementation of the relevant functions and REPL shortcuts at <http://paste.lisp.org/display/22487> with the behaviour that I expected (and a couple other folks too).

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

slime-devel July 2006