slime-devel July 2010

slime-devel@common-lisp.net

17 participants
23 discussions

[slime-devel] Daily ChangeLog diff
by Helmut Eller 27 Jul '10

27 Jul '10

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.398 slime/contrib/ChangeLog:1.402 --- slime/contrib/ChangeLog:1.398 Thu Jul 22 21:46:34 2010 +++ slime/contrib/ChangeLog Sat Jul 24 19:39:24 2010 @@ -1,3 +1,29 @@ +2010-07-24 Stas Boukarev <stassats(a)gmail.com> + + * slime-package-fu.el (slime-frob-defpackage-form): Accept a + symbol or a list of symbols. Optimize inserting several symbols at + a time. + (slime-search-exports-in-defpackage): Search forward until nothing + is found, otherwise it searching for FOO will stop after encountering + FOO-B. + (slime-export-class): Rename from slime-export-structure. + + * swank-package-fu.lisp (export-symbol-for-emacs): Fix typo. + (export-structure): Add support for CCL and for exporting + standard-class accessors using MOP. + + * slime-sprof.el (slime-sprof-start-alloc) + (slime-sprof-start-time): New functions to start profiling in + :alloc and :time mode. The default slime-sprof-start is :cpu. + * swank-sprof.lisp (swank-sprof-start): Accept :mode keyword. + +2010-07-24 Stas Boukarev <stassats(a)gmail.com> + + * slime-package-fu.el (slime-export-structure): New function, + export all constructors, accessors, etc. + * swank-package-fu.lisp (export-structure): Lisp side of the above + function, works only on SBCL for now. + 2010-07-23 Stas Boukarev <stassats(a)gmail.com> * swank-arglists.lisp (arglist-dispatch): Export it, so it may be

3 2

[slime-devel] Daily ChangeLog diff
by Helmut Eller 23 Jul '10

23 Jul '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2114 slime/ChangeLog:1.2115 --- slime/ChangeLog:1.2114 Wed Jul 21 09:40:32 2010 +++ slime/ChangeLog Thu Jul 22 09:45:46 2010 @@ -1,3 +1,8 @@ +2010-07-22 Vitaly Mayatskikh <v.mayatskih(a)gmail.com> + + * swank-lispworks.lisp (list-callers-internal): Fix for LW6. + (list-callees-internal): New function, use it. + 2010-07-21 Stas Boukarev <stassats(a)gmail.com> * swank-sbcl.lisp (quit-lisp): Use sb-thread:terminate-thread Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.396 slime/contrib/ChangeLog:1.398 --- slime/contrib/ChangeLog:1.396 Fri Jul 16 03:34:23 2010 +++ slime/contrib/ChangeLog Thu Jul 22 21:46:34 2010 @@ -1,3 +1,11 @@ +2010-07-23 Stas Boukarev <stassats(a)gmail.com> + + * swank-arglists.lisp (arglist-dispatch): Export it, so it may be + extended more easily. + + * slime-fancy-inspector.el (slime-fancy-inspector): Add + slime-parse dependency. + 2010-07-16 Stas Boukarev <stassats(a)gmail.com> * slime-repl.el (slime-call-defun): Handle setf-functions.

1 0

[slime-devel] LispWorks 6 and list-callers/list-callees
by Vitaly Mayatskikh 22 Jul '10

22 Jul '10

Hi! Here's patch to bring list-callers back to life for LW6 and implementation of list-callees. Tested on LW 6.0.1. --- ChangeLog.~1.2114.~ 2010-07-22 13:48:06.000000000 +0200 +++ ChangeLog 2010-07-22 14:12:24.900354346 +0200 @@ -1,3 +1,8 @@ +2010-07-22 Vitaly Mayatskikh <v.mayatskih(a)gmail.com> + + * swank-lispworks.lisp (list-callers-internal): Fix for LW6. + (list-callees-internal): New function, use it. + 2010-07-21 Stas Boukarev <stassats(a)gmail.com> * swank-sbcl.lisp (quit-lisp): Use sb-thread:terminate-thread --- swank-lispworks.lisp.~1.136.~ 2010-03-06 19:29:09.000000000 +0100 +++ swank-lispworks.lisp 2010-07-22 14:04:35.266247655 +0200 @@ -697,8 +697,10 @@ function names like \(SETF GET)." (defxref who-macroexpands hcl:who-calls) ; macros are in the calls table too (defxref calls-who hcl:calls-who) (defxref list-callers list-callers-internal) -;; (defxref list-callees list-callees-internal) +#+lispworks6 +(defxref list-callees list-callees-internal) +#-lispworks6 (defun list-callers-internal (name) (let ((callers (make-array 100 :fill-pointer 0 @@ -716,6 +718,24 @@ function names like \(SETF GET)." (list 'function object) (or (dspec:object-dspec object) object))))) +#+lispworks6 +(defun list-callers-internal (name) + ;; Delay dspec:object-dspec until after sweep-all-objects + ;; to reduce allocation problems. + (loop for object in (hcl::who-calls name) + collect (if (symbolp object) + (list 'function object) + (or (dspec:object-dspec object) object)))) + +#+lispworks6 +(defun list-callees-internal (name) + ;; Delay dspec:object-dspec until after sweep-all-objects + ;; to reduce allocation problems. + (loop for object in (hcl::calls-who name) + collect (if (symbolp object) + (list 'function object) + (or (dspec:object-dspec object) object)))) + ;; only for lispworks 4.2 and above #-lispworks4.1 (progn -- wbr, Vitaly

3 2

[slime-devel] Daily ChangeLog diff
by Helmut Eller 22 Jul '10

22 Jul '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2113 slime/ChangeLog:1.2114 --- slime/ChangeLog:1.2113 Tue Jul 6 08:09:19 2010 +++ slime/ChangeLog Wed Jul 21 09:40:32 2010 @@ -1,3 +1,10 @@ +2010-07-21 Stas Boukarev <stassats(a)gmail.com> + + * swank-sbcl.lisp (quit-lisp): Use sb-thread:terminate-thread + instead of sb-ext:quit :recklessly-p t. This way + sb-ext:*exit-hooks* will be run. + Reported by Lorenz Mösenlechner. + 2010-07-06 Helmut Eller <heller(a)common-lisp.net> Find definition for (%primitive NAME ...)

1 0

[slime-devel] quit-lisp on sbcl
by Lorenz Mösenlechner 21 Jul '10

21 Jul '10

Hi, I just figured out that QUIT-LISP exits lisp without executing any exit-hook on a recent sbcl 1.0.38. The problem is that it tries to call SB-EXT:QUIT with RECKLESSLY-P set to T in each thread. That call immediately quits sbcl on the first call. Is there a reason for QUIT-LISP not using SB-THREAD:TERMINATE-THREAD? Cheers, Lorenz

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 17 Jul '10

17 Jul '10

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.395 slime/contrib/ChangeLog:1.396 --- slime/contrib/ChangeLog:1.395 Sun Jul 4 11:55:29 2010 +++ slime/contrib/ChangeLog Fri Jul 16 03:34:23 2010 @@ -1,3 +1,7 @@ +2010-07-16 Stas Boukarev <stassats(a)gmail.com> + + * slime-repl.el (slime-call-defun): Handle setf-functions. + 2010-07-04 Stas Boukarev <stassats(a)gmail.com> * swank-asdf.lisp (asdf:operation-done-p): Fix reloading on

1 0

[slime-devel] Slime distributed with GNU Emacs
by Tobias C Rittweiler 09 Jul '10

09 Jul '10

Some people were contacted by one of the current GNU Emacs maintainers and requested to assign copyright to the FSF in order to allow Slime be distributed with GNU Emacs. I'm still having mail server issues (seems like I'm only able to send mail on full moon), and hence could not participate in a short internal roundabout mail between Luke and Helmut. Helmut rightly said that the matter should be discussed on the mailing list anyway. So below follows undelivered mail, and represent my thoughts right after having read about the request. Seems futile to me. Too many contributors over the time. Will the SWANK side not be included? (Then how is that supposed to be distributed?) SWANK is actually PD, not GPL. There's this saying that Germans (and maybe Austrians, too?) cannot legally give up their copyright and hence cannot actually release stuff in public domain. So the situation of the SWANK side is pretty muddy to begin with. Supposing all these legal matters get sorted out, and someone of the FSF actually does all the legal work and gets the necessary assignments (an Heracles job in my eyes), and Slime becomes part of Emacs, then there's still the issue how future contributions are supposed to be dealt with and whether need for assignment may be regarded as an obstacle for contribution. OTOH, I don't see major contribution happening. And then yet another issue is synchronization between a Slime development tree, and the putative Slime Emacs tree; no one has so far stepped up who actually wants to take care of regular releases of Slime which then may become a prerequisite of some sort. In short: too much hassle for too little gain. On the other hand, it's not really my hassle, so I don't really object to the idea. I think there's a new packaging system being developed for Emacs for third-party contributions. It strikes me that may a better fit for Slime. -T.

5 7

[slime-devel] Daily ChangeLog diff
by Helmut Eller 07 Jul '10

07 Jul '10

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2112 slime/ChangeLog:1.2113 --- slime/ChangeLog:1.2112 Tue Jun 22 06:02:49 2010 +++ slime/ChangeLog Tue Jul 6 08:09:19 2010 @@ -1,3 +1,11 @@ +2010-07-06 Helmut Eller <heller(a)common-lisp.net> + + Find definition for (%primitive NAME ...) + + * swank-cmucl.lisp (template-definitions, primitive-definitions): + New functions. + (find-definitions): Use them. + 2010-06-22 Helmut Eller <heller(a)common-lisp.net> * swank-loader.lisp (*architecture-features*): ECL uses :x86_64.

1 0

[slime-devel] Distributing SLIME with Emacs 24
by Helmut Eller 06 Jul '10

06 Jul '10

On the emacs-devel list they discussed recently[1] whether SLIME should be distributed with Emacs 24. Chong Yidong (he's one of the two Emacs lead maintainers) asked me who the contributors would be and I gave him a list with names and email addresses. I expected that he would give up because the list is too long :-), but he went on and asked those people if they are willing to assign copyrights to the FSF. Assigning copyrights is of course your own decision. What I actually want to discuss is what people here think about distributing SLIME with Emacs and making that the "official" release. Since they only make Emacs releases every couple of years or so, we would probably have enough time to get our act together. Gnus seems to be maintained by a different group than the Emacs core developers and is synchronized now and then with Emacs's trunk. I don't know how this would work us. There's also the question how/if the Common Lisp part should be distributed. Including it with Emacs seems problematic, but separating it isn't attractive either. I guess if SLIME were distributed with Emacs it would be a bit easier for us to get certain things fixed in Emacs (e.g. LOOP indentation) which so far have been ignored by the Emacs developers. It seems like it would be more work than it's now but I don't want to close the door too early. What do others think about it? Are there any volunteers to do the syncing with Emacs? Helmut [1] http://article.gmane.org/gmane.emacs.devel/126696

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 05 Jul '10

05 Jul '10

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.394 slime/contrib/ChangeLog:1.395 --- slime/contrib/ChangeLog:1.394 Fri Jul 2 07:44:15 2010 +++ slime/contrib/ChangeLog Sun Jul 4 11:55:29 2010 @@ -1,3 +1,8 @@ +2010-07-04 Stas Boukarev <stassats(a)gmail.com> + + * swank-asdf.lisp (asdf:operation-done-p): Fix reloading on + ASDF2. ASDF2 no longer has `asdf:around' method combination. + 2010-07-02 Stas Boukarev <stassats(a)gmail.com> * swank-fancy-inspector.lisp (docstring-ispec): Don't insert an

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 2010