January 2006 - slime-devel - mailman3.common-lisp.net

[slime-devel] Slime support for metaclasses...
by Pascal Costanza 31 Jan '06

31 Jan '06

Hi, I have seen a discussion on slot-boundp / slot-makunbound and their MOP counterparts. I have tried to respond, but wasn't subscribed to this mailing list yet. This time I should get through... Most CLOS MOP implementations indeed specialize the third arguments in the slot-xxx-using-class generic functions on slot definition metaobjects. There are two exceptions (one is well known, the other probably not so much): - LispWorks specializes on slot names. This is probably because the switch from slot names to slot definition metaobjects in the slot-xxx- using-class functions was a very late change in the CLOS MOP specification, and the LispWorks CLOS implementation seems to be based on some older version of PCL (?!?). - Allegro specializes slot-value-using-class and (setf slot-value- using-class) on slot definition metaobjects, BUT specializes slot- boundp-using-class and slot-makunbound-using-class on slot names!!! This is not documented anywhere, so it's probably a source of confusion. (I have only accidentally found out about this. I have already filed a bug report with Franz.) I don't know exactly the context of your discussion, but you could use the Closer to MOP compatibility layer which fixes the slot-xxx- using-class functions for LispWorks and Allegro so that all slot-xxx- using-class functions are specialized on slot definition metaobjects. However, this may have other drawbacks that I cannot judge from here... Cheers, Pascal -- Pascal Costanza, mailto:pc@p-cos.net, http://p-cos.net Vrije Universiteit Brussel, Programming Technology Lab Pleinlaan 2, B-1050 Brussel, Belgium

1 0

[slime-devel] comma shortcut doesn't work always
by Edi Weitz 31 Jan '06

31 Jan '06

Another minor glitch that I encountered is that the comma shortcut in the REPL doesn't always work. I can consistently reproduce it (GNU Emacs 21.3.50 and also 21.3 release) like this: 1. Make sure point is at the beginning of a REPL input line. 2. Split Emacs frame into two windows, one of them being the REPL the other one containing something else. 3. Hide REPL window with C-x 0. 4. Now type M-x slime. 5. Type the comma. At this point I'm not asked for a shortcut command in the minibuffer but the comma is simply entered into the REPL instead. Anybody else seeing this? Cheers, Edi.

10 25

[slime-devel] patch: show initargs/initforms in slime-echo-arglist
by Luís Oliveira 30 Jan '06

30 Jan '06

Hello, The attached patch extends slime-echo-arglist in order to display a classe's initargs and initforms when make-instance is detected. Here's an example: CL-USER> (defclass foo () ((a :initarg :initarg-a :initform nil) (b :initarg :initarg-b) c)) #<STANDARD-CLASS FOO> CL-USER> (make-instance<space> ;; minibuffer displays: (make-instance class &rest initargs) CL-USER> (make-instance 'foo<space> ;; minibuffer displays: (make-instance 'foo &key (initarg-a nil) initarg-b) The code is a bit ugly, specially on the elisp side. It could perhaps be generalized to support more cases like this. Suggestions on how to improve it are most welcome! (BTW, the ChangeLog seems to be use the latin-1 enconding, perhaps using utf-8 and/or adding a -*- coding: ... -*- line at the top would be a good idea?) -- Luís Oliveira luismbo (@) gmail (.) com Equipa Portuguesa do Translation Project http://www.iro.umontreal.ca/translation/registry.cgi?team=pt

2 2

[slime-devel] slime fails to start after Linux upgrade
by John Whittaker 30 Jan '06

30 Jan '06

I recently upgraded my Mandriva Linux system after reading about some KDE security flaw. I shouldn't have done it, but it's too late now. Among other problems, I now cannot run slime and cmucl from xemacs. I get the following error message: 'itimer "run-with-timer" signaled: (error "127.0.0.1/x^T-^H& Servname not supported for ai_socktype")' This behavior was the same for my old slime setup, which was 1.2.1, and my current which is the latest tarball from common-lisp.net. Does anyone have any idea what is causing this and how I can fix it? Thanks, John Whittaker

1 0

[slime-devel] Making potentially problematic changes
by Ian Eslick 27 Jan '06

27 Jan '06

I've updated the inspector in swank.lisp to support metaclass based slot-value overrides. This requires use of the mop functions slot-boundp-using-class and slot-value-using-class (and adding them to the swank-mop package) which may not be consistent over the various versions. Since the inspector is rather central to slime, I don't want to commit this change directly (and don't have write permissions to slime CVS (ieslick(a)common-lisp.net) Is there a proper procedure for testing a change like this if you don't have all the test configurations available locally? Thank you, Ian

2 1

[slime-devel] newbie REPL questions
by funkyj＠gmail.com 24 Jan '06

24 Jan '06

I'm not sure how the whole SLIME development cycle is suppose to work. One of the things that is confusing me is that selecting various options in SLDB causes the REPL to lose its connection with the inferior lisp, requiring me to do 'M-x slime' to reconnect. Here is an example: ;; starting slime for the first time in this emacs session M-x slime ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; messages Connecting to Swank on port 2369.. Process bridge is installed Connected. Hack and be merry! ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; repl buffer shows ; SLIME 2006-01-20 CL-USER> x ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; sldb clisp/0 buffer appears EVAL: variable X has no value [Condition of type SYSTEM::SIMPLE-UNBOUND-VARIABLE] Restarts: 0: [USE-VALUE] You may input a value to be used instead of X. 1: [STORE-VALUE] You may input a new value for X. 2: [ABORT-REQUEST] Abort handling SLIME request. 3: [ABORT] ABORT Backtrace: 0: #<SYSTEM-FUNCTION EVAL> 1: #<COMPILED-FUNCTION SWANK::EVAL-REGION> 2: #<COMPILED-FUNCTION SWANK::LISTENER-EVAL-1> 3: #<COMPILED-FUNCTION #:|273 275 (DEFINTERFACE CALL-WITH-SYNTAX-HOOKS (FN) ...)-31-3-1-1|> 4: #<COMPILED-FUNCTION SWANK::CALL-WITH-BUFFER-SYNTAX> 5: #<COMPILED-FUNCTION SWANK:LISTENER-EVAL> 6: #<SYSTEM-FUNCTION EVAL> 7: #<COMPILED-FUNCTION SWANK::EVAL-FOR-EMACS-1> 8: #<COMPILED-FUNCTION #:|472 477 (DEFINTERFACE CALL-WITH-DEBUGGER-HOOK (HOOK FUN) ...)-52-3-1-1|> 9: #<COMPILED-FUNCTION SWANK::EVAL-FOR-EMACS> 10: #<COMPILED-FUNCTION SWANK::READ-FROM-EMACS> 11: #<COMPILED-FUNCTION SWANK::HANDLE-REQUEST-1> 12: #<COMPILED-FUNCTION #:|472 477 (DEFINTERFACE CALL-WITH-DEBUGGER-HOOK (HOOK FUN) ...)-52-3-1-1|> 13: #<COMPILED-FUNCTION SWANK::CALL-WITH-CONNECTION-1> 14: #<COMPILED-FUNCTION SWANK::CALL-WITH-REDIRECTED-IO> 15: #<COMPILED-FUNCTION SWANK::MAYBE-CALL-WITH-IO-REDIRECTION> 16: #<COMPILED-FUNCTION SWANK::CALL-WITH-CONNECTION> 17: #<COMPILED-FUNCTION SWANK::HANDLE-REQUEST> 18: #<COMPILED-FUNCTION SWANK::SIMPLE-SERVE-REQUESTS-2> 19: #<COMPILED-FUNCTION SWANK::SIMPLE-SERVE-REQUESTS> 20: #<COMPILED-FUNCTION SWANK::SERVE-REQUESTS> 21: #<COMPILED-FUNCTION SWANK::SERVE-CONNECTION> 22: #<COMPILED-FUNCTION SWANK::SETUP-SERVER-SERVE> 23: #<COMPILED-FUNCTION SWANK::SETUP-SERVER> 24: #<COMPILED-FUNCTION SWANK:START-SERVER> 25: #<SYSTEM-FUNCTION SYSTEM::READ-EVAL-PRINT> 2 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; type '3' (abort) into the sldb buffer ;; the following minibuffer message appears: Lisp connection closed unexpectedly: exited abnormally with code 256 ;; type 'enter' in repl buf and I see: if: Not connected. Use `M-x slime' to start a Lisp. ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; Starting slime again with M-x slime works fine but takes a long ;; time. It seems a lot of actions cause the lisp connection to close unexpectedly. Is this normal? Should I just ignore it and restart slime? Restarting slime after each "mistake" (I make a lot of them) is rather tedious. FYI: I'm using clisp from cygwin: GNU CLISP 2.35 (2005-08-29) (built on winsteingoldlap.ad.alphatech.com [10.41.52.182]) Software: GNU C 3.4.4 (cygming special) (gdc 0.12, using dmd 0.125) gcc -W -Wswitch -Wcomment -Wpointer-arith -Wimplicit -Wreturn-type -Wmissing-declarations -Wno-sign-compare -O2 -fexpensive-optimizations -DUNICODE -DDYNAMIC_FFI -I. -x none -lintl libcharset.a libavcall.a libcallback.a -lreadline -lncurses -liconv -L/usr/local/libsigsegv-cygwin/lib -lsigsegv -L/usr/X11R6/lib -lX11 SAFETY=0 HEAPCODES STANDARD_HEAPCODES SPVW_PAGES SPVW_MIXED libsigsegv 2.2 libiconv 1.9 Features: (REGEXP SYSCALLS I18N LOOP COMPILER CLOS MOP CLISP ANSI-CL COMMON-LISP LISP=CL INTERPRETER SOCKETS GENERIC-STREAMS LOGICAL-PATHNAMES SCREEN FFI GETTEXT UNICODE BASE-CHAR=CHARACTER PC386 UNIX CYGWIN) C Modules: (clisp i18n syscalls regexp) Installation directory: /usr/lib/clisp/ User language: ENGLISH Machine: I686 (I686) jcano-lt.jnpr.net [192.168.1.103] and GNU Emacs 21.3.1 (i386-mingw-nt5.1.2600) of 2004-03-10 on NYAUMO SLIME, as you can see is the CVS head I checked out 2006-01-20. Thanks in advance, --jfc

2 2

[slime-devel] shortcuts to OPs other than ASDF's own
by Walter C. Pelissero 24 Jan '06

24 Jan '06

I was wondering if this patch could be considered for inclusion in SLIME. It allows to write repl shortcuts for operations not belonging to the ASDF package simply prefixing the operation string with the package name. diff -c -r1.92 swank-backend.lisp *** swank-backend.lisp 23 Oct 2005 08:47:54 -0000 1.92 --- swank-backend.lisp 24 Jan 2006 13:51:45 -0000 *************** *** 313,319 **** (error "ASDF is not loaded.")) (with-compilation-hooks () (let ((operate (find-symbol "OPERATE" :asdf)) ! (operation (find-symbol operation-name :asdf))) (when (null operation) (error "Couldn't find ASDF operation ~S" operation-name)) (apply operate operation system-name keyword-args)))) --- 313,323 ---- (error "ASDF is not loaded.")) (with-compilation-hooks () (let ((operate (find-symbol "OPERATE" :asdf)) ! (operation (let ((colon (position #\: operation-name))) ! (if colon ! (find-symbol (subseq operation-name (1+ colon)) ! (subseq operation-name 0 colon)) ! (find-symbol operation-name :asdf))))) (when (null operation) (error "Couldn't find ASDF operation ~S" operation-name)) (apply operate operation system-name keyword-args)))) -- walter pelissero http://www.pelissero.de

1 0

[slime-devel] New startup behavior
by Jonathon McKitrick 24 Jan '06

24 Jan '06

Hi all, I noticed that since the recent change in startup behavior, slime seems to compile a lot more code each time it is started. My older laptop takes a lot longer to start than before, when sbcl used asdf to load it, IIRC. Is there a way to avoid this delay with each start? -Jonathon -- "I am sure that Jesus would use a command prompt. Hello? Ten Commandments??"

1 0

[slime-devel] SLIME and CLX and CMUCL "lock up" problem
by David Trudgett 24 Jan '06

24 Jan '06

Hi everyone, I was directed to this mailing list by a fellow on comp.lang.lisp, so hopefully this is the right place to ask. I've been using SLIME for a while now (I'm just a Lisp newbie, however), and it has been really great. Since doing some CLX programming in CMUCL, I've noticed a lot of "freeze ups" where the Lisp subprocess just goes off and does its own thing and has to be sent a TERM signal, which means I also have to restart SLIME. Is this a known conflict between the CLX library and SLIME? If any of you guys has a suggestion about how I might work around this, or ideas about what else might be causing problems, it would be greatly appreciated. My environment is: GNU Emacs 21.4.1 SLIME 2005-12-27 (from CVS) CMUCL 19a CLX X Library MIT R5.02, (the one that comes with CMUCL) Thanks for doing a great job with SLIME! Bye for now, David -- David Trudgett http://www.zeta.org.au/~wpower/ Hickory Dickory Dock, The mice ran up the clock, The clock struck one, The others escaped with minor injuries.

1 0

[slime-devel] [Feature Request:] keeping history for several macroexpansions in a row
by Tobias C. Rittweiler 23 Jan '06

23 Jan '06

Hi Slime-hackers, I'd like to suggest keeping a stack-like history for several macroexpansions in a row such that you can easily go back one macroexpansion after you've macroexpanded a form /inside/ the macroexpansion buffer. Erm, yeah, sure, here's a small example: Let's say you have two macros like these: (defmacro with-foo (() &body body) `(flet ((foo (arg) arg)) ,@body)) (defmacro with-bar ((arg) &body body) `(with-foo () (prog1 (foo ,arg) ,@body))) Ah, well, hopefully yours will do something less stupid. Anyway, let's further say that you want to macroexpand (`C-c RET') the following form: (with-bar ('foobar) (let ((x 1)) (* x x))) the macroexpansion buffer should pop up containing something like (WITH-FOO NIL (PROG1 (FOO 'FOOBAR) (LET ((X 1)) (* X X)))) Now you want to expand that form as well, so you place point before `(WITH-FOO' and press `C-c RET' again, now the old expansion buffer disappears and instead a new one pops up in some other window. While it's cool that macroexpansion works from within the macroexpansion buffer, the current behaviour is unfortunately somewhat annoying, because upon exit from the expansion buffer (via pressing `q'), -- while the old buffer constellation is nicely recovered as of before any macroexpansion --, the point won't afterwards be in the same window where you started macroexpansion. (Split your emacs horizentally, on top the REPL, on the bottom the slime-scratch buffer; now paste the above form in the REPL and do a macroexpansion there, slime-scratch will be replaced by the expansion buffer; now expand again from inside the macroexpansion buffer, a new expansion buffer will be placed into the REPL window; there press `q' and the point will be within slime-scratch although you started from the REPL.) I suggest a slightly different behaviour when expanding a form from within the macroexpansion buffer: the old buffer should be reused and you should be able to go back to the previous macroexpansion (like what `l' does in the inspector.) Additionally I'd like to suggest a way to macroexpand into a newly created macroexpansion buffer that's independant from a possibly other expansion buffer that got created prior. The purpose is simply for side-by-side comparasions of the same macro but that was modified in between. Finally, I'm often hit by not releasing the control-key fast enough such that a `C-c C-RET' key sequence is generated. Maybe that could be made an alias for `C-c RET'? Keep up the great work! -*- tcr

2 1