slime-devel

slime-devel@common-lisp.net

4231 discussions

[slime-devel] Re: [slime-cvs] CVS update: slime/b0rk.lisp
by Luke Gorrie 11 Dec '03

11 Dec '03

Dan Barlow <dbarlow(a)common-lisp.net> writes: > Added Files: > b0rk.lisp > Log Message: > source-path-file-position broken > Date: Thu Dec 11 01:59:07 2003 > Author: dbarlow I'm guessing this was an accident :-)

1 0

[slime-devel] window-point vs buffer point: making slime-goto-next-xref work
by Daniel Barlow 11 Dec '03

11 Dec '03

I don't know enough about emacs internals to explain why, but on my "GNU Emacs 21.2.1 (i386-debian-linux-gnu, X toolkit, Xaw3d scroll bars) of 2002-03-22 on raven, modified by Debian", I need the attached patch to make slime-goto-next-xref actually move the cursor in the xref buffer: without it, I never get past the first xref. Does it work for other people? I notice that there's no keybinding for slime-goto-next-xref or slime-next-location which usually calls it, so I could understand if it hadn't been tested. Or then again, maybe my setup is weird. Anyone want to take a stab at explaining? Two other points, seeing as I'm here - - pretty(ish) picture of SLIME building an ASDF system, via <URL:http://ww.telent.net/diary/2003/12/#11.16680>. - we have a problem with source path parsing which causes slime to be unable to compile Marty Hall's Memoization package. I've just checked in the relevant function as b0rk.lisp: try C-c C-k on this and (at least in SBCL) you get READER-ERROR at 1120 (line 28, column 36) on #<FILE-STREAM for "file \"/home/dan/src/cl.net/slime/b0rk.lisp\"" {9BD6389}>: unmatched close parenthesis which is clearly not the case. I suspect #' is confusing it, but I'm going to bed now anyway Enjoy Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.139 diff -u -w -r1.139 slime.el --- slime.el 11 Dec 2003 04:57:33 -0000 1.139 +++ slime.el 11 Dec 2003 06:12:07 -0000 @@ -3331,9 +3331,10 @@ (defun slime-goto-next-xref () "Goto the next cross-reference location." (let ((location (with-current-buffer (slime-xref-buffer) - (display-buffer (current-buffer) t) - (goto-char (next-single-char-property-change - (point) 'slime-location)) + (let ((w (display-buffer (current-buffer) t))) + (goto-char (1+ (next-single-char-property-change + (point) 'slime-location))) + (set-window-point w (point))) (cond ((eobp) (message "No more xrefs.") nil) -dan -- http://web.metacircles.com/ - Open Source software development and support

2 2

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 11 Dec '03

11 Dec '03

Index: slime/ChangeLog diff -u slime/ChangeLog:1.145 slime/ChangeLog:1.150 --- slime/ChangeLog:1.145 Sun Dec 7 20:44:08 2003 +++ slime/ChangeLog Wed Dec 10 23:57:33 2003 @@ -1,3 +1,74 @@ +2003-12-11 Daniel Barlow <dan(a)telent.net> + + * swank.lisp (swank-compiler): new function abstracts commonality + between swank-compile-{file, string}. + (swank-load-system): call swank-compiler to load asdf system + + * swank-sbcl.lisp (compiler-note-location and elsewhere): + remove all trace of *compile-filename* + (compile-*-for-emacs): shorten + + * swank-backend.lisp (call-with-compilation-hooks): new GF + should set up all appropriate error condition loggers etc + to do a compilation preserving the notes. Implement for + sbcl, cmucl + + * slime.el (slime-find-asd, slime-load-system): new command + to compile and load an ASDF system with all the usual compiler + notes and stuff + (slime-compilation-finished): if more than one file has new + errors/notes, create an xref buffer to show them all + (slime-remove-old-overlays): bug fix: now removes overlays even + at start of buffer + (slime-overlay-note): do nothing quietly if + slime-choose-overlay-region returns nil + (slime-choose-overlay-region): return nil if note has no location + +2003-12-11 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * slime.el (slime-repl-previous-prompt, slime-repl-next-prompt): + New commands. Suggested by Håkon Alstadheim. + (slime-repl-beginning-of-defun, slime-repl-end-of-defun): New + commands. Suggested by Andreas Fuchs. + (slime-repl-insert-prompt): Mark the prompt with a + slime-repl-prompt text property. + (slime-repl-eol): New function. Mostly for symmetry. + (slime-repl-in-input-area-p, slime-repl-at-prompt-end-p): New + predicates. + (slime-repl-find-prompt, slime-search-property-change-fn): New + functions. + (slime-ir1-expand): New command. + + * swank-cmucl.lisp (accept-connection, request-loop): Don't use + fd-handlers. The code for the request-loop itself is now almost + the same as in the Allegro version. + (print-ir1-converted-blocks, expand-ir1-top-level): New functions. + +2003-12-10 Daniel Barlow <dan(a)telent.net> + + * swank-sbcl.lisp (serve-request): more fiddling with serve-event + descriptors + + * slime.el (slime-repl-return): slime-check-connected, otherwise + pressing Return in an unconnected repl gets a bit weird + +2003-12-10 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank-allegro.lisp, swank-lispworks.lisp, swank-openmcl.lisp, + swank-sbcl.lisp (create-swank-server): Accept an announce-function + keyword argument. + + * swank.lisp (start-server): Pass an announce callback function to + create-swank-server. Works better with single threaded + implementations. + (announce-server-port, simple-announce-function): New functions. + (alistify): Doc fix. + + * swank-cmucl.lisp (create-swank-server): Use announce callback. + (sldb-disassemble): New function. + + * slime.el (sldb-disassemble): New command. Bound to D. + 2003-12-08 Luke Gorrie <luke(a)bluetail.com> * swank-cmucl.lisp (*debug-definition-finding*): Now nil by

1 0

[slime-devel] Re: [slime-cvs] CVS update: slime/swank-cmucl.lisp
by Helmut Eller 11 Dec '03

11 Dec '03

Dan Barlow <dbarlow(a)common-lisp.net> writes: > + ;; XXX is this _ever_ used? By what? *compile-file-truename* > + ;; should be set by the implementation inside any call to compile-file This is mostly used when a compilation condition is signaled at the end of a compilation unit, e.g, "those function <list of functions> are undefined" and also for notes about missing exported symbols in a defpackage form. The same situation occurs in SBCL with the "redefining function <foo>" style warnings. Set a breakpoint at this location to see the details. Helmut.

1 0

[slime-devel] Lispworks; Windows gotcha
by Alain.Picard＠memetrics.com 10 Dec '03

10 Dec '03

Currently, we have: (defun create-swank-server (port &key reuse-address) "Create a Swank TCP server on `port'. Return the port number that the socket is actually listening on." (declare (ignore reuse-address)) (comm:start-up-server-and-mp :announce *terminal-io* :service port :process-name "Swank Request Processor" :function 'swank-accept-connection ) ; tsk tsk; terrible style! port) >From the fine manual: Note: start-up-server-and-mp is implemented only on Unix/Linux/Mac OS platforms. Which means the current CREATE-SWANK-SERVER for lispworks will fail on Windows. Now, I hate windows as much as the next guy, but this seems unnecessarily rude; I suggest one of the following things: * revert to just comm:start-up-server. This has the advantage that many users either need to start up MP themselves later on. Once comm:start-up-server-and-mp runs, you can no longer launch MP yourself from a chosen entry point. What was the rationale of using x-and-mp? * #+Unix #-Unix the code to only use x-an-mp on unices * do your own checks, e.g. (if MP::*MULTIPROCESSING* ; I coulda sworn this used to be public in 4.2 (comm:start-up-server ...) (comm:start-up-server-and-mp ...)) I'm not sure which is the best solution because I'm not sure why it currently uses comm:start-up-server-and-mp. This may or may not be related to another problem I'm having, which is all works fine when I manually evaluate (load "/home/ap/Lisp/slime/swank-loader.lisp") (swank::create-swank-server 4005) in a running lispworks and connecting via slime-connect, but this fails when the above forms are in my .lispworks file and I try C-u M-x slime [lisp?] my-lispworks-image The process shows up: XOS 5 > (mp:ps) #<MP:PROCESS Name "Swank Request Processor" Priority 3 State "Waiting for connection"> #<MP:PROCESS Name "Top level loop" Priority 0 State "Running"> #<MP:PROCESS Name "The idle process" Priority -8388608 State "Running"> NIL XOS 6 > But noone is actually listening on the socket. (as can be verified with telnet). Clues? Thanks,

3 4

[slime-devel] some conveniences for slime-repl
by Andreas Fuchs 10 Dec '03

10 Dec '03

hi all, I have found a few things that I miss in the SLIME repl: * a key binding to erase all input before point; much like comint's C-c C-u. alternatively, * a key binding to jump to just after the input marker; and * slightly different RET handling. currently, SLIME will insert the RET at (point) and then send the line if I am inside a complete sexp. this can lead to badness like: ZY> (un trace my-function) or just ugly cmdlines in the history when I use M-( I think inserting the \n only when (point) is at the end of the finished sexp would work better. Thanks, -- Andreas Fuchs, <asf(a)boikor.net>, asf(a)jabber.at, antifuchs

4 5

[slime-devel] Tentative small refactoring stuff
by Daniel Barlow 10 Dec '03

10 Dec '03

I'd like to add SLIME support for compiling entire (multi-file) systems, and am finding that each backend handles compilation slightly differently. I believe I need to add a new function compile-system-for-emacs that looks not unlike compile-{file,string}-for-emacs : set up compiler condition hook stuff, then call (asdf:operate 'asdf:oop systemname) You'd think I would be able to do that once instead of n times for n backends, but apparently not. I don't have n lisps installed nor the will to test them all either. So, I have (committage imminent) - created a new GF call-with-compilation-hooks and a macro with-compilation-hooks that expands into a call to it (It's implemented for SBCL and CMUCL, because they already have and use with-compilation-hooks) - created a compile-system-for-emacs for SBCL which uses with-compilation-hooks. I believe the contents are portable CL and will work on any backend, if the maintainer of that backend has implemented call-with-compilation-hooks. If anyone is found copying compile-system-for-emacs into their own backend file and hacking it, they'd better have a good reason. Let's not needlessly duplicate entities, hey? -dan -- http://web.metacircles.com/ - Open Source software development and support

2 1

[slime-devel] Some patches
by Sven Van Caekenberghe 10 Dec '03

10 Dec '03

I think someone forgot the loop around the let in create-swank-server in swank-openmcl.lisp - without the loop you can only connect once and the socket is lost Also for all servers, I would suggest an option to restrict connections to localhost only, and even make that the default - in openmcl, the test is simply (= (remote-host client-socket) (local-host client-socket)) right after the accept With these patches, you could leave a swank-server running all the time, and forward the port over ssh to access remotely running images, with all the comfort of slime! BTW, I think slime is very cool! Sven -- Sven Van Caekenberghe - mailto:sven@beta9.be Beta Nine - software engineering - http://www.beta9.be .Mac - svc(a)mac.com - http://homepage.mac.com/svc "Lisp isn't a language, it's a building material." - Alan Kay

2 1

[slime-devel] Rape/Steal/Pillage
by Alain.Picard＠memetrics.com 10 Dec '03

10 Dec '03

Dear Slimers, Well, it hadda happen, I started using SLIME in earnest, as a full-blown replacement to ILISP. Of course, this immediately lead to annoyance at the many missing features. Now, it turns out that many of these missing features are really just editing conveniences, and don't have a lot (anything?) to do with emacs<-->lisp communication, so I thought I'd ask: what is the avowed policy of this project with regards to borrowing¹ the missing features of ilisp? I'm thinking in particular of all the buffer-movement, narrow-to-X and reindentation commands, for a start. It seems silly to rewrite all this, given that it all works. All that would be needed is some poor schmuck² to import only the strict minimum, track down all the dependencies, and hook in the keybindings, taking care not to trounce too much exising ones without reason or debate. Opinions? ¹ Ok, maybe the subject line is more accurate than this euphemism. ² Yeah, I know. Probably _me_. p.s. Is it "slimers" or "slimeys"? :-)

7 13

[slime-devel] eval compile dwim
by jan 10 Dec '03

10 Dec '03

I don't like jumping back and forth between a buffer and the repl when testing small snippets of code, I prefer to temporarily insert them into the buffer and eval them in place. This creates problems in slime because some features (who-calls, compiler notes) only work properly with compilation, which means I have to use both eval and compile for this strategy to be effective. This leads to errors when done manually so I defined this function: (defun slime-eval/compile-defun-dwim (&optional arg) "Call the computation command you want (Do What I Mean). Look at defun and determine whether to call `slime-eval-defun' or `slime-compile-defun'. A prefix of `-' forces evaluation, any other prefix forces compilation." (interactive "P") (case arg ;; prefix is `-', evaluate defun ('- (slime-eval-defun)) ;; no prefix, automatically determine action ('nil (if (string-match "^(def" (slime-defun-at-point)) (slime-compile-defun) (slime-eval-defun))) ;; prefix is not `-', compile defun (otherwise (slime-compile-defun)))) I use this command all the time so I bind it on M-j. This is bound to indent-new-comment-line by default but it isn't missed since it is also bound on C-M-j. Having the command bound on a Meta key is useful because I don't have to release Meta to M-n or M-p if I get compiler notes. I thought this might be helpful, but maybe I'm missing something. How do other people typically interact with lisp? What are the most commonly used commands? -- jan

3 5

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

slime-devel