March 2008 - slime-devel - mailman3.common-lisp.net

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 17 Mar '08

17 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1320 slime/ChangeLog:1.1321 --- slime/ChangeLog:1.1320 Fri Mar 14 09:05:04 2008 +++ slime/ChangeLog Sun Mar 16 05:46:33 2008 @@ -1,3 +1,16 @@ +2008-03-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (load-file-set-package): Removed; the function was + only used in `slime-load-file-set-package' which invokes + `slime-repl-set-package' which set the package for a second time. + + * slime.lisp (slime-load-file-set-package): Don't call + SWANK:LOAD-FILE-SET-PACKAGE, but merely call SWANK:LOAD-FILE, + then invoke `slime-repl-set-package' which will set the package. + + (slime-pretty-find-buffer-package): Removed. (Nowhere used.) + (slime-set-package): Ditto. + 2008-03-14 Tobias C. Rittweiler <tcr(a)freebits.de> * swank.lisp (classify-symbol, symbol-classification->string): Add

1 0

[slime-devel] Two trivial patches for cmucl and ccl
by Hans Hübner 16 Mar '08

16 Mar '08

Hi, the attached patch works for me to get Slime CVS running with CCL 1.2 (pre-release trunk version) and cmucl (19d). The cmucl patch will need revision with 19E, but for the moment, it helps. -Hans

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 15 Mar '08

15 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1319 slime/ChangeLog:1.1320 --- slime/ChangeLog:1.1319 Thu Mar 13 18:45:35 2008 +++ slime/ChangeLog Fri Mar 14 09:05:04 2008 @@ -1,3 +1,9 @@ +2008-03-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (classify-symbol, symbol-classification->string): Add + classification of symbols denoting type specifier, and denoting + constants. + 2008-03-13 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-eval-macroexpand): Indent expansion. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.97 slime/contrib/ChangeLog:1.100 --- slime/contrib/ChangeLog:1.97 Sat Mar 8 03:42:23 2008 +++ slime/contrib/ChangeLog Fri Mar 14 09:39:36 2008 @@ -1,3 +1,29 @@ +2008-03-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-fancy-inspector.lisp (add-slots-for-inspector): Remove + IGNORE declaration of non-existing argument. + +2008-03-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-fuzzy.lisp (fuzzy-convert-matching-for-emacs): Return a + string representation of the classifications rather than the + classifications themselves. (Notice this propagates up to (and + consequently changes the return value of) the RPC function + FUZZY-COMPLETIONS. (incompatible api change.) + + Rationale: The number of supported classification can be changed + without having to adapt its display at the client. + + * slime-fuzzy.lisp (slime-fuzzy-insert-completion-choice): + (slime-fuzzy-fill-completions-buffer): Adapted to API change. + +2008-03-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-fancy-inspector.lisp (make-symbols-listing :classification): + Add support for typespec and constant classification; don't + silently ignore symbols that can't be usefully classified, but + group them under "MISC". + 2008-03-08 Helmut Eller <heller(a)common-lisp.net> Don't blindly override the inspect method for functions.

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 14 Mar '08

14 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1312 slime/ChangeLog:1.1319 --- slime/ChangeLog:1.1312 Sat Mar 8 03:42:21 2008 +++ slime/ChangeLog Thu Mar 13 18:45:35 2008 @@ -1,3 +1,43 @@ +2008-03-13 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-eval-macroexpand): Indent expansion. + +2008-03-13 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-edit-definition-cont): If no definition could + be found, print also the package name in the error message where + the definition was tried to be found in. + +2008-03-13 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-region-for-defun-function): Deleted. + (slime-region-for-defun-at-point): Use beginning-of-defun + and not beginning-of-sexp. + (slime-flash-region): New function. + (slime-compile-region): Use it. + +2008-03-13 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-xref-group): Renamed from slime-location-to-string. + Handle source-form locations. + +2008-03-12 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-find-definitions-function): Renamed from + slime-edit-definition-fallback-function. + (slime-find-definitions): Use it. + (slime-find-tag-if-tags-table-visited): Deleted. + +2008-03-12 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-inspector-operate-on-point): Signal an error if + there is no object to operate on. + +2008-03-08 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (sldb-toggle-details): Inhibit point-motion-hooks. + This is a workaround for problems with the --more-- field. + 2008-03-07 Helmut Eller <heller(a)common-lisp.net> Be GC friendlier when parsing net packets.

2 1

[slime-devel] inexistent function slime-restore-window-configuration
by Stelian Ionescu 13 Mar '08

13 Mar '08

the keybinding C-c M-0 points to a nonexistent function: slime-restore-window-configuration -- Stelian Ionescu a.k.a. fe[nl]ix Quidquid latine dictum sit, altum videtur.

1 0

[slime-devel] slime-compiler-macroexpand problem
by Stelian Ionescu 12 Mar '08

12 Mar '08

slime-compiler-macroexpand and slime-compiler-macroexpand, when invoked in a macroexpansion buffer, replace the entire buffer instead of replacing only the expanded form -- Stelian Ionescu a.k.a. fe[nl]ix Quidquid latine dictum sit, altum videtur.

1 0

[slime-devel] (asdf:oos 'asdf:load-op :swank)(create-server) doesn't load contribs.
by Numeromancer 09 Mar '08

09 Mar '08

Hi, Just a note, and an implied question, I guess. I start a web server using sbcl, and start a swank server in it. I just started using a later version of slime from cvs (SLIME 2008-02-28). Slime no longer loads the contribs, as you know, which caused the server to generate an error when connecting to it remotely from an emacs which is using the contribs. .sbclrc used to say: (asdf:oos 'asdf:load-op :swank) (swank:create-server :port 4006 :dont-close t) And it worked. I now must say: (asdf:oos 'asdf:load-op :swank) (swank-loader:init :load-contribs t) (swank:create-server :port 4006 :dont-close t) I thought I would mention this here, as I could find nothing in the docs, or anything in the changelog, news, or readme files which mentioned this. If anyone can say "You should be doing it this way instead...", that'd be good. Could swank.asd have something like this? (just off the top of my head, untested): (defclass swank-with-contribs-loader-file (swank-loader-file) ()) (defmethod asdf:operation-done-p ((o asdf:compile-op) (f swank-loader-file)) t) (defmethod asdf:perform ((o asdf:load-op) (f swank-with-contribs-loader-file)) (load (asdf::component-pathname f)) (funcall (read-from-string "swank-loader::init") :reload (asdf::operation-forced o) :delete (asdf::operation-forced o) :load-contribs t :setup nil)) (asdf:defsystem :swank-with-contribs :default-component-class swank-with-contribs-loader-file :components ((:file "swank-loader"))) I know there is a better way to do this. Tim S

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 09 Mar '08

09 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1311 slime/ChangeLog:1.1312 --- slime/ChangeLog:1.1311 Tue Mar 4 10:48:40 2008 +++ slime/ChangeLog Sat Mar 8 03:42:21 2008 @@ -1,3 +1,10 @@ +2008-03-07 Helmut Eller <heller(a)common-lisp.net> + + Be GC friendlier when parsing net packets. + + * slime.el (slime-net-read): Instead of consing a fresh string, + use narrow-to-region and read the packet out of the buffer. + 2008-03-04 Helmut Eller <heller(a)common-lisp.net> * test.sh: Updated for Emacs 23. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.96 slime/contrib/ChangeLog:1.97 --- slime/contrib/ChangeLog:1.96 Sun Mar 2 17:41:45 2008 +++ slime/contrib/ChangeLog Sat Mar 8 03:42:23 2008 @@ -1,3 +1,11 @@ +2008-03-08 Helmut Eller <heller(a)common-lisp.net> + + Don't blindly override the inspect method for functions. + + * swank-fancy-inspector.lisp (emacs-inspect function): Define this + method only if the backend hasn't defined one. + (inspect-function): New function. + 2008-03-02 Tobias C. Rittweiler <tcr(a)freebits.de> M-. now works on presentations.

1 0

Re: [slime-devel] Mercurial
by Helmut Eller 08 Mar '08

08 Mar '08

Tobias C. Rittweiler" wrote: > Hi Helmut, > > I am currently in vacations, and don't have access to > my machine, and as I don't think that I'm subscribed > to the mailinglist (I usually use the gmane interface), > I write this mail to you directly (feel free to forward.) > > you wrote: > > > So lets make a poll. Those people with cvs commit rights can > > vote. Say > > which system you prefer; if you don't want to switch, vote for > > CVS. > > You can cast your ballot until 2008-March-07 00:00 UTC. > > I'm reluctant to anything that there isn't an Emacs interface > for that doesn't suck. Does one exist for Mercurial, and > do you have experience using it? In Emacs 23 (CVS Emacs) the vc interface supports hg, git, bzr and maybe others. But that is rather new and has still some rough edges. E.g. the vc-annotate for hg only shows the the revision number and the date but not the author who wrote the line. Another annoyance is that they synchronously call subprocesses the find-file-hook, apparently to display the version number in the mode-line. This can be rather slow (not so much for SLIME, but for big repos). To create a changeset one must mark the relevant files in the vc-dired buffer and the marked files will then be committed together (not individually as in older versions). The diff command seem to work but it doesn't honor `diff-switches' properly. I think the log command works only on one file if called in the buffer or on multiple files if called in the vc-dired buffer. There are no Emacs equivalents for the status, pull, and push commands. For git there exist some other Emacs packages, one is pcl-cvs-a-like. I wasn't impressed by those. The bzr-mode.el shipped with Bazaar was even written by the illustrious Luke Gorrie. Like his darcs-mode.el it's simple but easily hackable. Currently, I use hg primarily from the command line. Helmut.

2 1

[slime-devel] SoC 2008: any projects you'd like to get funded?
by Heow Eide-Goodman 05 Mar '08

05 Mar '08

Guys, Once again, LispNYC is getting ready for Google's Summer of Code. Since we all use and love SLIME, getting a project funded would be great. If you have any project ideas, drop us a line here ( http://lispnyc.org/soc-idea.clp ), we're also looking for a few good mentors. :-) Cheers, - Heow -- ;;-------------------------------------------------------------------------- (defun reply-to-this-email (addresses) (apply 'concatenate 'string (mapcar #'reverse (nth (random (length addresses)) addresses)))) (reply-to-this-email '(("woeH" "@" "cnIskeeGahplA" "." "moc") ("woeH" "@" "CYNpsiL" "." "gro")))

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 05 Mar '08

05 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1309 slime/ChangeLog:1.1311 --- slime/ChangeLog:1.1309 Sun Mar 2 10:10:34 2008 +++ slime/ChangeLog Tue Mar 4 10:48:40 2008 @@ -1,3 +1,12 @@ +2008-03-04 Helmut Eller <heller(a)common-lisp.net> + + * test.sh: Updated for Emacs 23. + +2008-03-04 Andreas Fuchs <asf(a)boinkor.net> + + * swank.asd (asdf:output-files, asdf:perform): Make compile-op on + swank-loader-file a noop. + 2008-03-02 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-edit-definition-hooks): This variable can be

1 0

[slime-devel] Fancy inspector wishlist items
by Nikodemus Siivola 04 Mar '08

04 Mar '08

Loading the fancy inspector overrides the SBCL specific EMACS-INSPECT method for functions. It would be nice if it didn't, but rather co-operated with it. At some point in the past C-I worked on presentations, which it doesn't anymore. This was very useful, and it would be nice to have it back in the fancy inspector. While I don't miss the full DWIM magic of C-I, there is one thing I would like there: if the point is at the start of a list starting with FOO, insert #'FOO instead of FOO in the minibuffer. Cheers, -- Nikodemus

2 2

[slime-devel] SBCL-specific function inspection
by Ariel Badichi 04 Mar '08

04 Mar '08

The following patch is a work-around (excuse the pun) to have the fancy-inspector use the SBCL-specific method for function inspection. Ariel diff -u -r1.193 swank-sbcl.lisp --- swank-sbcl.lisp 28 Feb 2008 19:44:14 -0000 1.193 +++ swank-sbcl.lisp 3 Mar 2008 04:09:39 -0000 @@ -1041,7 +1041,7 @@ (loop for value in parts for i from 0 append (label-value-line i value)))))))) -(defmethod emacs-inspect ((o function)) +(defmethod emacs-inspect :around ((o function)) (let ((header (sb-kernel:widetag-of o))) (cond ((= header sb-vm:simple-fun-header-widetag) (label-value-line*

3 2

[slime-devel] Re: SLIME and new windows
by Rommel Martinez 04 Mar '08

04 Mar '08

Sorry for bothering you guys. I found a solution, but I don't think it's the most elegant one. What I did was I changed the call to `pop-to-buffer' to `switch-to-buffer' in `slime-hide-inferior-lisp-buffer' like so: ---------------------------------------------------------------------- --- slime.el.orig 2008-03-04 12:42:05.000000000 +0800 +++ slime.el 2008-03-04 12:42:41.000000000 +0800 @@ -1528,7 +1528,7 @@ (window (set-window-buffer window repl-buffer)) (t - (pop-to-buffer repl-buffer) + (switch-to-buffer repl-buffer) (goto-char (point-max)))))) ---------------------------------------------------------------------- Is there a better way to do this? -- Rommel M. Martinez <ebzzry(a)gmail.com>

1 0

[slime-devel] SLIME and new windows
by Rommel Martinez 04 Mar '08

04 Mar '08

How do I make SLIME *not* create a new window when it connects to a Swank server via `slime-connect'? Specifically, how do I make SLIME switch to the REPL buffer (output buffer) using the current window? Thanks in advance. -- Rommel M. Martinez <ebzzry(a)gmail.com>

1 0

[slime-devel] slime-macroexpand issue/wishlist item
by Nikodemus Siivola 03 Mar '08

03 Mar '08

If a macroexpansion communicates with other parts of the system using eg. structures or instances, expanding nested macros doesn't work in Slime: ...assuming you have CFFI installed and loaded... (cffi:defcallback foo :void ((data :string)) (error data)) C-RET expands it to (PROGN (CFFI-SYS:%DEFCALLBACK FOO :VOID (DATA) (:POINTER) (LET ((DATA (CFFI:TRANSLATE-FROM-FOREIGN DATA #<CFFI::FOREIGN-STRING-TYPE :UTF-8>))) (BLOCK FOO (ERROR DATA))) :CALLING-CONVENTION :CDECL) 'FOO) where the %DEFCALLBACK is also a macro. Trying to re-expand that using C-RET breaks when it tries to READ #<. It would be nice if (a) the second expansion worked (2) the #<...> objects in the buffer were presentations. Cheers, -- Nikodemus

1 0

[slime-devel] Debugging SBCL startup issue.
by David Brown 03 Mar '08

03 Mar '08

I'm running SBCL 1.0.14 built on gentoo, using the recent CVS of Slime. When I start slime, it begins in the *inferior-lisp* buffer, loads Swank, and then sits at a CL-USER[2]: prompt. The minibuffer says: "Connecting to Swank on port nnnn.." This inferior lisp buffer accepts commands. I've discovered that by generating an error in the inferior lisp, SBCL enters its debugger, which then prints "Adding fd handler: 6" and then slime starts up fine. Any ideas on how to debug this, or what the problem might be? It seems to work once I generate the error. I'm going to try rebuilding SBCL with threading support enabled to see if this helps. Thanks, David Brown

1 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 03 Mar '08

03 Mar '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1308 slime/ChangeLog:1.1309 --- slime/ChangeLog:1.1308 Thu Feb 28 14:46:13 2008 +++ slime/ChangeLog Sun Mar 2 10:10:34 2008 @@ -1,3 +1,10 @@ +2008-03-02 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-edit-definition-hooks): This variable can be + used to hook into the M-. machinery. + (slime-edit-definition): Run above hooks until one succeeds. By + default, try to find a definition for the symbol at point. + 2008-02-28 Tobias C. Rittweiler <tcr(a)freebits.de> * swank.lisp (find-definition-for-thing): New DEFSLIMEFUN. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.94 slime/contrib/ChangeLog:1.96 --- slime/contrib/ChangeLog:1.94 Thu Feb 21 15:49:31 2008 +++ slime/contrib/ChangeLog Sun Mar 2 17:41:45 2008 @@ -1,3 +1,22 @@ +2008-03-02 Tobias C. Rittweiler <tcr(a)freebits.de> + + M-. now works on presentations. + + Additionally, a Find Definition entry is presented in the menu + appearing on right clicking on a presentation. + + * slime-presentations.lisp (slime-M-.-presentation): New function. + (slime-M-.-presentation-at-mouse): New function. + (slime-M-.-presentation-at-point): New function. + (slime-maybe-M-.-presentation-at-point): New function. + (slime-menu-choices-for-presentation): New entry "Find Definition". + (slime-presentation-easy-menu): New entry "Find Definition". + (slime-presentations-init): Hook into `slime-edit-definition-hooks'. + +2008-03-02 Helmut Eller <heller(a)common-lisp.net> + + * swank-mit-scheme.scm: New file. + 2008-02-21 Tobias C. Rittweiler <tcr(a)freebits.de> Having the `slime-presentations' contrib enabled, (princ 10)

1 0

[slime-devel] #+foo (toplevel-form ...), C-c C-c
by Tobias C. Rittweiler 02 Mar '08

02 Mar '08

Hi, on something like #+foo (defun foo (...) ...) | or #+foo (defun foo (...) ...) | Do you think C-c C-c should take the reader conditional into account? As `slime-region-for-defun-at-point' happens to be implemented right now, it isn't. -T.

4 3