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

[slime-devel] Daily ChangeLog diff
by Helmut Eller 08 Dec '08

08 Dec '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1592 slime/ChangeLog:1.1593 --- slime/ChangeLog:1.1592 Fri Dec 5 20:01:54 2008 +++ slime/ChangeLog Sun Dec 7 12:25:38 2008 @@ -1,3 +1,9 @@ +2008-12-07 Nikodemus Siivola <nikodemus(a)random-state.net> + + * slime.el (slime-qualify-cl-symbol-name): Clean up the package + name using `slime-pretty-package-name', so that packages named with + strings are not left with the extra doublequotes. + 2008-12-05 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-macroexpansion-minor-mode-map): Bind keybindings

1 0

[slime-devel] Problems with slime using sbcl on winXP (emacsW32)
by Matthew D. Swank 06 Dec '08

06 Dec '08

Starting in late September, whenever I try to exit out of the debugger the repl prompt doesn't come back. Example: *slime-repl sbcl* CL-USER> o *sldb sbcl/0* The variable O is unbound. [Condition of type UNBOUND-VARIABLE] Restarts: 0: [RETRY] Retry SLIME REPL evaluation request. 1: [ABORT] Return to SLIME's top level. <-----------pick this one 2: [CLOSE-CONNECTION] Close SLIME connection 3: [ABORT] Exit debugger, returning to top level. The repl buffer comes back with no prompt, and all further requests are pipelined. This does not happen in clisp. Some other info: *slime-events* (:emacs-rex (swank:listener-eval "o\n") "COMMON-LISP-USER" :repl-thread 14) (:debug 0 1 ("The variable O is unbound." " [Condition of type UNBOUND-VARIABLE]" nil) (("RETRY" "Retry SLIME REPL evaluation request.") ("ABORT" "Return to SLIME's top level.") ("CLOSE-CONNECTION" "Close SLIME connection") ("ABORT" "Exit debugger, returning to top level.")) ((0 "(SB-INT:SIMPLE-EVAL-IN-LEXENV O #<NULL-LEXENV>)") (1 "(SWANK::EVAL-REGION \"o\\n\")") (2 "((LAMBDA ()))") (3 "(SWANK::TRACK-PACKAGE #<CLOSURE (LAMBDA #) {25B5C3E5}>)") (4 "(SWANK::CALL-WITH-RETRY-RESTART \"Retry SLIME REPL evaluation request.\" #<CLOSURE (LAMBDA #) {25B5C375}>)") (5 "(SWANK::CALL-WITH-BUFFER-SYNTAX NIL #<CLOSURE (LAMBDA #) {25B5C35D}>)") (6 "(SWANK::REPL-EVAL \"o\\n\")") (7 "(SB-INT:SIMPLE-EVAL-IN-LEXENV (SWANK:LISTENER-EVAL \"o\\n\") #<NULL-LEXENV>)") (8 "(SWANK::EVAL-FOR-EMACS (SWANK:LISTENER-EVAL \"o\\n\") \"COMMON-LISP-USER\" 14)") (9 "(SWANK::PROCESS-REQUESTS NIL)") (10 "((LAMBDA ()))") (11 "((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN)) #<FUNCTION SWANK:SWANK-DEBUGGER-HOOK> #<CLOSURE (LAMBDA #) {241035BD}>)") (12 "(SWANK::CALL-WITH-REDIRECTED-IO #<SWANK::CONNECTION {2401F269}> #<CLOSURE (LAMBDA #) {241035CD}>)") (13 "(SWANK::CALL-WITH-CONNECTION #<SWANK::CONNECTION {2401F269}> #<CLOSURE (LAMBDA #) {241035BD}>)") (14 "(SWANK::HANDLE-REQUESTS #<SWANK::CONNECTION {2401F269}> NIL)") (15 "((LAMBDA ()))") (16 "((LAMBDA (SWANK-BACKEND::HANDLER FUNCTION)) #<FUNCTION (LAMBDA #) {23DC867D}> #<CLOSURE (LAMBDA #) {24103555}>)") (17 "(SWANK::SIMPLE-SERVE-REQUESTS #<SWANK::CONNECTION {2401F269}>)") (18 "(SWANK::SERVE-CONNECTION #<SB-BSD-SOCKETS:INET-SOCKET descriptor -1 {23FB59F9}> NIL NIL :ISO-8859-1)") (19 "(SWANK::SETUP-SERVER 0 #<CLOSURE (LAMBDA #) {23F4527D}> NIL NIL :ISO-8859-1)")) (14)) (:debug-activate 0 1 nil) (:emacs-rex (swank:invoke-nth-restart-for-emacs 1 1) nil 0 15) (:return (:abort) 15) (:debug-return 0 1 nil) These are the current contribs I'm using: (require 'slime-fancy) (require 'slime-asdf) Just using straight socket for communication: swank:*communication-style* => nil Matt -- "You do not really understand something unless you can explain it to your grandmother." -- Albert Einstein.

3 7

[slime-devel] Prefix-Arg of slime-toggle-trace-fdefinition
by Tobias C. Rittweiler 06 Dec '08

06 Dec '08

Hi, is there a particular reason why C-u C-c C-t is not the default behaviour of C-c C-t? Perhaps it's better if C-c C-t always tries to look at the context, but then offers that choice to the user in a way that is correctable by the user. Say it prompts for (flet foo) for a local function. -T.

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 06 Dec '08

06 Dec '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1590 slime/ChangeLog:1.1592 --- slime/ChangeLog:1.1590 Tue Dec 2 22:00:41 2008 +++ slime/ChangeLog Fri Dec 5 20:01:54 2008 @@ -1,3 +1,17 @@ +2008-12-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-macroexpansion-minor-mode-map): Bind keybindings + of `undo' to `slime-macroexpansion-undo'. + (slime-show-macroexpansion): Renamed to + `slime-initialize-macroexpansion-buffer' + (slime-initialize-macroexpansion-buffer): Make sure that the user + can't undo past the initial insertion. + (slime-macroexpand-undo): New function. Tries to undo-only. + +2008-12-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-trace-query): SPEC argument can be a symbol. + 2008-12-02 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-compute-modeline-connection-state): Fix

1 0

[slime-devel] [PATCH] more reader cleanups
by Tobias C. Rittweiler 05 Dec '08

05 Dec '08

* make-d-macro-char.diff Currently, invoking MAKE-DISPATCH-MACRO-CHARACTER on a character that's already been made a dispatch macro character results in an error. All other implementations I tested, do not signal an error, but just override a possibly existing dispatch table. I think that's sensible, as the the CLHS entry for MAKE-D-M-CHAR says Exceptional Situations are None. * get-dtable-pair.diff The implementation for retrieving an entry from the dispatch tables is currently duplicated on all call sites. This patch introduces an explicit getter function. Also, in SET-DISPATCH-MACRO-CHARACTER the function argument was COERCED to a function (thus allowing literal lambda expressions), use %COERCE-TO-CALLABLE instead. * inline-assert-not-std-rt.diff Inline ASSERT-NOT-STANDARD-READTABLE, and MAKE-DISPATCH-CHAR-TABLE. * char-attr-table-type.diff Introduce an explicit type for character attributes (CHAR-ATTR), and rename the type ATTRIBUTE-TABLE to CHAR-ATTR-TABLE. * ref-cat-entry.diff Fix more FIXMES and code duplication by introducing a low-level getter REF-CAT-TABLE which is then used at the appropriate places. -T.

1 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 03 Dec '08

03 Dec '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1587 slime/ChangeLog:1.1590 --- slime/ChangeLog:1.1587 Sun Nov 23 22:11:36 2008 +++ slime/ChangeLog Tue Dec 2 22:00:41 2008 @@ -1,3 +1,23 @@ +2008-12-02 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-compute-modeline-connection-state): Fix + computation of debugged requests. + +2008-11-30 Gábor Melis <mega(a)retes.hu> + + * slime.el (slime-compute-modeline-connection-state): Print the + number of debugged requests if non-zero. + (slime-connection-state-as-string): Removed unused branches. + (slime-compute-connection-state): Removed :debugged branch making + most likely return :connected instead. + +2008-12-02 Tobias C. Rittweiler <tcr(a)freebits.de> + + Modeline wouldn't display {debugged..} after `slime-interrupt'. + + * slime.el (slime-debugged-connection-p): Can't rely on + `sldb-debugged-continuations' to be non-null in every SLDB buffer. + 2008-11-23 Helmut Eller <heller(a)common-lisp.net> `q' in temp buffers should only delete the window if it was newly

1 0

[slime-devel] [BUG] quitting from sldb after error in thread
by Tobias C. Rittweiler 03 Dec '08

03 Dec '08

On SBCL, evaluating (swank::spawn #'(lambda () (error "FOO"))) at the REPL, then pressing `q' in SLDB results in an invocation of (FIND-RESTART NIL) which is an error in itself. -T.

1 0

[slime-devel] Reincarnate "eval..."
by Gábor Melis 02 Dec '08

02 Dec '08

Running cvs HEAD I now realize that I grew to rely on checking the modeline of lisp buffers to see whether there are outstanding requests. Often, I start evalutations that take minutes or hours to finish and it's useful to have this at a glance without having to worry about missing when the return value is printed or having to check the running threads. The attached patch does not reinstate "eval...", instead it displays the number of outstanding requests like "CON:sbcl{3}" if it's greater than zero. Currently there is a loop through buffer-list, it that's a problem maybe there is a better way to arrange for the modeline to be updated using a global mode-line-format variable as a trigger to signal the change in connection-state that is to be recomputed in each slime buffer. The patch does not solve the minor problems caused by the use of run-with-idle-timer, namely that the modeline may not be updated if the status of the connection changes in the background while emacs is idle. Cheers, Gábor

3 3

[slime-devel] slime-undefmethod contrib
by Madhu 02 Dec '08

02 Dec '08

Attached is my first contrib submission. M-x slime-undefine-function does not help when you wish to remove a specific method. This contrib adds a new function slime-undefmethod. Invoking M-x slime-undefmethod with the point inside a defmethod definition in your source will find and remove the specific method. I've posted the code before on usenet, but felt it may be more useful if packaged up with slime. Comments welcome. -- Madhu

2 1

[slime-devel] Call stack when using Lisp debugger from sldb
by Raymond Toy 01 Dec '08

01 Dec '08

Sometimes when I'm in sldb, I want to use the actual Lisp debugger because sometimes it more convenient to have a real repl. Pressing B works nicely. Previously, however, I would be in the debugger, and the backtrace would be exactly what I expected, with the top of stack being exactly where the error occurred. Now, however, the actual error is some 10-20 frames away because there are lots of slime things in between. This is kind of nice too because now I can return to sldb, which I couldn't before. But this also means I now have to go hunt around for the real offending frame. Is there a way to the old behavior? Ray

3 5