January 2004 - slime-devel - mailman3.common-lisp.net

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 20 Jan '04

20 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.223 slime/ChangeLog:1.226 --- slime/ChangeLog:1.223 Sun Jan 18 16:54:06 2004 +++ slime/ChangeLog Mon Jan 19 16:49:49 2004 @@ -1,3 +1,61 @@ +2004-01-19 Alan Ruttenberg <alanralanr(a)comcast.net> + *swank-openmcl.lisp in frame-catch-tags, ppc32::catch-frame.catch-tag-cell -> 0, + ppc32::catch-frame.csp-cell -> 3. FIXME when this code is more stable in openMCL. + +2004-01-19 Michael Weber <michaelw(a)foldr.org> + + * slime.el (slime-close-all-sexp): New command to close all + unmatched parens in the current defun. Bound to `C-c C-]'. With + prefix argument, only operate in the region (for closing + subforms). + +2004-01-19 Luke Gorrie <luke(a)bluetail.com> + + * swank-openmcl.lisp (thread-id, thread-name): Fixed silly bugs + (thanks Marco Baringer). + + * swank-loader.lisp: Call (swank:warn-unimplemented-interfaces). + + * swank.lisp (ed-in-emacs): New command with the same interface + as CL:ED. + + * swank-cmucl.lisp, swank-sbcl.lisp, swank-lispworks.lisp, + swank-openmcl.lisp, swank-allegro.lisp, swank-clisp.lisp: Updated + to use `defimplementation'. + + * swank-backend.lisp (definterface, defimplementation): New macros + as sugar around defgeneric/defmethod. This supports conveniently + supplying a default (on NO-APPLICABLE-METHOD). Because the + underly mechanism is still generic functions this doesn't break + code that isn't updated. + (warn-unimplemented-interfaces): Print a list of backend functions + that are not implemented. + (xref and list-callers): Defined interfaces for these functions. + (describe-definition): New function that takes over from the many + other describe-* functions called from apropos listing. Takes the + type of definition (as returned by describe-symbol-for-emacs) as + an argument. + + * slime.el (sldb-enable-styled-backtrace): This is now true by + default. + (slime-keys): Bound `slime-inspect' to `C-c I'. + (slime): `M-x slime' now offers to keep existing connections + alive (else disconnect them). If you disconnect them, the new + connection gets to reuse the existing REPL. + (slime-connection): Error if the connection is closed. + (slime-handle-oob): New message (:ED WHAT) for `slime-ed'. + (slime-display-output-buffer): Don't pop up the REPL if it is + already visible in any frame. + (slime-find-asd): Handle case where (buffer-file-name) is nil. + (slime-ed): Elisp backend for (CL:ED WHAT). + (slime-apropos): Add a summary line to apropos listings. + (slime-print-apropos): Replaced `action' property (name of lisp + describe function) with `type' (argument to pass to unified + swank:describe-definition function). + (slime-apropos-package): New command on `C-c P'. Presents apropos + listing for all external (with prefix also internal) symbols in a + package. + 2004-01-18 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank-lispworks.lisp (sigint-handler): Bind a continue restart.

1 0

[slime-devel] close-all-sexp for SLIME
by Michael Weber 19 Jan '04

19 Jan '04

Hi, Attached is a function for SLIME which when run automatically balances all open sexps at point. It is similar to ILISP's CLOSE-ALL-LISP, which is bound to C-c ] or M-], I think, but smaller and easier to understand. It is a little smarter than C-<Ret> in slime-repl, handles comments, reformats trailing right parens into one line, and DWIM when a region is active. Strings are not yet handled because it can be quite costly, I think. Take it if you like it :) (License is same as SLIME.) Cheers, Michael -- /~\ ASCII ribbon | "If we knew what it was we were doing, it would not be \ / campaign | called research, would it?" X against | -- Albert Einstein / \ HTML mail |

3 6

[slime-devel] repl reused when in sldb
by Alan Ruttenberg 19 Jan '04

19 Jan '04

Suppose you have a repl, execute a form that causes the debugger to be invoked then do M-x slime. You now get a new process reusing the repl of the process in the debugger. I suggest the behaviour should be that a new repl is created, in the same way that a new one is created when the repl is busy. -Alan

3 2

[slime-devel] definterface/defimplementation
by Luke Gorrie 19 Jan '04

19 Jan '04

Ahoy, I added some sugar on top of defgeneric/defmethod: definterface defines the generic function. It keeps track of which ones have been defined, so they can be audited. It also can have body forms which become a NO-SUCH-METHOD default. defimplementation defines the backend method. It updates the list of implemented interfaces. swank-loader finishes by printing a list of all unimplemented interfaces. (A nod to Peter for suggesting this.) I updated cmucl/sbcl/lispworks/openmcl. However, I wasn't able to test OpenMCL because of another incompatibility: Read error between positions 12276 and 12727 in /Users/luke/slime/swank-openmcl.lisp. > Error in process listener(1): There is no package named "PPC32" . Possibly because it's running on a 64-bit machine (G5)? Updating the other backends should just be a matter of selective search-and-replace from defmethod to defimplementation. However, since it's the same thing under the hood, it still works without being updated. Cheers, Luke

2 2

[slime-devel] openmcl patches
by Marco Baringer 19 Jan '04

19 Jan '04

hi, there's been for a while a warning about an unused variable in swank-openmcl.lisp, this fixes that. then there's a thinko regarding the use of the elisp function assq as opposed to assoc. regarding the variable swank::*known-processes* it would seem that every new process gets put on that list twice, is this the expected behaviour? -- Marco Ring the bells that still can ring. Forget the perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

1 0

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 19 Jan '04

19 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.219 slime/ChangeLog:1.223 --- slime/ChangeLog:1.219 Sun Jan 18 02:59:49 2004 +++ slime/ChangeLog Sun Jan 18 16:54:06 2004 @@ -1,5 +1,32 @@ 2004-01-18 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + * swank-lispworks.lisp (sigint-handler): Bind a continue restart. + (make-dspec-location): Handle strings like pathnames. + Some multithreading support. + + * slime.el (compile-defun): Don't use keywords. The keyword + package is locked in Lispworks and causes the test-suite to hang. + (slime-eval-with-transcript): Fix bug triggered when 'package' is + a buffer local variable. Reported by Janis Dzerins. + (slime-batch-test): Wait until the connection is ready. + +2004-01-18 Alan Ruttenberg <alanralanr(a)comcast.net> + + * swank-openmcl: Implement frame-catch-tags. Added debugger functions + sldb-restart-frame, sldb-return-from-frame. Should probably be added to backend.lisp + but let's discuss first. Do other lisps support this? + + * slime.el sldb-restart-frame, sldb-return-from-frame + +2004-01-18 Wolfgang Jenkner <wjenkner(a)inode.at> + + * swank-clisp.lisp (call-without-interrupts): Evaluate + linux:SIGFOO at read time since the macro with-blocked-signals + expects a fixnum. + (compile-file-for-emacs): Comment fix. + +2004-01-18 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + * swank-sbcl.lisp (make-fn-streams): Deleted. Already defined in swank-gray.lisp.

1 0

[slime-devel] emacs complains about 'package' variable
by Janis Dzerins 18 Jan '04

18 Jan '04

Hello slime hackers. I'm having trouble with latest version of Slime -- I can't evaluate (as in C-M-x) or macroexpand (C-c C-m) forms. Emacs complains with message: 'Symbol's value as variable is void: package'. I've traced the problem to the function SLIME-EVAL-WITH-TRANSCRIPT. I debugged it with EDebug and here's what I found: when entering the function, arguments FORM, PACKAGE and FN have their intended values. I checked it with EDebug's eval facility. Then, after passing the SLIME-WITH-OUTPUT-END-MARK, I tried evaluating PACKAGE again, and it no longer had a value. This might have something to do with the WITH-LEXICAL-BINDINGS macro, but I'm not any good at debugging emacs code. So now I stick without macroexpand and use C-c C-c (which works!) instead of C-M-x. Any suggestions? -- Janis Dzerins If million people say a stupid thing, it's still a stupid thing.

2 3

[slime-devel] restart frame, return from frame
by Alan Ruttenberg 18 Jan '04

18 Jan '04

I have checked in changes that implement this for openmcl. sldb-return-from-frame: "reads an expression in the minibuffer and causes the function to return that value, evaluated in the context of the frame" sldb-restart-frame: "causes the frame to restart execution with the same arguments as it was called originally" You can use these functions in an sldb buffer with cursor on the frame you want to restart or return from. Does this functionality exist in other lisps? Also checked in implementation of frame-catch-tags for openmcl -Alan

3 2

[slime-devel] reloading swank-loader.lisp
by Alan Ruttenberg 18 Jan '04

18 Jan '04

Fails in openmcl, at least. SWANK:EVAL-IN-FRAME is defined as something other than a generic function. This poses a problem if you want to save an image into which slime is already loaded. If this is openmcl specific, let me know and I will patch. -Alan

2 1

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 18 Jan '04

18 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.213 slime/ChangeLog:1.219 --- slime/ChangeLog:1.213 Fri Jan 16 16:54:54 2004 +++ slime/ChangeLog Sun Jan 18 02:59:49 2004 @@ -1,3 +1,46 @@ +2004-01-18 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank-sbcl.lisp (make-fn-streams): Deleted. Already defined in + swank-gray.lisp. + + * swank.lisp (find-symbol-or-lose, format-arglist): New functions. + (without-interrupts): New macro. + (send-to-emacs): Use it. + + * swank-backend.lisp, swank-clisp.lisp, swank-lispworks.lisp, + swank-openmcl.lisp, swank-sbcl.lisp, swank-allegro.lisp: + (arglist-string): Refactor common code to swank.lisp. + (call-without-interrupts, getpid): Are now generic functions. + + * slime.el (arglist): Test slot readers and closures. + + * swank-cmucl.lisp (arglist-string): Use + pcl:generic-function-lambda-list for generic functions. Handle + closures. Print arglist in lower case. + (inspected-parts-of-value-cell): Was lost during the inspector + refactoring. + +2004-01-18 Wolfgang Jenkner <wjenkner(a)inode.at> + + * swank-clisp.lisp (compile-file-for-emacs, + split-compiler-note-line): Revert last change. + (handle-notification-condition): Don't signal the condition. + (*compiler-note-line-regexp*): Fix and rewrite it as extended + regexp. + + * slime.el (slime-changelog-date): Use file-truename of + byte-compile-current-file. + +2004-01-17 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * slime.el (slime-format-arglist): Add some sanity checks and + print zero argument functions nicer. Suggested by Ivan Boldyrev. + (slime-test-expect): Take test predicate as argument. + (arglist): Test generic functions. + + * swank-cmucl.lisp (arglist-string): Handle generic functions + better. Reported by Ivan Boldyrev. + 2004-01-16 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank-allegro.lisp: Multiprocessing support.

1 0

[slime-devel] Method arglist generates error
by Paolo Amoroso 18 Jan '04

18 Jan '04

Suppose I type the following at a *slime-repl<n>* buffer: ; CMU Common Lisp Port: (localhost 32816) Pid: 805 CL-USER> (defclass myclass () ((slot :initarg :slot :accessor slot))) #<STANDARD-CLASS MYCLASS {485E7535}> CL-USER> (slot If I place the cursor just after "(slot" and press space, I get the following SLIME error: :NO-LAMBDA-LIST fell through ETYPECASE expression. Wanted one of (NULL CONS STRING). [Condition of type CONDITIONS::CASE-FAILURE] Restarts: 0: [ABORT] Return to SLIME toplevel. 1: [ABORT] Return to Top-Level. Backtrace: 0: ("DEFMETHOD ARGLIST-STRING (T)" #<unused-arg> #<unused-arg> "slot") 1: ("DEFMETHOD ARGLIST-STRING (T)" 3 #<unused-arg> #<unused-arg> "slot")[:EXTERNAL] --more-- I use SLIME CVS sources updated today, CMUCL 18e and GNU Emacs 20.7.2 under Debian Woody. Paolo -- Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film

2 1

[slime-devel] Yet another patch
by Ivan Boldyrev 17 Jan '04

17 Jan '04

When generic function is defgeneric'ed, but not defmethod'ed, pcl::generic-function-pretty-arglist retruns nil on it. I doubt that using any other function instead of pcl::generic-function-pretty-arglist will help, because even common-lisp:describe doesn't show any useful information. Slime with this tiny patch handles this situation better: Before Slime described such methods as (test I) (where 'I' is (substring "NIL" 1 -1) :) Now it describes it as (test <no info>). May be, <no info> is not good string, feel free to replace it with better one :) -- Ivan Boldyrev Is 'morning' a gerund?

2 1

[slime-devel] [PATCH] swank:arglist-string retruns NIL for generic functions
by Ivan Boldyrev 17 Jan '04

17 Jan '04

My configuration is: Emacs 21.3.1 and CMU CL 18e under Linux. swank:arglist-string always returns NIL for generic functions. I have investigated the problem and here is a tiny patch: --- swank-cmucl.lisp.~1.50.~ 2004-01-16 00:54:45.000000000 +0600 +++ swank-cmucl.lisp 2004-01-17 04:18:09.000000000 +0600 @@ -781,7 +781,7 @@ (cond ((eval:interpreted-function-p fun) (eval:interpreted-function-arglist fun)) ((pcl::generic-function-p fun) - (pcl::gf-pretty-arglist fun)) + (pcl::generic-function-pretty-arglist fun)) (arglist arglist) ;; this should work both for ;; compiled-debug-function and for It seems pcl::gf-pretty-arglist always returns nil :( If you want to apply the patch and to mention my name in ChangeLog, use unmodified e-mail (with +nospam@). Or do not mention e-mail at all. -- Ivan Boldyrev Onions have layers. Unix has layers too.

2 1

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 17 Jan '04

17 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.210 slime/ChangeLog:1.213 --- slime/ChangeLog:1.210 Fri Jan 16 02:28:23 2004 +++ slime/ChangeLog Fri Jan 16 16:54:54 2004 @@ -1,10 +1,23 @@ 2004-01-16 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + * swank-allegro.lisp: Multiprocessing support. + + * swank-openmcl.lisp, swank-cmucl.lisp, swank-backend.lisp, + swank.lisp: Refactor inspector code. + * swank.lisp (changelog-date): Use *compile-file-truename* instead of *compile-file-pathname*. (with-I/O-lock, with-a-connection): The usual CLISP fixes. + (create-swank-server): Patch by Marco Baringer <mb(a)bese.it>. + Bring it back again. + (create-connection): Use return the dedicated output stream if + available. * slime.el: Numerous REPL related fixes. + (slime-update-state-name): Take state as argument. + (slime-repl-beginning-of-defun, slime-repl-end-of-defun): Fix + typos. + (sldb-insert-restarts): Remove duplicate definition. 2004-01-16 Luke Gorrie <luke(a)bluetail.com>

1 0

[slime-devel] create-swank-server
by Marco Baringer 16 Jan '04

16 Jan '04

this adds back the create-swank-server function (why was it removed?) -- Marco Ring the bells that still can ring. Forget the perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

2 1

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 16 Jan '04

16 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.205 slime/ChangeLog:1.210 --- slime/ChangeLog:1.205 Wed Jan 14 18:44:40 2004 +++ slime/ChangeLog Fri Jan 16 02:28:23 2004 @@ -1,3 +1,86 @@ +2004-01-16 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank.lisp (changelog-date): Use *compile-file-truename* instead + of *compile-file-pathname*. + (with-I/O-lock, with-a-connection): The usual CLISP fixes. + + * slime.el: Numerous REPL related fixes. + +2004-01-16 Luke Gorrie <luke(a)bluetail.com> + + * swank-openmcl.lisp: Multiprocessing support. + + * swank.lisp (changelog-date): make-pathname portability fix + (from alanr). + (with-io-redirection): Use (current-connection) instead of + *dispatching-connection* (from alanr). + + * slime.el (slime-init-output-buffer): XEmacs portability fix, and + use header-line-format to show info about Lisp in Emacs21. + +2004-01-15 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank-sbcl.lisp, swank-cmucl.lisp (remove-input-handlers): New + method. + + * swank-allegro.lisp (excl:stream-read-char-no-hang): Import it. + (emacs-connected): Add default method. The method for + no-applicable-method doesn't seem to work. ACL bug? + + * swank-loader.lisp (compile-files-if-needed-serially): Don't + handle compilation errors. We must compile everything because + changelog-date requires *compile-file-truename*. + + * slime.el: (slime-changelog-date) + (slime-check-protocol-version): New functions. + (slime-handle-oob): Handle :check-protocol-version event. + (slime-init-output-buffer): Print some info about the remote Lisp. + (slime-connect): Use it. + (slime-note-transcript-start): Renamed from + slime-insert-transcript-delimiter. + (slime-note-transcript-end): New function. + (slime-with-output-end-mark, slime-repl-insert-prompt) + (slime-repl-show-result, slime-compile-file) + (slime-show-evaluation-result): Insert output from eval commands + after the prompt and asynchronous output before the prompt. Needs + documentation. + (repl-test, repl-read, interactive-eval-output): New tests. + (slime-flush-output): Accept output from all processes. + + * swank.lisp (serve-requests): New function. + (setup-server): Use it. + (start-server): Pass backgroud to setup-server. + (create-connection): Check the protocol version. + (changelog-date): New function. + (make-output-function): Use write-string instead of princ. + + * swank-backend.lisp (remove-input-handlers): New function. + +2004-01-15 Luke Gorrie <luke(a)bluetail.com> + + * slime.el (slime-aux-connect, slime-handle-oob): Support for + (:open-aux-connection port) message where Lisp requests that + Emacs make a connection. These are "auxiliary" connections which + don't (or at least shouldn't) have their own REPL etc. + + * swank.lisp: New support for multiprocessing and multiple + connections + commentary. + (with-a-connection): Macro to execute some forms "with a + connection". This is used in the debugger hook to automatically + create a temporary connection if needed (i.e. if the current + thread doesn't already have one). + (open-aux-connection): Helper function to create an extra + connection to Emacs. + + * swank-sbcl.lisp: Implemented multiprocessing. Not perfect. + + * swank-cmucl.lisp: Implemented new multiprocessing interface. + (create-socket): Make FDs non-blocking when multiprocessing is + enabled. + (startup-multiprocessing): Set *swank-in-background* to :spawn. + + * swank-backend.lisp: Changed multiprocessing interface. + 2004-01-15 Wolfgang Jenkner <wjenkner(a)inode.at> * swank-clisp.lisp (with-blocked-signals): New macro.

1 0

[slime-devel] Bug report
by Ivan Boldyrev 16 Jan '04

16 Jan '04

Resent Slime from CVS HEAD has bug: 1. Load Lisp file. 2. Run slime (CMU CL). 3. Switch to Lisp text buffer. 4. Load file with C-c C-l And REPL is not usable: Assertion failed: (<= (point) slime-repl-input-end-mark) BTW, is it proper place for bugreports? -- Ivan Boldyrev Perl is a language where 2 x 2 is not equal to 4.

3 3

[slime-devel] sldb-quit and debugger for non-repl process
by Alan Ruttenberg 16 Jan '04

16 Jan '04

Following onto my previous message, if I do get a debugger, when I execute sldb-quit I get another error: Can't throw to tag SWANK::SLIME-TOPLEVEL -Alan

2 1

[slime-devel] issues in recent changes
by Alan Ruttenberg 15 Jan '04

15 Jan '04

Just trying them out in openmcl. First thing I run into is that changelog date broke when *compile-file-pathname* was a logical host. In changelog date I think you want (make-pathname :name "ChangeLog" :directory dir :host (pathname-host *compile-file-pathname*))) I'm not making changes myself because logical pathnames can be platform dependent and I don't want to break another platform. Second: I run (process-run-function "foo" (lambda() (invoke-debugger)) ) and this doesn't pop up a sldb window. Upon investigation a problem seems to be that in slime-debug, there is (when (open-stream-p (connection.socket-io *main-connection*)) (with-a-connection () (with-io-redirection ((current-connection)) (swank-debugger-hook c next))))))) *dispatching-connection* is nil with-a-connection checks (current-connection) which succeeds, since there is *main-connection* But then with-io-redirection looks at *dispatching-connection* for its connection. Don't know if this is due to swank-openmcl not being up to date. Is someone currently working on swank-openmcl? -Alan

1 0

[slime-devel] Re: [slime-cvs] CVS update: slime/ChangeLog
by Helmut Eller 15 Jan '04

15 Jan '04

Wolfgang Jenkner <wjenkner(a)common-lisp.net> writes: > + * swank-clisp.lisp (with-blocked-signals): New macro. > + (without-interrupts): Use it. I get now errors/warnings like: WARNING: clos::make-init-form[linux::sigset_t][common-lisp::simple-error]: clos::no-applicable-method: When calling #1=#Y(clos::make-load-form #22Y(00 00 00 00 01 00 00 00 11 17 69 00 00 69 00 01 69 00 02 9C 02 03) (common-lisp::t . common-lisp::t) #(clos::no-method-caller clos::no-applicable-method #1#) #(1. 1. common-lisp::nil common-lisp::nil common-lisp::nil common-lisp::nil) (0.) common-lisp::nil) with arguments WARNING: clos::make-init-form[linux::sigset_t][common-lisp::simple-error]: clos::no-applicable-method: When calling ... Do others see this too? Helmut.

2 1

[slime-devel] Lisp is not ready for requests from the REPL, can it be any other way?
by Dirk Gerrits 15 Jan '04

15 Jan '04

Hi. I've noticed that when one's program is running, one is still free to use SLIME. However, attempting to evaluate an expression produces the message "Lisp is not ready for requests from the REPL". Pretty understandable, because Lisp can only do one thing at a time. But I was wondering if it is possible to 'fix' this. Ie I would like to be able to inspect variables and redefine functions etc. *while the program is running*. Can SLIME help me do this? Or should I post a question to comp.lang.lisp? Thanks in advance, Dirk Gerrits

6 7

[slime-devel] Feature request
by Ignas Mikalajunas 15 Jan '04

15 Jan '04

As we can see there are lot's of fake bug reports associated with using incompatible slime.el version. I think there should be some way to put a function into swank like (equired-slime-version) so slime.el could check whether it is up to date on connect and recompile itself or at least warn the user. Ignas Mikalajunas

4 3

[slime-devel] Partial multiprocessing support, take 2
by Luke Gorrie 15 Jan '04

15 Jan '04

Ahoy, I checked in some very basic multiprocessing support. It works with SBCL and CMUCL and it should be easy to port. Threads automatically create new connections on-demand and can all talk to Emacs at the same time. There is no flow-control, so if 20 threads hit the debugger at once then you can expect 20 layers of SLDB buffers to pop up :-). There is no real policy for standard I/O streams or for interrupting yet -- "it does what it does". There is some commentary in swank.lisp about how multiple-connections work from the Lisp side (short version: one per thread created on demand). I also changed (shrunk) the multiprocessing interface in swank-backend, it should be implementable in 20 lines or so and all that's needed for a port. For people already using threads it should be a great improvement. For people not using threads it shouldn't break anything. There's a caveat for SBCL: spawning threads from SLIME (e.g. the REPL) doesn't work properly. When you spawn a thread in SBCL it inherits the dynamic bindings from its parent, but it seems to get some things that were supposed to be thread-specific/temporary in the bargain. I'll ask Dan about this stuff when he gets back online. BTW, there is a job opening for a threads user-interface designer/hacker. Helmut and I aren't threads users, so we're not the guys to do this stuff. It would be better for some threads hacker(s) to muck around and figure out a good UI, and we can help with the infrastructure (which is hopefully mostly there now). Cheers, Luke

2 3

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 15 Jan '04

15 Jan '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.204 slime/ChangeLog:1.205 --- slime/ChangeLog:1.204 Wed Jan 14 01:54:11 2004 +++ slime/ChangeLog Wed Jan 14 18:44:40 2004 @@ -1,3 +1,11 @@ +2004-01-15 Wolfgang Jenkner <wjenkner(a)inode.at> + + * swank-clisp.lisp (with-blocked-signals): New macro. + (without-interrupts): Use it. + (*use-dedicated-output-stream*, *redirect-output*): Don't set them + here, use the default settings. + Make :linux one of *features* if we find the "LINUX" package. + 2004-01-14 Luke Gorrie <luke(a)bluetail.com> * swank-openmcl.lisp (emacs-connected): Typo fix (missing

1 0

[slime-devel] Backend carnage! Battle stations!
by Luke Gorrie 14 Jan '04

14 Jan '04

Howdy, I just checked in some incompatible changes to the backend interface. I've only updated the CMUCL backend so far, and I'm hoping to fix SBCL (and maybe LispWorks) this evening. Backend hackers please join in :-). The ChangeLog gives the gist, and diff'ing swank-cmucl.lisp against FAIRLY-STABLE will show how the CMUCL backend was updated. Everyone else might want to stick to FAIRLY-STABLE (just updated) for a day or two 'til the dust settles. The basic change is to move most of the protocol logic into portable code. Now the backends provide a generic callback-based TCP server interface which is used by swank.lisp to implement the basic protocol, parts of I/O redirection, and the dedicated output channel. Hopefully it also buys enough generality that multithreading experiments can go forward without breaking backends again. It should be mostly a matter of deleting code and adapting what remains. The new interfaces are documented in swank-backend.lisp. Questions & flames welcome :-) Cheers, Luke

5 10