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

Slime startup sometimes: Undefined function SWANK::SYMBOL-INDENTATION
by Mark H. David 12 Dec '15

12 Dec '15

I frequently see this (see debugger backtrace below) when starting SLIME. Not all the time. I suspect it's a timing issue. Anyone see this? Anyone looked into fixing? Thanks, --Mark Undefined function SWANK::SYMBOL-INDENTATION called with arguments (:BREAK-OPTIONS-STRING) . [Condition of type CCL::UNDEFINED-FUNCTION-CALL] Restarts: 0: [CONTINUE] Retry applying SWANK::SYMBOL-INDENTATION to (:BREAK-OPTIONS-STRING). 1: [USE-VALUE] Apply specified function to (:BREAK-OPTIONS-STRING) this time. 2: [STORE-VALUE] Specify a function to use as the definition of SWANK::SYMBOL-INDENTATION. 3: [ABORT-BREAK] Reset this thread 4: [ABORT] Kill this thread Backtrace: 0: ((:INTERNAL SWANK::CONSIDER SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS) :BREAK-OPTIONS-STRING) 1: (#<Anonymous Function #x302000BD1E4F> :BREAK-OPTIONS-STRING) 2: (CCL::APPLY-TO-HTAB-SYMS #<COMPILED-LEXICAL-CLOSURE #x2389553F> #(:NFRAMES #<Unbound> :CODE 0 :PAUSE-PERIOD 0 ...)) 3: (CCL::ITERATE-OVER-PRESENT-SYMBOLS #<Package "KEYWORD"> #<COMPILED-LEXICAL-CLOSURE #x2389553F>) 4: (CCL::ITERATE-OVER-ALL-SYMBOLS #<COMPILED-LEXICAL-CLOSURE #x2389553F>) 5: (SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS #<HASH-TABLE :TEST EQ size 228/305 #x302000AC574D> (#<Package "X86-DARWIN64"> #<Package "X8664"> #<Package "KEYWORD"> #<Package "CCL"> ..) #<Package "COMMON-.. 6: (SWANK::PERFORM-INDENTATION-UPDATE #<MULTITHREADED-CONNECTION #x302000AC56AD> (#<Package "X86-DARWIN64"> #<Package "X8664"> #<Package "KEYWORD"> #<Package "CCL"> #<Package "COMMON-LISP"> ..) #<Package.. 7: ((:INTERNAL SWANK::INDENTATION-CACHE-LOOP)) 8: (SWANK/BACKEND:CALL-WITH-DEBUGGER-HOOK #<Compiled-function SWANK:SWANK-DEBUGGER-HOOK #x302000A414FF> #<COMPILED-LEXICAL-CLOSURE (:INTERNAL SWANK::INDENTATION-CACHE-LOOP) #x302000ACFD1F>) 9: (SWANK::INDENTATION-CACHE-LOOP #<MULTITHREADED-CONNECTION #x302000AC56AD>) 10: (CCL::RUN-PROCESS-INITIAL-FORM #<PROCESS swank-indentation-cache-thread(6) [Active] #x302000AC982D> (#<COMPILED-LEXICAL-CLOSURE (:INTERNAL CCL::%PROCESS-RUN-FUNCTION) #x302000AC95BF>)) 11: ((:INTERNAL (CCL::%PROCESS-PRESET-INTERNAL (PROCESS))) #<PROCESS swank-indentation-cache-thread(6) [Active] #x302000AC982D> (#<COMPILED-LEXICAL-CLOSURE (:INTERNAL ..))) 12: ((:INTERNAL CCL::THREAD-MAKE-STARTUP-FUNCTION))

4 5

help writing a contrib
by Jim Newton 03 Dec '15

03 Dec '15

I hope someone can help me with a few things I'm confused on. I've trying to write a slime contrib (for the first time). I have already written the common-lisp function which works fine. But I want to write the elisp front-end to this function. The elisp function should get the symbol at the point create a temporary file name, then it should call the common lisp function passing those two arguments. After the CL function returns, emacs should open the file named by the tmp file name. There are several issues I'm not sure how to answer. 1. the symbol at the point has an implicit package name, I want the CL function to receive the correct symbol. E.g. suppose the symbol is xyzzy which is CL parses as abc::xyzzy (because if an in-package earlier in the file), I want the CL function called with 'abc::xyzzy not with 'CL::xyzzy. 2. If I try to call the CL function using slime-eval with a string, then CL understand the correct symbol, but slime-eval returns t before the CL function returns, and when I call find-file, the file is not ready to open yet. 3. If I try to call the CL function with slime-eval-async, then the CL function receives the wrong symbol: CL::xyzzy rather than abc::xyzzy. 4. I would like to put the .lisp file also in the contrib directory, but the CL code depends on c2mop. Does this pose a problem? What do I need to assure that if a user loads the file that c2mop has already been loaded? I.e., is there an asd file I need to create or modify? If someone could point me in the right direction, it would be great. BTW here is the elisp function as I currently have it. (defun class-graph (class-name) (interactive "P") (let* ((class-name (or class-name (slime-symbol-at-point))) (tmp-file (make-temp-file (format "%s-" class-name) nil ".png"))) (cond ((slime-eval (format "(class-graph:class-graph-for-slime '%s %S)" class-name tmp-file)) ;; this does not work, becasue tmp-file is not ready as class-graph-for-slime has not yet finished. (find-file tmp-file) (image-mode-fit-frame)) (t (error "failed to create graph in %s" tmp-file)))))

2 1

help writing a contrib
by Jim Newton 03 Dec '15

03 Dec '15

I hope someone can help me with a few things I'm confused on. I've trying to write a slime contrib (for the first time). I have already written the common-lisp function which works fine. But I want to write the elisp front-end to this function. The elisp function should get the symbol at the point create a temporary file name, then it should call the common lisp function passing those two arguments. After the CL function returns, emacs should open the file named by the tmp file name. There are several issues I'm not sure how to answer. 1. the symbol at the point has an implicit package name, I want the CL function to receive the correct symbol. E.g. suppose the symbol is xyzzy which is CL parses as abc::xyzzy (because if an in-package earlier in the file), I want the CL function called with 'abc::xyzzy not with 'CL::xyzzy. 2. If I try to call the CL function using slime-eval with a string, then CL understand the correct symbol, but slime-eval returns t before the CL function returns, and when I call find-file, the file is not ready to open yet. 3. If I try to call the CL function with slime-eval-async, then the CL function receives the wrong symbol: CL::xyzzy rather than abc::xyzzy. 4. I would like to put the .lisp file also in the contrib directory, but the CL code depends on c2mop. Does this pose a problem? What do I need to assure that if a user loads the file that c2mop has already been loaded? I.e., is there an asd file I need to create or modify? If someone could point me in the right direction, it would be great. BTW here is the elisp function as I currently have it. (defun class-graph (class-name) (interactive "P") (let* ((class-name (or class-name (slime-symbol-at-point))) (tmp-file (make-temp-file (format "%s-" class-name) nil ".png"))) (cond ((slime-eval (format "(class-graph:class-graph-for-slime '%s %S)" class-name tmp-file)) ;; this does not work, becasue tmp-file is not ready as class-graph-for-slime has not yet finished. (find-file tmp-file) (image-mode-fit-frame)) (t (error "failed to create graph in %s" tmp-file)))))

1 0

New Maintainer.
by Helmut Eller 27 Nov '15

27 Nov '15

I'm resigning as maintainer. The new maintainer is Stas Boukarev. Helmut

2 2

reconnecting and *GLOBALLY-REDIRECT-IO*
by mega＠retes.hu 26 Nov '15

26 Nov '15

Hi If I connect to a swank server starting simply with: (require :swank) (swank:create-server :port 4095 :dont-close t) and evaluate an output producing form such as: (loop (format t "~S~%" *standard-output*) (sleep 2)) then the output is printed to the repl, but after disconnecting and connecting again the output no longer makes it there. This is seemingly regardless of the value of *GLOBALLY-REDIRECT-IO* (set in ~/.swank.lisp) which is contradicting the documentation at https://common-lisp.net/project/slime/doc/html/Global-IO-Redirection.html Is there a way to get standard streams redirected to the new repl upon reconnect? Thanks, Gabor

5 10

Re: [named-readtables] in-readtable does not touch swank:*readtable-alist* anymore (#8)
by 73budden . 06 Nov '15

06 Nov '15

> Skimming through that slime-devel thread, it's not clear there is a consensus on whether this is the right approach. There is consensus about two things: 1. Understanding readtables in SLIME is useful and desired 2. Frobbing made by (in-readtable) is kludgy and potentially harmful. But I'm not sure people understood that the change I suggest is breaking. Currently, once readtable is associated to a package, every file with that package implies that readtable due to record in *readtable-alist*. So it is enough to issue (in-readtable) once in the package and it sticks to that package until changed. After suggested modification to SLIME and NAMED-READATBLES, explicit in-readtable form is required at every file, and also every time we issue in-package form in REPL. We are also free to add records to *readtable-alist* manually. How will user adapt to suggested modifications? Simplest way is to fill *readtable-alist* in his initialization file.

2 1

using (in-readtable) to select a readtable
by 73budden . 05 Nov '15

05 Nov '15

Hi! I've made some changes so that one can specify a readtable on a per-file basis. It seems to work. Code can be found here: https://bitbucket.org/budden/budden-slime Sorry I'm new to git so I had some problems with cr/lf on Windows. I'd like to get some feedback.

6 24

pprint-lambda-list
by Peter Stirling 19 Oct '15

19 Oct '15

Hi, As part of implementing slime-style macroexpansion in a web app (code[1], demo[2]), I've noticed an oddity in sbcl's pretty printing: CL-USER> (pprint '(lambda (&key (bar 'baz)))) (LAMBDA (&KEY (BAR (QUOTE BAZ)))) ; No value CL-USER> (pprint '(lambda (foo &key (bar 'baz)))) (LAMBDA (FOO &KEY (BAR 'BAZ))) ; No value Looking in the source for SB-PRETTY:PPRINT-LAMBDA-LIST[3] it appears to be written as a state machine, and there are 2 bugs: first the state is set once, and only once, because the state change CASE only runs when FIRST is null (and therefore if you have a positional argument, then all subsequent arguments are printed as if positional), the second is to do with printing default values. 1 https://github.com/pjstirling/pastecode 2 http://pjstirling.plus.com/paste/expand/?id=5&entry=5 3 sbcl/src/code/pprint.lisp:1031 --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

2 1

Reinstalling slime...
by Marco Antoniotti 04 Oct '15

04 Oct '15

Hi I just reinstalled slime on the latest Aquamacs and I have a few questions to bring me back up to speed. Question 1: How can you get the latest Clozure CL to run under slime? Invoking the standard command starts the GUI. (Ok, I should have asked this to the ccl guys…) Question 2: Was the cute initial animation canned? It does not seem to work for me. Thanks Marco -- Marco Antoniotti

3 3

Documentation typos.
by Dmitry Igrishin 15 Sep '15

15 Sep '15

Hello, I've read the documentation and found these typos: https://common-lisp.net/project/slime/doc/html/Inferior_002dlisp.html#Infer… "(see Loading Contribs for information on how to enable the REPL.)" should be "(See Loading Contribs for information on how to enable the REPL.)" https://common-lisp.net/project/slime/doc/html/Editing-Commands.html https://common-lisp.net/project/slime/doc/html/Macro_002dexpansion.html https://common-lisp.net/project/slime/doc/html/Scratch-Buffer.html "sexp" better be "S-expression" https://common-lisp.net/project/slime/doc/html/Compilation.html#Compilation "(See (emacs)Compilation Mode)" should be "(see (Emacs)Compilation Mode.)" https://common-lisp.net/project/slime/doc/html/Multithreading.html#Multithr… "add a binding the swank:*default-worker-thread-bindings*" should be "add a binding to the swank:*default-worker-thread-bindings*" https://common-lisp.net/project/slime/doc/html/Evaluation.html#Evaluation https://common-lisp.net/project/slime/doc/html/Command-Index.html#Command-I… "slime-eval-last-expression-display-output" -- no such command in SLIME https://common-lisp.net/project/slime/doc/html/slime_002dmrepl.html#slime_0… "slime-open-listener" -- no such command in SLIME https://common-lisp.net/project/slime/doc/html/Fuzzy-Completion.html#Fuzzy-… "slime-complete-symbol-function" -- obsolete, so ‘slime-completion-at-point-functions’ should be referenced instead. https://common-lisp.net/project/slime/doc/html/Xref-and-Class-Browser.html#… "This command asks for a class name and displays inheritance tree of for the class." should be "This command asks for a class name and displays inheritance tree of the class." https://common-lisp.net/project/slime/doc/html/SLIME-Enhanced-M_002d_002e.h… "slime-mdot-fu" -- no such contrib. "slime-mdot" is here. https://common-lisp.net/project/slime/doc/html/SLIME-Trace-Dialog.html#SLIM… "the the *trace-output* stream" should be "the *trace-output* stream" https://common-lisp.net/project/slime/doc/html/Tips-and-Tricks.html#Tips-an… "lisp" should be "Lisp" https://common-lisp.net/project/slime/doc/html/Connecting-to-a-remote-lisp.… "swank" should be "Swank" "lisp" should be "Lisp" https://common-lisp.net/project/slime/doc/html/Setting-up-the-lisp-image.ht… "swank" should be "Swank" "lisp" should be "Lisp" "ssh" should be "SSH" "emacs" should be "Emacs" "See See Communication style." should be "See Communication style." "On the emacs side you will use something like" should be "On the Emacs side you will use something like" "(slime-connect "127.0.0.1" 4005))" extra right parenthesis https://common-lisp.net/project/slime/doc/html/Setting-up-Emacs.html#Settin… "ssh" should be "SSH" "swank" should be "Swank" https://common-lisp.net/project/slime/doc/html/Setting-up-pathname-translat… "swank" should be "Swank" "ssh" should be "SSH" "if we want things like" should be "If we want things like" "(See See (tramp)TRAMP User Manual.)" should be "(See (tramp)TRAMP User Manual.)" https://common-lisp.net/project/slime/doc/html/Banner.html#Banner "( See (elisp)Header Lines.)" should be "(see (elisp)Header Lines)" https://common-lisp.net/project/slime/doc/html/Editing-Commands.html#Editin… "s-expression" should be "S-expression" "s-expressions" should be "S-expressions" "sexp" should be "S-expression" https://common-lisp.net/project/slime/doc/html/Typeout-frames.html#Typeout-… "(see Loading Contribs.)" should be "(See Loading Contribs.)" https://common-lisp.net/project/slime/doc/html/TRAMP.html#TRAMP "(see Setting up pathname translations)" should be "(See Setting up pathname translations.)" https://common-lisp.net/project/slime/doc/html/slime_002dsprof.html#slime_0… "swank" should be "Swank" "(callers, calls to)" should be "(callers, callees)" I think these changes will make the documentation to looks better. Also, it would be nice to replace everywhere: lisp -> Lisp emacs -> Emacs slime -> SLIME repl -> REPL ssh -> SSH swank -> Swank sexp -> S-expression

2 2