June 2005 - slime-devel - mailman3.common-lisp.net

[slime-devel] password handling
by Christian Lynbech 10 Jun '05

10 Jun '05

Is there a simple way to stuff strings back into slime? Suppose one has a function for connecting to a database. This connection will need a password and suppose further that one is a bit conscious about saving the password in clear text inside the source file ro typing it directly into the slime repl buffer (where it also would appear in cleartext). For shell buffers, I have solved this problem by indentifying the shell modes filter function and (this on the meacs side) adding the advice (defadvice term-emulate-terminal (after term-emulate-terminal-advice activate) "Advice `term-emulate-terminal' to prompt for passwords. The variable `term-password-regexp' determines how to match prompt strings. The variable `term-password-handling' determines whether passwords are actually handled or not." (if (and term-password-handling (string-match term-password-regexp str)) (term-read-password (match-string 0 str)))) Together with the function: (defun term-read-password (&optional prompt) "Read a password and send it to the buffers process. Optional argument PROMPT specifies alternative prompt string." (interactive) (process-send-string nil (my-read-passwd (or prompt "Password: "))) (process-send-string nil "\n")) this allows emacs to recognise that a password is needed, prompt for the password in the minibuffer using 'read-passwd' and the send the password directly to the inferior process without the password appearing directly in the buffer. I am not claiming this to be perfect security but I think it is a little better than nothing. Now I would like to retrofit this functionality to slime, but as the lisp process is not running directly within the slime repl buffer I am somewhat unsure on how to best hook in to the I/O flow. (Some stuff has been left out for brevity, let me know if you need the full emacs code.) ------------------------+----------------------------------------------------- Christian Lynbech | christian #\@ defun #\. dk ------------------------+----------------------------------------------------- Hit the philistines three times over the head with the Elisp reference manual. - petonic(a)hal.com (Michael A. Petonic)

2 1

[slime-devel] PATCH: Support for SBCL HEAD (and future 0.9.2)
by Nikodemus Siivola 10 Jun '05

10 Jun '05

The attached patch provides support for 0.9.1.25 (coming soon to an anoncvs near you), which includes the newfangled WITH-COMPILATION-UNIT :SOURCE-PLIST stuff -- and breaks unpatched Slime. The patch should work with all versions between 0.9.1 and 0.9.1.25 as well. I also took the liberty of documenting Slime as supporting only the latest official release of SBCL, as opposed to a specific version.... Cheers, -- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."

2 1

[slime-devel] SLIME from CVS on Jun 10 2005 won't compile with OpenMCL's 050607 image
by David Steuber 10 Jun '05

10 Jun '05

I updated both today, so I don't know which is at fault. From *inferior-lisp* (load "/Users/david/usr/src/slime/swank-loader.lisp" :verbose t) (swank:start-server "/tmp/slime.1152" :external-format :iso-latin-1-unix) Welcome to OpenMCL Version (Beta: Darwin) 0.14.3-050607! ? ;Loading #P"/Users/david/usr/src/slime/swank-loader.lisp"... ;Compiling "/Users/david/usr/src/slime/swank-backend.lisp"... ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/swank- backend.dfsl"... ;Compiling "/Users/david/usr/src/slime/nregex.lisp"... ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/nregex.dfsl"... ;Compiling "/Users/david/usr/src/slime/metering.lisp"... ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/ metering.dfsl"... ;Compiling "/Users/david/usr/src/slime/swank-openmcl.lisp"... ;Compiler warnings for "/Users/david/usr/src/slime/swank-openmcl.lisp" : ; Unused lexical variable STREAM, in (MAKE-STREAM-INTERACTIVE (T)). ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/swank- openmcl.dfsl"... ;Compiling "/Users/david/usr/src/slime/swank-gray.lisp"... ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/swank- gray.dfsl"... ;Compiling "/Users/david/usr/src/slime/swank.lisp"... ;Loading #P"/Users/david/.slime/fasl/openmcl-0.14-darwin-powerpc/swank.dfsl"... > Error in process listener(1): value 0 is not of the expected type SYMBOL. > While executing: CCL::%SYMPTR->SYMBOL > Type :POP to abort. Type :? for other options. 1 > :b (F0135AB0) : 0 "CCL::%SYMPTR->SYMBOL" 36 (F0135AB0) : 1 "CCL::%USE-PACKAGE-CONFLICT-CHECK" 1664 (F0135AC0) : 2 NIL NIL (F0135AD0) : 3 "CCL::USE-PACKAGE-1" 204 (F0135AE0) : 4 "USE-PACKAGE" 140 (F0135AF0) : 5 "CCL::%DEFINE-PACKAGE" 588 (F0135B00) : 6 "Anonymous Function #x602792E" 48 (F0135B10) : 7 "CCL::%FASLOAD" 860 (F0135B30) : 8 NIL NIL (F0135B40) : 9 "CCL::ATTEMPT-LOAD" 44 (F0135B50) : 10 "CCL::%LOAD" 1132 (F0135B60) : 11 NIL NIL (F0135B70) : 12 "LOAD" 636 (F0135B90) : 13 "COMPILE-FILES-IF-NEEDED-SERIALLY" 256 (F0135BA0) : 14 NIL NIL (F0135BB0) : 15 "CCL::CALL-CHECK-REGS" 72 (F0135BC0) : 16 "CCL::LOAD-FROM-STREAM" 220 (F0135BD0) : 17 "CCL::%LOAD" 1328 (F0135BF0) : 18 NIL NIL (F0135C00) : 19 "LOAD" 636 (F0135C20) : 20 NIL NIL (F0135C30) : 21 "CCL::CALL-CHECK-REGS" 72 (F0135C40) : 22 NIL NIL (F0135C50) : 23 "CCL::TOPLEVEL-EVAL" 152 (F0135C60) : 24 "CCL::READ-LOOP" 844 (F0135CA0) : 25 "CCL:TOPLEVEL-LOOP" 44 (F0135CC0) : 26 "Anonymous Function #x60CFB06" 68 (F0135CD0) : 27 NIL NIL (F0135CE0) : 28 "Anonymous Function #x61196AE" 612 (F0135D00) : 29 "CCL::RUN-PROCESS-INITIAL-FORM" 400 (F0135D30) : 30 NIL NIL (F0135D40) : 31 "Anonymous Function #x60D7E46" 152 (F0135D60) : 32 "Anonymous Function #x60CC21E" 172 1 >

1 2

[slime-devel] can't M-x slime on CMUCL 18c+
by Jack Andrews 08 Jun '05

08 Jun '05

hi. i recently downloaded slime-1.2.1 but it fails to work on, CMUCL 3.0.8 18c+ (debian): i don't know if its because: a) i'm using lisp-safe.core (more typechecks?) b) i should be using an older slime to match the older CMUCL c) ... i'd prefer to use an older slime than a newer cmucl, unless i'm persuaded otherwise. help much appreciated, jack M-x slime Error in function C::DO-CALL: Execution of a form compiled with errors: (DEFSTRUCT (SLIME-INPUT-STREAM (:INCLUDE STRING-STREAM (COMMON-LISP::IN #'SIS/IN) (COMMON-LISP::MISC #'SIS/MISC)) (:CONC-NAME SIS.) (:PRINT-FUNCTION %PRINT-SLIME-OUTPUT-STREAM) (:CONSTRUCTOR MAKE-SLIME-INPUT-STREAM (INPUT-FN SOS))) (INPUT-FN NIL :TYPE FUNCTION) (SOS NIL :TYPE SLIME-OUTPUT-STREAM) (BUFFER "" :TYPE STRING) (INDEX 0 :TYPE KERNEL:INDEX)) Restarts: 0: [CONTINUE] Return NIL from load of #p"home:.slime/fasl/cmu-release x86-linux 3.0.8 18c+ 31 december 2001 build 3030-linux-x86/swank-cmucl.x86f". 1: Return NIL from load of "/home/jack/slime-1.2.1/swank-loader.lisp". 2: [ABORT ] Return to Top-Level.

4 3

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 07 Jun '05

07 Jun '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.708 slime/ChangeLog:1.709 --- slime/ChangeLog:1.708 Fri Jun 3 22:00:28 2005 +++ slime/ChangeLog Tue Jun 7 12:08:04 2005 @@ -1,3 +1,15 @@ +2005-06-07 Espen Wiborg <espenhw(a)grumblesmurf.org> + + * swank-corman.lisp: Convert to Unix line-endings. + (create-socket): Pass through the port argument unmodified, + gettting a random port if 0. Requires supporting change in + <ccl>/modules/sockets.lisp. + (inspect-for-emacs): defimplementation instead of defmethod. + +2005-06-06 Espen Wiborg <espenhw(a)grumblesmurf.org> + + * doc/slime.texi, PROBLEMS: Added notes about CCL. + 2005-06-03 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-background-activities-enabled-p): Allow

1 0

[slime-devel] Re: [Maxima] Slime vs Maxima's nregex
by Nikodemus Siivola 06 Jun '05

06 Jun '05

On Mon, 6 Jun 2005, Raymond Toy wrote: > While playing with Maxima and slime, I've run into a problem. Both > maxima and slime use the nregex package. This would be ok, but we've > modified nregex slightly to make it do what we want, and the changes > are compatible with slime's version. > > To prevent collision, I propose we change the package that nregex is > in from NREGEX to MAXIMA-NREGEX. This shouldn't be a problem and > should prevent collisions with other NREGEX packages in use. Might it be less work for both Slime and Maxima in the long run to register nregex on sourceforge or clnet, or even just keep it in a separate module within either project's CVS, so that there would once again be an upstream for it? Cheers, -- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."

3 3

Re: [slime-devel] password handling
by Hoehle, Joerg-Cyril 06 Jun '05

06 Jun '05

Christian Lynbech wrote: >For shell buffers, I have solved this problem by indentifying the >shell modes filter function > (defadvice term-emulate-terminal (after >term-emulate-terminal-advice activate) Interesting hook function, but did you know that comint provides M-x send-invisible to that effect? M-x send-invisible asks for the hidden text to be sent to the process, echos * instead in the echo area and does not keep input in the history (but still in C-h l aka. view-lossage). It works in comint, inf-lisp and all other modes that are based on comint (originally called cmushell/cmulisp.el by Olin Shivers around 1992). Of course, send-invisible does not work in slime either, so the problem still remains. Regards, Jorg Hohle.

1 0

[slime-devel] Arglist not displayed when typing SPACE in repl
by Paolo Amoroso 04 Jun '05

04 Jun '05

Suppose I type the following at a SLIME repl (# is the cursor): CL-USER> (my-function-name# When I type space, is SLIME supposed to display the function's arglist in the minibuffer? Am I dreaming? Did this work in the past? If so, it no longer does with the version from CVS sources updated less than an hour ago. I use CMUCL Snapshot (19B) (06-2005) with GNU Emacs 21.3.2 under Slackware Linux 10.0. I have: (slime-setup :autodoc t) in my ~/.emacs, and indeed the arglist is displayed in a Lisp mode buffer. Paolo -- Lisp Propulsion Laboratory log - http://www.paoloamoroso.it/log

3 5

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 03 Jun '05

03 Jun '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.706 slime/ChangeLog:1.708 --- slime/ChangeLog:1.706 Wed Jun 1 19:00:13 2005 +++ slime/ChangeLog Fri Jun 3 22:00:28 2005 @@ -1,3 +1,10 @@ +2005-06-03 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-background-activities-enabled-p): Allow + background stuff in repl-mode buffers too. + + * swank-cmucl.lisp (sis/misc): Return t for :interactive-p. + 2005-06-01 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-load-system, slime-oos): Fix bug related to file

1 0

[slime-devel] SLIME supporting Corman Lisp 2.51
by Espen Wiborg 02 Jun '05

02 Jun '05

After much sweat and cursing, I now have SLIME working on Corman Lisp (only tested on 2.51). I've attached the necessary swank-corman.lisp as well as a patch against CVS HEAD. I actually had this working back in December, but due to a lack of round tuits I haven't made it official before. Scissored from swank-corman.lisp: ;;; Notes ;;; ===== ;;; You will need CCL 2.51, and you will *definitely* need to patch ;;; CCL with the patches at ;;; http://www.grumblesmurf.org/lisp/corman-patches, otherwise SLIME ;;; will blow up in your face. You should also follow the ;;; instructions on http://www.grumblesmurf.org/lisp/corman-slime. ;;; ;;; The only communication style currently supported is NIL. ;;; ;;; Starting CCL inside emacs (with M-x slime) seems to work for me ;;; with Corman Lisp 2.51, but I have seen random failures with 2.5 ;;; (sometimes it works, other times it hangs on start or hangs when ;;; initializing WinSock) - starting CCL externally and using M-x ;;; slime-connect always works fine. ;;; ;;; Sometimes CCL gets confused and starts giving you random memory ;;; access violation errors on startup; if this happens, try dumping a ;;; new image. ;;; -- Espen Wiborg <espenhw(a)grumblesmurf.org> The Unix philosophy: Make your program do one thing, and do it well.

5 14