I'm currently using Kubuntu 7.10.
I installed emacs, and clisp from the package manager, and I installed
slime through cvs.
My .emacs file is
(add-to-list 'load-path "/home/max/slime/slime.el")
(require 'slime)
(add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
(add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
t)))
(setq inferior-lisp-program "usr/bin/clisp")
But when I type M-x slime [no match] appears next to my command.
Any help would be appreciated.
On Nov 10, 2:48 am, ·········@gmail.com wrote:
> I'm currently using Kubuntu 7.10.
>
> I installed emacs, and clisp from the package manager, and I installed
> slime through cvs.
>
> My .emacs file is
>
> (add-to-list 'load-path "/home/max/slime/slime.el")
> (require 'slime)
> (add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
> (add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
> t)))
> (setq inferior-lisp-program "usr/bin/clisp")
>
> But when I type M-x slime [no match] appears next to my command.
>
> Any help would be appreciated.
/usr/bin/clisp ?
On Nov 9, 5:50 pm, ·······@corporate-world.lisp.de" <······@corporate-
world.lisp.de> wrote:
> On Nov 10, 2:48 am, ·········@gmail.com wrote:
>
>
>
> > I'm currently using Kubuntu 7.10.
>
> > I installed emacs, and clisp from the package manager, and I installed
> > slime through cvs.
>
> > My .emacs file is
>
> > (add-to-list 'load-path "/home/max/slime/slime.el")
> > (require 'slime)
> > (add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
> > (add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
> > t)))
> > (setq inferior-lisp-program "usr/bin/clisp")
>
> > But when I type M-x slime [no match] appears next to my command.
>
> > Any help would be appreciated.
>
> /usr/bin/clisp ?
The executable file clisp is in /usr/bin
On Nov 10, 2:54 am, ·········@gmail.com wrote:
> On Nov 9, 5:50 pm, ·······@corporate-world.lisp.de" <······@corporate-
>
>
>
> world.lisp.de> wrote:
> > On Nov 10, 2:48 am, ·········@gmail.com wrote:
>
> > > I'm currently using Kubuntu 7.10.
>
> > > I installed emacs, and clisp from the package manager, and I installed
> > > slime through cvs.
>
> > > My .emacs file is
>
> > > (add-to-list 'load-path "/home/max/slime/slime.el")
> > > (require 'slime)
> > > (add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
> > > (add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
> > > t)))
> > > (setq inferior-lisp-program "usr/bin/clisp")
>
> > > But when I type M-x slime [no match] appears next to my command.
>
> > > Any help would be appreciated.
>
> > /usr/bin/clisp ?
>
> The executable file clisp is in /usr/bin
usr/bin or /usr/bin
······@corporate-world.lisp.de wrote:
>>>> (setq inferior-lisp-program "usr/bin/clisp")
>>>> But when I type M-x slime [no match] appears next to my command.
>>>> Any help would be appreciated.
>>> /usr/bin/clisp ?
>> The executable file clisp is in /usr/bin
>
> usr/bin or /usr/bin
>
Yeah, the missing / at the front is what jumps out at me too. Try
making sure that's in there and try it again.
On Sat, 10 Nov 2007 01:48:56 +0000, max.baroi wrote:
> I'm currently using Kubuntu 7.10.
>
> I installed emacs, and clisp from the package manager, and I installed
> slime through cvs.
>
> My .emacs file is
>
> (add-to-list 'load-path "/home/max/slime/slime.el") (require 'slime)
> (add-hook 'lisp-mode-hook (lambda () (slime-mode t))) (add-hook
> 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode t)))
> (setq inferior-lisp-program "usr/bin/clisp")
>
> But when I type M-x slime [no match] appears next to my command.
>
> Any help would be appreciated.
Try putting (setq inferior...) before (require 'slime), and, more
importantly, try putting (slime-setup) somewhere after it. You might also
note that your load path can (and maybe should) point to
"/home/max/slime/".
mb> I'm currently using Kubuntu 7.10.
mb> I installed emacs, and clisp from the package manager, and I installed
mb> slime through cvs.
mb> My .emacs file is
mb> (add-to-list 'load-path "/home/max/slime/slime.el")
i believe you need to add "/home/max/slime", not the el file
mb> (require 'slime)
mb> But when I type M-x slime [no match] appears next to my command.
you can try executing (require 'slime) from .emacs or from scratch buffer --
if it cannot find it it will barf something, if it can it will print slime.
mb> I installed emacs, and clisp from the package manager, and I installed
mb> slime through cvs.
and why did not you install package slime via package manager? that will
automatically add it to the path
"Alex Mizrahi" <········@users.sourceforge.net> writes:
> mb> I installed emacs, and clisp from the package manager, and I installed
> mb> slime through cvs.
>
> and why did not you install package slime via package manager? that will
> automatically add it to the path
Because the last official slime release was over an year ago.
I generally advice against installing Slime from one's distribution
package manager---unless it provides a special slime-cvs package that
automatically fetches current CVS HEAD. (The reason is that the first
thing someone, who goes against this advice, hears is to upgrade to CVS
anyway when he encounters problems.)
-T.
mb>>> I installed emacs, and clisp from the package manager, and I
mb>>> installed slime through cvs.
??>>
??>> and why did not you install package slime via package manager? that
??>> will automatically add it to the path
TCR> Because the last official slime release was over an year ago.
nevertheless SLIME included in a package might have been tested with Lisps
included into that system and with XEmacs/Emacs versions, so it's quite
likely that at least basic functionality will work.
TCR> I generally advice against installing Slime from one's distribution
TCR> package manager---unless it provides a special slime-cvs package that
TCR> automatically fetches current CVS HEAD. (The reason is that the first
TCR> thing someone, who goes against this advice, hears is to upgrade to
TCR> CVS anyway when he encounters problems.)
recently i've erroneously upgraded to CVS HEAD (while i actually needed to
configure XEmacs), and now i don't have inspection of return values in
REPL -- it's mysteriously have gone. i liked that feature..
what would you recommend - to downgrade to stable version that was used for
many monthes without a single trouble, or try updating to CVS HEAD several
more times until it minimally works?
certainly after using SLIME for years it won't be much problem for me to
pick version that works correctly with functionality i need.
however "newbies" do not actually know what is correct behaviour for SLIME,
how stable it should be etc. so if there are more chances that stable
version included in distro will work -- because it was tested, and basic
functionality *should* work -- they should first try distro version.
once they'll need additional features, or if they will encounter problems,
or they'll upgraded Lisp implementation -- they can try to fetch current CVS
HEAD.
Alex Mizrahi wrote:
>
> recently i've erroneously upgraded to CVS HEAD (while i actually
> needed to configure XEmacs), and now i don't have inspection of return
> values in REPL -- it's mysteriously have gone. i liked that feature..
> what would you recommend - to downgrade to stable version that was
> used for many monthes without a single trouble, or try updating to CVS
> HEAD several more times until it minimally works?
>
This is probably the slime core/contribs split rearing its head again,
you probably just have to do
(slime-setup '(slime-fancy slime-asdf))
in your .emacs (unless they've changed things again since I last
updated).
While I understand the split is handy for slime developers, having
slime starting up out-of-box with large chunks of "considered cool"
functionality turned off is a pity in some ways.
David Golden <············@oceanfree.net> writes:
> While I understand the split is handy for slime developers, having
> slime starting up out-of-box with large chunks of "considered cool"
> functionality turned off is a pity in some ways.
The contribs are not (politically) part of Slime. This means that,
strictly speaking, they are not even maintained by the slime
developers. The contribs are supposed to be maintained by whoever cares
about the specific functionality a contrib provides (and this just happens
to subsume slime developers.)
I.e. the contribs are essentially maintained by you.
But more to your point: Are you subscribed to slime-devel? If so, why
haven't you spoken up the time I raised this issue and actually
advocated an easier migration path to an all-included configuration?
-T.
Tobias C. Rittweiler wrote:
> But more to your point: Are you subscribed to slime-devel?
No. I just scan the archives every so often. I'm just a user of slime,
not to date a developer of it. None of this is major criticism, it's
just a niggle about discoverability of the contribs that could be
solved by one-liner changes to the main docs. A readme *in* the contrib
dir is not quite adequate if you want to draw people's attention to the
existence of the contrib dir in the first place - And of course, its
special importance in the slime case seeing as it's where a bunch of
the stuff that used to be in slime core has gone, rather than being
e.g. contributed patches to get the project working on obscure
platforms like many "contrib" dirs in open source projects.
In the slime case the contrib dir could perhaps better be
called "plugins" or "addons" or "extensions", "contrib" reflects how
slime developers see them, "addons" would reflect how end-users
(who happen to be developers) might see them (hey, as contrib/README
says 'They are essentially "add-ons"').
> If so, why haven't you spoken up the time
Well (a) I wasn't on the list and (b) when I hit the issue (in my
unlucky case after the changes to contrib/ but before the changes to
slime-setup) I was considering doing so when people were still talking
about requiring people to write lambda exprs and load hooks in
their .emacs just to load the contribs... but then slime-setup was
adjusted to take a list of contrib names and slime-fancy appeared, so I
figured (slime-setup '(slime-fancy ...)) was adequate, apart from the
documentation issue.
Before you say it, I know it makes most sense for the online web docs to
match the latest stable release tarball from april 2006 (so before the
split) rather than unstable cvs, at least if there's only one version
of the online docs. It's just right now, people who use cvs but
blindly refer to the online manual (or even the cvs toplevel README or
cvs version of the manual...) may be briefly caught out.
Yes, there's a README *in* the contrib dir and mentions scattered
through the ChangeLog. But you have to not overlook the contrib dir
(whether in search for the functionality that from the perspective of a
user of an old slime has disappeared or as a new user) before you find
that README. Upgraders will presumably work out fairly quickly once
they see the talk of "moved to contrib" in the ChangeLog (that they're
of course looking through because some beloved functionality has
mysteriously vanished), but there's just a bit of a discoverability
gap: A reminder of the existence of contrib/, perhaps the coolness of
the contribs therein and the straightforward slime-setup argument
allowing their loading, in the manual and the toplevel (not
subdirectory!) README is probably all that is needed e.g. after
the setup instructions, something like
"slime-setup can take a list of contribs to load for additional non-core
functionality such as fuzzy completion and asdf packaging system
support. See contrib/README for full details."
"Alex Mizrahi" <········@users.sourceforge.net> writes:
> TCR> I generally advice against installing Slime from one's distribution
> TCR> package manager---unless it provides a special slime-cvs package that
> TCR> automatically fetches current CVS HEAD. (The reason is that the first
> TCR> thing someone, who goes against this advice, hears is to upgrade to
> TCR> CVS anyway when he encounters problems.)
>
> recently i've erroneously upgraded to CVS HEAD (while i actually needed to
> configure XEmacs), and now i don't have inspection of return values in
> REPL -- it's mysteriously have gone. i liked that feature..
> what would you recommend - to downgrade to stable version that was used for
> many monthes without a single trouble, or try updating to CVS HEAD several
> more times until it minimally works?
It has not mysteriously gone. A lot of functionality has been moved from
the core into a contrib system some months ago. Please see the file
`contrib/README' in the Slime checkout for more information.
The functionality you're missing can be found in the
`slime-presentation-streams' contrib. Additionally, you probably are
also interested in `slime-fancy', `slime-asdf', and maybe `slime-tramp'.
HTH,
-T.
Okay,
my .emacs file is now
(add-to-list 'load-path "/home/max/slime/")
(require 'slime)
(add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
(add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
t)))
(setq inferior-lisp-program "/usr/bin/sbcl")
When I enter M-x slime, the prompt comes up, but weird things happen
like if the pormpt is at the top, after I enter something the buffer
scrolls so that the prompt is at the bottom. And in the message
buffer, there's this message:
"Polling "/tmp/slime.6621".. (Abort with 'M-x-slime-abort-
connection'.)
Which is there whether or not I'm typing anything.
mb> Okay,
mb> my .emacs file is now
mb> (add-to-list 'load-path "/home/max/slime/")
mb> (require 'slime)
mb> (add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
mb> (add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
mb> t)))
mb> (setq inferior-lisp-program "/usr/bin/sbcl")
mb> When I enter M-x slime, the prompt comes up, but weird things happen
mb> like if the pormpt is at the top, after I enter something the buffer
mb> scrolls so that the prompt is at the bottom. And in the message
mb> buffer, there's this message:
mb> "Polling "/tmp/slime.6621".. (Abort with 'M-x-slime-abort-
mb> connection'.)
i'm not sure i've understood you.. first buffer SLIME opens shows
compilation/loading process. you shouldn't write anything there.
once it's done, it will show you a fancy prompt.
if it hangs there, please shows us last messages you see, or ones you find
relevant.
on my system first messages look like this:
(progn (load "/usr/share/common-lisp/source/slime/swank-loader.lisp"
:verbose t) (funcall (read-from-string "swank:start-server")
"/tmp/alex/slime.3466" :coding-system "utf-8-unix"))
This is SBCL 1.0.9, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.
The compilation/loading process does complete, but when I type an
expression into the prompt and then press enter, the buffer
automatically scrolls so the prompt is now at the bottom of the
screen.
And I get the "Polling "/tmp/slime.6621".. (Abort with 'M-x-slime-
abort-connection'.)" message in that little message buffer at the
bottom instead of the little procedure form reminders like '(+ &rest
args)'.
·········@gmail.com wrote:
>
> The compilation/loading process does complete, but when I type an
> expression into the prompt and then press enter, the buffer
> automatically scrolls so the prompt is now at the bottom of the
> screen.
>
That sounds almost like normal behaviour (except it usually doesn't
kick in until you've got more than a screenful of output): slime repl
has comint-like show-maximum-output functionality. Dunno if it's
customisable or if that's what's doing it on your system.
And why wouldn't you want that, though? There's never going to be
anything after the final active repl prompt except what you're typing
in?
Your prompt, by the way, should probably look like "CL-USER> ",
and the emacs modeline say "slime-repl". If your prompt looks like "* "
and your emacs modeline says "inferior lisp", you're not in slime, but
in an inferior-lisp-mode buffer (which slime uses underneath as an
implementation detail), which is a comint mode window, and as such has
comint show-maximum-output functionality (which is customizable
somewhat: you can turn it off and if it's on it places the repl prompt
variable "scroll-margin"'s value lines from the bottom).
On Nov 12, 6:27 am, David Golden <············@oceanfree.net> wrote:
> ·········@gmail.com wrote:
>
> > The compilation/loading process does complete, but when I type an
> > expression into the prompt and then press enter, the buffer
> > automatically scrolls so the prompt is now at the bottom of the
> > screen.
>
> That sounds almost like normal behaviour (except it usually doesn't
> kick in until you've got more than a screenful of output): slime repl
> has comint-like show-maximum-output functionality. Dunno if it's
> customisable or if that's what's doing it on your system.
>
> And why wouldn't you want that, though? There's never going to be
> anything after the final active repl prompt except what you're typing
> in?
I find it annoying that if the prompt is at the top of the page and I
type
(if <i>test</i>
then if I press enter so I can type the true clause on the next line,
the
buffer scrolls to the bottom of the page. This never happened when I
used lispbox
in windows, or the scheme interpreter on my school's computers.
> Your prompt, by the way, should probably look like "CL-USER> ",
> and the emacs modeline say "slime-repl". If your prompt looks like "* "
> and your emacs modeline says "inferior lisp", you're not in slime, but
> in an inferior-lisp-mode buffer (which slime uses underneath as an
> implementation detail), which is a comint mode window, and as such has
> comint show-maximum-output functionality (which is customizable
> somewhat: you can turn it off and if it's on it places the repl prompt
> variable "scroll-margin"'s value lines from the bottom).
My prompt is '0] '
·········@gmail.com wrote:
> My prompt is '0] '
Er. And what buffer are you in? What does your emacs modeline say?
That IS a typical sbcl (and several other lisps) debug prompt though.
slime normally takes over the debugger once loaded, so my guess is it's
dropping into the sbcl debugger before swank (the common lisp side of
slime) is finished loading.
I can certainly simulate such an event (by placing (error "pants") in my
~/.swank.lisp ...), and it drops me into aa 0] debug prompt in
*inferior-lisp*, with the usual comint show-maximum-output behaviour
(which indeed causes warp on "(if <enter" ).
My guess is you've got a whole lot of stuff /before/ the 0] prompt ?
(and your modeline says *inferior-lisp* ?)
Really, this might be better handled in a slime-specific forum, but WHAT
is before the 0] prompt? something a little like the below, maybe?
debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread"
{A6DD779}>:
pants
Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL.
restarts (invokable by number or by possibly-abbreviated name):
0: [ABORT] Exit debugger, returning to top level.
(SB-INT:SIMPLE-EVAL-IN-LEXENV (ERROR "pants") #<NULL-LEXENV>)
0]
> Er. And what buffer are you in? What does your emacs modeline say?
*inferior-lisp*
> My guess is you've got a whole lot of stuff /before/ the 0] prompt ?
> (and your modeline says *inferior-lisp* ?)
>
> Really, this might be better handled in a slime-specific forum, but WHAT
> is before the 0] prompt? something a little like the below, maybe?
>
> debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread"
> {A6DD779}>:
> pants
>
> Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL.
>
> restarts (invokable by number or by possibly-abbreviated name):
> 0: [ABORT] Exit debugger, returning to top level.
>
> (SB-INT:SIMPLE-EVAL-IN-LEXENV (ERROR "pants") #<NULL-LEXENV>)
> 0]
Yeah, when I start up, I get this:
<b>
WARNING: These Swank interfaces are unimplemented:
(CALLS-WHO DISASSEMBLE-FRAME SLDB-BREAK-AT-START SLDB-BREAK-ON-RETURN
WHO-SPECIALIZES)
debugger invoked on a SB-BSD-SOCKETS:SOCKET-ERROR in thread #<THREAD
"initial thread" {A806681}>:
Socket error in "bind": 99 (Cannot assign requested address)
Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL.
restarts (invokable by number or by possibly-abbreviated name):
0: [ABORT] Exit debugger, returning to top level.
(SB-BSD-SOCKETS:SOCKET-ERROR "bind")
0]
</b>
·········@gmail.com wrote:
> *inferior-lisp*
Right. So you're not in slime yet.
> debugger invoked on a SB-BSD-SOCKETS:SOCKET-ERROR in thread #<THREAD
> "initial thread" {A806681}>:
> Socket error in "bind": 99 (Cannot assign requested address)
>
And that is an error message (things that programmers tend not to have
the luxury of ignoring I'm afraid...).
That points to the problem: Swank in SBCL can't open a communications
socket on your system. Why? Well, see above. But Whyyyyy? Well, I
dunno exactly, maybe something weird about your network settings. Does
127.0.0.1 actually exist? i.e. if you open up a shell window and
type "ping 127.0.0.1", do you get responses?
On Nov 13, 8:05 am, David Golden <············@oceanfree.net> wrote:
> ·········@gmail.com wrote:
> > *inferior-lisp*
>
> Right. So you're not in slime yet.
>
> > debugger invoked on a SB-BSD-SOCKETS:SOCKET-ERROR in thread #<THREAD
> > "initial thread" {A806681}>:
> > Socket error in "bind": 99 (Cannot assign requested address)
>
> And that is an error message (things that programmers tend not to have
> the luxury of ignoring I'm afraid...).
>
> That points to the problem: Swank in SBCL can't open a communications
> socket on your system. Why? Well, see above. But Whyyyyy? Well, I
> dunno exactly, maybe something weird about your network settings. Does
> 127.0.0.1 actually exist? i.e. if you open up a shell window and
> type "ping 127.0.0.1", do you get responses?
No, I don't get a response when I type "ping 127.0.0.1".
·········@gmail.com wrote:
> No, I don't get a response when I type "ping 127.0.0.1".
Then your linux IPv4 network configuration is screwy, not really a slime
problem except that slime currently relies on it working: 127.0.0.1 is
the conventional IPv4 loopback address of your own machine, commonly
called "localhost". Slime uses IPv4 sockets between the emacs and the
common lisp side - head over to ubuntu/kubuntu forums and ask them why
you don't have 127.0.0.1. It may have something to do with
the "Network Manager" service.
While it's entirely possible to deliberately have a different/strange
(or no) IPv4 loopback network configuration under linux, I doubt that's
the case here.
·········@gmail.com wrote:
> Okay,
> my .emacs file is now
>
> (add-to-list 'load-path "/home/max/slime/")
> (require 'slime)
> (add-hook 'lisp-mode-hook (lambda () (slime-mode t)))
> (add-hook 'inferior-lisp-mode-hook (lambda () (inferior-slime-mode
> t)))
> (setq inferior-lisp-program "/usr/bin/sbcl")
>
Try a slime-setup? Here's my setup in .emacs (there's also some
custom-set-variables entries for slime). slime-setup establishes
the hooks for you, so you don't need most of the above.
(beware if you're using the distro's sbcl that depending on
the distro it can be very out of date). Note that
"lisp" is a symlink to "sbcl" on my system, so I don't alter
inferior-lisp-program, you'll probably still need that.
-8<-----------------------------------------------------
; SLIME setup
(add-to-list 'load-path "/home/david/src/slime")
(require 'slime)
; if from slime cvs
(slime-setup '(slime-fancy slime-asdf))
; if the slime release
; (slime-setup)
; dunno if all of these are still right.
(setq slime-multiprocessing t)
(setq slime-truncate-lines nil)
(setq slime-net-coding-system 'utf-8-unix)
(global-set-key "\C-cs" 'slime-selector)
;; HyperSpec access package
(setq common-lisp-hyperspec-root
"file:/home/david/doc/cl-hyperspec/HyperSpec/")
-8<-----------------------------------------------------