[slime-devel] slime-repl does not appear
Hello, I'm have a fresh 32-bit Ubuntu 8.10 (Intrepid) install. I obtained both the 2009-01-03 and 2009-01-04 slime tarfile / cvs snapshot. Both of them exhibit the same issue. With both SBCL and CMUCL, the slime repl does not appear. The inferior lisp window is displayed. The 'connected' message, along with its jaunty colloquial appendage, shows in the mini-buffer. Sometimes I would get a swank disconnected (end of file) message on a socket, other times it appeared as if there were no error-related messages at all. I did 'apt-get slime', pointed to that one, and it works. This is Emacs 22.2.1. I believe the apt repository's version comes from 2008-02-23. Is the comment in that slime.el that "SLIME is compatible with GNU Emacs 20 and 21" still valid, i.e. is slime not compatible with Emacs 22? (Even if so, that doesn't explain why the apt version does work while the cvs version does not, other than that the change may have been intentional.) Jeff
"Jeff Caldwell" <jeffrey.d.caldwell@gmail.com> writes:
Hello,
I'm have a fresh 32-bit Ubuntu 8.10 (Intrepid) install. I obtained both the 2009-01-03 and 2009-01-04 slime tarfile / cvs snapshot. Both of them exhibit the same issue. With both SBCL and CMUCL, the slime repl does not appear. The inferior lisp window is displayed. The 'connected' message, along with its jaunty colloquial appendage, shows in the mini-buffer. Sometimes I would get a swank disconnected (end of file) message on a socket, other times it appeared as if there were no error-related messages at all.
The REPL moved to a contrib. Instead of (slime-setup), place (slime-setup '(slime-fancy slime-asdf)) into your ~/.emacs. -T.
* Jeff Caldwell [2009-01-04 19:02+0100] writes:
Hello,
I'm have a fresh 32-bit Ubuntu 8.10 (Intrepid) install. I obtained both the 2009-01-03 and 2009-01-04 slime tarfile / cvs snapshot. Both of them exhibit the same issue. With both SBCL and CMUCL, the slime repl does not appear. The inferior lisp window is displayed. The 'connected' message, along with its jaunty colloquial appendage, shows in the mini-buffer.
The slime-repl is no longer loaded by default. You can do that with (slime-setup '(slime-repl)) in your .emacs. Alternatively you can try the 'inferior-slime contrib which adds symbol completion and package tracking to the *inferior-lisp*. Or 'slime-mrepl which lets you open multple listeners with M-x slime-open-listener.
Sometimes I would get a swank disconnected (end of file) message on a socket, other times it appeared as if there were no error-related messages at all.
We haven't figured out yet what is causing this disconnect. It may well be an Emacs problem or bug.
Is the comment in that slime.el that "SLIME is compatible with GNU Emacs 20 and 21" still valid, i.e. is slime not compatible with Emacs 22? (Even if so, that doesn't explain why the apt version does work while the cvs version does not, other than that the change may have been intentional.)
Emacs 20 is no longer supported. Emacs 21, 22, 23 and XEmacs should work (modulo bugs of course). Helmut.
Helmut Eller <heller@common-lisp.net> writes:
Alternatively you can try the 'inferior-slime contrib which adds symbol completion and package tracking to the *inferior-lisp*. Or 'slime-mrepl which lets you open multple listeners with M-x slime-open-listener.
Does it allow copying presentations between listeners of the same connection? -T.
* Tobias C. Rittweiler [2009-01-04 22:26+0100] writes:
Alternatively you can try the 'inferior-slime contrib which adds symbol completion and package tracking to the *inferior-lisp*. Or 'slime-mrepl which lets you open multple listeners with M-x slime-open-listener.
Does it allow copying presentations between listeners of the same connection?
Copying from slime-repl to slime-mrepl will probably work. But slime-mrepl doesn't write the result as presentation. Helmut.
Thanks, Helmut. I had grabbed both the cvs snapshot tarfile and the pdf link under documentation from the common-lisp.net slime page. After reading your response, I checked. The project page's pdf is from 2005 while the cvs snapshot was current. I didn't see a current ref. manual in the doc subdir of the cvs snapshot. Even the snapshot's README describes the old setup method, not the new one you describe below. Is there a current manual? What you described is significant new functionality, which I'm happier to have than a new manual. I'm just curious about the current state of affairs and e.g. status of the manual. Perhaps if the project page warned the manual is out of date and to read the (hopefully updated) README for setup info? Thanks, Jeff On Sun, Jan 4, 2009 at 2:40 PM, Helmut Eller <heller@common-lisp.net> wrote: ...
The slime-repl is no longer loaded by default. You can do that with (slime-setup '(slime-repl)) in your .emacs.
Alternatively you can try the 'inferior-slime contrib which adds symbol completion and package tracking to the *inferior-lisp*. Or 'slime-mrepl which lets you open multple listeners with M-x slime-open-listener.
* Jeff Caldwell [2009-01-06 12:04+0100] writes:
Thanks, Helmut. I had grabbed both the cvs snapshot tarfile and the pdf link under documentation from the common-lisp.net slime page. After reading your response, I checked. The project page's pdf is from 2005 while the cvs snapshot was current. I didn't see a current ref. manual in the doc subdir of the cvs snapshot. Even the snapshot's README describes the old setup method, not the new one you describe below.
Is there a current manual? What you described is significant new functionality, which I'm happier to have than a new manual. I'm just curious about the current state of affairs and e.g. status of the manual.
The most current manual is always in the doc/ directory. On Linux, you can type "make" to build it. It's currently not very much in sync with the code--the bits about the REPL are missing--but it's more up-to-date than the one on the webpage.
Perhaps if the project page warned the manual is out of date and to read the (hopefully updated) README for setup info?
The README is current. It just describes the minimal configuration. Some people don't like that but, since it is minimal, it has the best chance to work the first time. I think that a working setup, even though minimal, is better than everything that doesn't work. Helmut.
participants (3)
-
Helmut Eller -
Jeff Caldwell -
Tobias C. Rittweiler