Fwd: CLISP EMACS SWANK remote connection crash
C Habs <[email protected]> Sat, 10 Aug 2019 17:06:17 +0000
| Newsgroups | gmane.lisp.clisp.general |
|---|---|
| Message-ID | <CANe8y8Y=kNKKSXaquPNj1AUY-O7QqYT6Q1so6iOJ3tCzMLj05w@mail.gmail.com> |
Just to add: I have installed SBCL and started that lisp in place of CLISP, but the setup and same steps taken below apply. I didn't change anything re:SBCL - just started it as installed (and loaded swank etc using quicklisp just as I did with CLISP). Everything working fine and i am given restart options etc on submitting a syntax error. To me that is indicating it is something to do with the CLISP relationship with SLIME/SWANK - but I will defer to more keener minds. I forgot to mention that I am using GNU CLISP 2.49.93+ thank you Habs ---------- Forwarded message --------- From: C Habs <[email protected]> Date: Sat, 10 Aug 2019 at 10:06 Subject: CLISP EMACS SWANK remote connection crash To: <[email protected]> Hello list@ I will start by saying 'sorry' in advance, if there is already previous information about this issue that I have missed. Essentially, with CLISP running on a remote server and when connecting to it remotely from Emacs on a client, with any sort of syntax error typed at the REPL, the session is disconnected and CLISP at the remote end terminated. There seems to be no handling of these type of errors and I am wondering if it is a setting I am missing, or simply 'one of those incompatibility' things yet to be resolved as such ? I welcome any suggestions to try trace where the problem might be in the setup (which I believe is fine according to all the material I have read to create such a remote setup arrangement) and 'fix' it, if this type of setup works fine for others without issue that is ? Regards and thank you Habs Brief account of setup: Both client and server machines are linux, both with all the same package versions of CLISP emacs slime/swank, ssh etc. The only difference is server is 32Bit linux and client is 64bit. I have CLISP running on a remote 'server' machine; started with, /usr/bin/clisp -m 10M. I have tried with the '-I' flag too. Once the [remote] CLISP is running, at its REPL prompt I have issued, (asdf:load-system :swank), which loads fine as it seems; followed by, (swank:create-server), which starts up on default port 4005. On the client machine, creating the ssh tunnel from client to server as, ssh -L 4005:127.0.0.1:4005 user@server. offers no problem and sets up fine. Then starting Emacs, which incidentally has set, in .emacs, (setq inferior-lisp-program "/usr/bin/clisp -m 10MB -I") ... all appears well with that too. Once Emacs is running, issuing the 'slime-connect' command and entering the requisite host (127.0.0.1) and port (4005) values connects to the remote CLISP session and REPL prompt appears, as expected. All firewall ports open etc. Typing an incorrect syntax expression causes a crash and it seems there is not error handling or recovery operating. If there are no syntax errors in anything submitted, all appears fine. _______________________________________________ clisp-list mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/clisp-list