Can someone explain to me the purpose of distinguishing the :EMACS-INTERRUPT wire message from the SWANK:SIMPLE-BREAK RPC, or why `slime-repl-read-break' on the Emacs side sends an :EMACS-REX of the latter rather than just sending an :EMACS-INTERRUPT message?
* Taylor R Campbell [2006-11-24 08:49+0100] writes:
Can someone explain to me the purpose of distinguishing the :EMACS-INTERRUPT wire message from the SWANK:SIMPLE-BREAK RPC, or why `slime-repl-read-break' on the Emacs side sends an :EMACS-REX of the latter rather than just sending an :EMACS-INTERRUPT message?
It's likely that :EMACS-INTERRUPT didn't yet exist at the time when that code was written. Interrupting was done by sending Unix signals, and it is desirable to avoid signals, if possible. Today it could probably rewritten in terms of :EMACS-INTERRUPT (and READ-CHAR-NO-HANG from the REPL should also be fixed).
Helmut.
Date: Fri, 24 Nov 2006 15:39:21 +0100 From: Helmut Eller heller@common-lisp.net
It's likely that :EMACS-INTERRUPT didn't yet exist at the time when that code was written. Interrupting was done by sending Unix signals, and it is desirable to avoid signals, if possible. Today it could probably rewritten in terms of :EMACS-INTERRUPT (and READ-CHAR-NO-HANG from the REPL should also be fixed).
Could someone make this small change, or perhaps just make `slime-repl-read-break' an alias for `slime-interrupt'? This would be very helpful for SLIME48.
Thanks! I think now `C-c C-c' will finally work reliably in SLIME48, even when waiting for input, which it didn't for a long time.
-
Helmut Eller -
Taylor R Campbell