[slime-devel] Re: What does "pipelined request" mean?

[Helmut Eller]

Peter Seibel <peter at javamonkey.com> writes:

> Peter Seibel <peter at javamonkey.com> writes:
> 
> > Peter Seibel <peter at javamonkey.com> writes:
> > 
> > > I've got a busted connection to CLISP (along with several other
> > > connections to other Lisps which seem to be working) Whenever I type
> > > anything in the CLISP REPL I get a message from emacs of the form:
> > > 
> > >   ; pipelined request... (swank:listener-eval "
> > >   ")
> > > 
> > > What does this mean?

This message is printed when Emacs sends two subsequent requests, but
the reply for the first request has not arrived at the time the second
request is send.  The message is there to warn the user that the
second request gets probably queued on the Lisp side.  The message is
also printed when the Lisp side hits a bug and enters the debugger
(because the first request is not completed).

> > Ah, here's another clue. In the CLISP *inferior-lisp* CLISP this
> > message appears:
> > 
> >   *** - Lisp stack overflow. RESET
> > 
> > I'm not sure what provokes that. It doesn't happen write away. (And
> > I'm not certain that I haven't somehow mucked something up in my copy
> > of slime though I don't have much changed.)
> 
> Okay. So something seems to be screwed up in CLISP's error handling.
> With CLISP as my only SLIME connection, and having done nothing else,
> I can type: (error "boo!") at the REPL and get the stack overflow
> message in the *inferior-lisp* buffer. The same thing does *not*
> happen when running CLISP outside of SLIME. (For that matter I can
> signal an error in the *inferior-lisp* buffer, and it behaves as
> expected.)
> 
> Anyone know what's up with that?

It is/was caused by the code that prints the arglist.  I now removed
the whole thing that used *print-pprint-dispatch*, because it didn't
work at all in CLISP and the result was different in SBCL and
Lispworks. 

Helmut.