[slime-devel] Slime behavior different (SBCL vs CMUCL)?

[Jeff Cunningham]

John Wiegley wrote:
> On Nov 22, 2007, at 4:48 PM, Jeff Cunningham wrote:
>
>> What about when you aren't running Hunchentoot? Can you interact with 
>> a running program by re-compiling code without stopping it? (assuming 
>> threads are enabled).
>
> When Hunchentoot is not running, I don't run functions that don't 
> exit.  I can imagine that in that case, it may be possible for SLIME 
> to hang waiting for a response from the inferior Lisp.  I just built a 
> threaded SBCL yesterday, so until now I've only used the REPL for 
> commands that return immediately.
>
> John
>
I guess I'm not being very clear. Suppose the following scenario: I am 
testing a loop function that monitors external hardware status and 
responds to state changes. The loop runs continuously, but has been 
launched by me in the REPL by calling a defun. Now while this is 
functioning I determine that I need to change the value of a 
defparameter. Currently, I cannot re-evaluate the defparameter using C-c 
C-c without stopping the loop that is running in the REPL. When I used 
to use CMUCL I could do this. Does this make any sense?

The reason why this is nice to do, is the loop requires half an hour of 
operation to get to the point where the parameter comes into play.

Regards,
Jeff