[slime-devel] large data set

Helmut Eller heller at common-lisp.net
Wed Mar 28 07:10:29 UTC 2012


* Joe Corneli [2012-03-27 21:41] writes:

> Hello:
>
> I'm processing a pretty big data set, but I was surprised to see SLIME
> fail with this message:
>  Evaluation aborted on #<SWANK::SWANK-ERROR {1003C395A1}>.
>
> ... while the result of my computation continued to be printed on the
> *inferior-lisp* buffer,
>   dispatch-event: (:WRITE-STRING "((0 0 0 0 0 0 [... 158M worth of 0's and 1's]
>
> Here is the rest of what I found there, in case it helps with
> debugging.

The maximum message length is 2^24 bytes.  Do you really want to send
158MB?  Or would be be happy if SLIME detects the problem and enters the
debugger instead of disconnecting?

> (Are there any configuration options you would recommend when working
> with large-scale data?)

Setting printer variables like *print-length* to fairly small values is
usually a good idea.  

Helmut





More information about the slime-devel mailing list