Darryl> sperber(a)informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor]) Darryl> wrote: Darryl> While it could be an XEmacs bug, it could also be memory-related: Darryl> * How much RAM do you have in your box? 320 Megs :-) Darryl> * What's the size of the long-running XEmacs process? About 30 Megs. Darryl> * What are the values of `gc-cons-threshold' and `auto-gc-threshold'? The defaults: gc-cons-threshold is 500000. auto-gc-threshold 10000 Darryl> Basically, if your long-running XEmacs is swapping, or is doing many Darryl> gc's (your benchmark eats memory), it may not be surprising that the Darryl> long-running XEmacs takes longer. I don't think it's that: I see the clock GC pointer appearing, and the interval between GCs clearly gets longer. -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla