From: Mike S. <m...@pe...> - 2005-09-20 05:17:41
|
On Mon, 19 Sep 2005, David Britton wrote: > I noticed a perf bottleneck in one of the methods when doing a profile > of my code. I have attched the patch. Basically, there was an > uncessary eval/etc going on in the case where we were not in an > init_and_watch situation. Since this method tends to be called a lot, > it added up into a lot of cumulative seconds (during initialization). Thanks for the patch, it's been applied and will be released with 1.01. The patch will increase performance only during init(), you must be calling init() fairly often to see a difference. Since init() should only be called at startup, you must have an unusual use case I'd like to hear about. Thanks! -- Mike Mike Schilli m...@pe... |