Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to email@example.com.
Kelly F. Hickel wrote: > Well, keeping it in the header would be much better (assuming you can > assure atomicity, which presumably you can). But, it would have been a > much bigger deal and probably wouldn't be acceptable to the cvs > maintainers. I did run cvs under Rational Quantify to determine that > this was the hotspot. I haven't bothered to run the patched version > because it's "fast enough" (at the moment). I don't know what criteria Arthur is using for the cvsnt stable branch at the moment, but for evs as long as a patch doesn't break anything putting it into a testing release is no big deal - that's at least in part what the testing release is for. I suggest submitting the patch to cvsnt to Arthur in the first case since that's where it'll be most immediately useful. evs has a number of other bottlenecks at the moment so it isn't as easy to see the results of speedups (including a significant pause as it enters a directory, that I need to track down). Tony