So I tried to figure out why rejecting a bundle is so slow. It triggers a call to repowidget.reload which calls thgrepo.thginvalidate. This in turn invalidates the dirstate among other things.
It's seems like a pretty drastic thing to do when setting and clearing bundles. Especially since the repo object in both cases has JUST been constructed like 1 second prior to being invalidated. Removing the invalidation from these cases results in a noticeable speedup, but I don't have the code overview to know what implications such a change might have. Addressing the problem from the opposite end, in which cases IS it actually necessary to call thginvalidate from reload? reload is used for setting and clearing revset filters as well, and a few other things. -Sune ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity, and more. Splunk takes this data and makes sense of it. IT sense. And common sense. http://p.sf.net/sfu/splunk-novd2d _______________________________________________ Tortoisehg-develop mailing list Tortoisehg-develop@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/tortoisehg-develop
