On 5/19/2013 11:52 PM, Shawn Heisey wrote: > As you can see, the tophints data was all from tests that I didn't run > this time. It happens on both Linux and Windows, and for tests under > lucene as well as solr.
After I sent this, I noted the existence of a .cache directory with many timehints.txt files. This tickled a memory about test timing statistics that survive multiple test runs on jenkins. I can think of two ways to approach fixing this cosmetic problem that don't involve losing the statistics. We could implement one or both of them: The first idea is to have the display of tophints data reflect only the tests that were actually run, not the cached data about every test in the group. The second idea touches on another topic that interests me -- having a build target that will return the tree to a completely pristine state, just like when it is first checked out from svn. I just saw Dawid's reply. It confirmed my vague memory. Thanks for the info about tweaking my local install, I will look into that. Thanks, Shawn --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org