I am noticing that you seem to be incrementing the lastModified time by 1 ms in SortingVisitorTest. You do know that the smallest unit of time you can count on for that is 1 second? Various file systems support different precisions. So if you are going to sort by lastModified time you either need to use 1 second increments or vary the precision depending on the file system (Good luck with that).
From a quick glance I suspect the other tests are also not using 1 second as a minimum. Ralph > On Nov 28, 2015, at 9:45 PM, Remko Popma <remko.po...@gmail.com> wrote: > > Well this is interesting: on my Windows 10 box all tests pass. > Jenkins is running Ubuntu, and I believe Ralph is running OS X, and both see > the same errors. > So the issue has something to do with NIO on unix file systems. > > What I can see so far is that sorting paths by lastModifiedTime does not seem > to work. > I'm trying to find out more details. > > I'll do my best but I only get feedback on my changes once an hour so it may > take some time to resolve this... > > > > > On Sun, Nov 29, 2015 at 9:19 AM, Remko Popma <remko.po...@gmail.com > <mailto:remko.po...@gmail.com>> wrote: > Strange, all tests passed on my machine before I merged. > I'll take a look. > > Sent from my iPhone > > On 2015/11/29, at 8:43, Ralph Goers <ralph.go...@dslextreme.com > <mailto:ralph.go...@dslextreme.com>> wrote: > >> I am getting test failures in the stuff you checked in. >> >> Ralph >> >>> On Nov 28, 2015, at 9:08 AM, Ralph Goers <ralph.go...@dslextreme.com >>> <mailto:ralph.go...@dslextreme.com>> wrote: >>> >>> Go ahead. I will review it on master. >>> >>> Ralph >>> >>>> On Nov 28, 2015, at 8:37 AM, Remko Popma <remko.po...@gmail.com >>>> <mailto:remko.po...@gmail.com>> wrote: >>>> >>>> Any objections if I merge LOG4J2-435-delete-on-rollover into master? >>> >> >
