The question is, what do we do about File.lastModified() calls in Commons non-test code?
On Sat, 8 Aug 2020 at 16:50, Gary Gregory <garydgreg...@gmail.com> wrote: > > FWIW, there recently was a bug found by Commons Lang in Java 15 and 16. > Java 16 has now fixed the bug but it will not be fixed in Java 15, at > least not before GA. > > > Gary > > On Sat, Aug 8, 2020, 11:45 Matt Sicker <boa...@gmail.com> wrote: > > > Even if it’s fixed upstream, it’ll never get backported most likely. Even > > compiler bugs aren’t backported (see for example a bug related to > > reflection on intersection types that was fixed in Java 9; I can’t find an > > exact reference, but multiple bugs related to this were all address in 9). > > > > On Fri, Aug 7, 2020 at 12:03 sebb <seb...@gmail.com> wrote: > > > > > If so, then it is not sufficient to run tests on Windows, it looks > > > like we need to run tests on Windows with the appropriate version of > > > Java. > > > > > > Presumably that is why Jenkins did not show the error. > > > > > > However, fixing the test won't fix the code - there may be instances > > > of lastModified elsewhere. > > > > > > The Java bug needs to be fixed. > > > > > > > > > On Fri, 7 Aug 2020 at 17:12, Gary Gregory <garydgreg...@gmail.com> > > wrote: > > > > > > > > On Fri, Aug 7, 2020 at 10:04 AM Matt Sicker <boa...@gmail.com> wrote: > > > > > > > > > That sounds like a file modification time granularity issue. Does > > > Windows > > > > > not support milliseconds or finer for file modification times? It > > > could be > > > > > that the test executed too fast! > > > > > > > > > > > > > This might be a classic bug I've run into at work: > > > > https://bugs.openjdk.java.net/browse/JDK-8177809 > > > > > > > > Basically, never use File.lastModified(), use > > > Files.getLastModifiedTime(). > > > > > > > > Gary > > > > > > > > > > > > > > > > > > On Fri, Aug 7, 2020 at 09:01 Gary Gregory <garydgreg...@gmail.com> > > > wrote: > > > > > > > > > > > As of recently, I am seing: > > > > > > > > > > > > [INFO] Running org.apache.commons.io.FileUtilsTestCase > > > > > > [ERROR] Tests run: 142, Failures: 2, Errors: 0, Skipped: 1, Time > > > elapsed: > > > > > > 5.613 s <<< FAILURE! - in org.apache.commons.io.FileUtilsTestCase > > > > > > [ERROR] testCopyFile2WithoutFileDatePreservation Time elapsed: > > > 0.038 s > > > > > > <<< FAILURE! > > > > > > org.opentest4j.AssertionFailedError: Check last modified date not > > > same as > > > > > > input ==> expected: not equal but was: <1596807101505> > > > > > > at > > > > > > org.apache.commons.io > > > > > > > > > > > > > > > > .FileUtilsTestCase.testCopyFile2WithoutFileDatePreservation(FileUtilsTestCase.java:1229) > > > > > > > > > > > > [ERROR] testCopyDirectoryPreserveDates Time elapsed: 0.027 s <<< > > > > > FAILURE! > > > > > > org.opentest4j.AssertionFailedError: expected: <true> but was: > > > <false> > > > > > > at > > > > > > org.apache.commons.io > > > > > > > > > > > > > > > > .FileUtilsTestCase.testCopyDirectoryPreserveDates(FileUtilsTestCase.java:1403) > > > > > > > > > > > > I am wondering if GitHub/Travis can also be set up to run at least > > > one > > > > > > build on Windows so we can catch issues like these? > > > > > > > > > > > > Gary > > > > > > > > > > > -- > > > > > Matt Sicker <boa...@gmail.com> > > > > > > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > > For additional commands, e-mail: dev-h...@commons.apache.org > > > > > > -- > > Matt Sicker <boa...@gmail.com> > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
