I think the locale/ timezone used should be simply dumped as part of repeat string. That array we pick from doesn't need to be sorted, it can contain different entries depending on the JVM, etc.
D. On Wed, Apr 25, 2012 at 12:26 AM, Uwe Schindler <[email protected]> wrote: > Hi, > > Since the last openjdk6,7 ports upgrades, the timezones update may change > with every single FreeBSD update, because JDK timezone files are now a > separate package shared across all ports. So it does not depend on the JDK > version. Die to a bug in the openjdk port this is disabled for openjdk6 at > the moment, but will come back! > > Uwe > > ----- > Uwe Schindler > H.-H.-Meier-Allee 63, D-28213 Bremen > http://www.thetaphi.de > eMail: [email protected] > > >> -----Original Message----- >> From: [email protected] [mailto:[email protected]] On Behalf Of >> Dawid Weiss >> Sent: Wednesday, April 25, 2012 12:22 AM >> To: [email protected] >> Subject: Re: [JENKINS] Lucene-Solr-tests-only-trunk-java7 - Build # 2324 - >> Failure >> >> I think the problem with repeatability here may be this, Chris: >> https://issues.apache.org/jira/browse/LUCENE-4020 >> >> The log on Jenkins for this seed says: >> >> NOTE: test params are: codec=Lucene40: {}, >> sim=RandomSimilarityProvider(queryNorm=false,coord=false): {}, >> locale=th_TH_TH_#u-nu-thai, timezone=Asia/Karachi >> >> and on my machine: >> >> NOTE: test params are: codec=Lucene40: {}, >> sim=RandomSimilarityProvider(queryNorm=false,coord=false): {}, >> locale=zh_CN, timezone=Australia/Yancowinna >> >> Enforcing these on my machine still didn't result in that original error >> though. I >> tried on Jenkins though (with the same original JVM) >> and: >> >> ant test-core -Dtests.class=*.TimeZoneUtilsTest -Dtests.method=testRandom - >> Dtests.seed=A928CC75CAAEE111 >> -Dtests.multiplier=3 -Dargs="-Dfile.encoding=US-ASCII" >> -Dtests.locale=th_TH_TH#u-nu-thai -Dtests.timezone=Asia/Karachi >> >> sure does reproduce every single time... >> >> Dawid >> >> On Tue, Apr 24, 2012 at 7:46 PM, Chris Hostetter <[email protected]> >> wrote: >> > >> > : FWIW: I can not reproduce this using hte specified seed... >> > >> > Even when i added explicit checks for the input that the test >> > complained about ("GMT-00") i still can't reproduce ... but i'm using >> > Java6 and i don't have Java7. >> > >> > I double checked the code, and i seriously don't understand how this >> > failure could happen -- unless multiple calls to >> > TimeZone.getTimeZone("GMT-00") return diff objects. (the only diff >> > between the expected and actual code paths are that in the "actual" >> > path, the helper method can return null instead of calling >> > TimeZone.getTimeZone if the input string doesn't match some rules. >> > >> > Can anybody reproduce this failure? >> > >> > you shouldn't even need to trust the seed, just run the entire >> > TimeZoneUtilsTest class with the changes to "testCustom" i just >> > committed that explicitly test "GMT-00" >> > >> > : ant test -Dtests.class=*.TimeZoneUtilsTest -Dtests.method=testRandom - >> Dtests.seed=A928CC75CAAEE111 -Dtests.multiplier=3 -Dargs="- >> Dfile.encoding=US-ASCII" >> > : >> > : And the same test passed on the latest trunk (non java7) jenkins build... >> > : >> > : https://builds.apache.org/job/Lucene-Solr-tests-only-trunk/13455/ >> > >> > -Hoss >> > >> > --------------------------------------------------------------------- >> > To unsubscribe, e-mail: [email protected] For >> > additional commands, e-mail: [email protected] >> > >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: [email protected] For additional >> commands, e-mail: [email protected] > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
