No problem. The issue was actually breaking with existing tests, with a specific random seed. Which passed after the fix. A new test could also be added to test for this though. So I see your point.
Joel Bernstein Search Engineer at Heliosearch On Sat, Aug 23, 2014 at 12:47 PM, Ryan Ernst <[email protected]> wrote: > Hi Joel, > > I appreciate your efforts in fixing SOLR-6407, but I don't think this > warrants a respin for a couple reasons: > * SortingResponseWriter is a new class in 4.10, so no existing users > could be broken by the bug > * The commit did not include any new tests. This does not give me > confidence the change actually fixes the issue. > * The tests for SortingResponseWriter are minimal. This worries me > that there could be other bugs for which another respin might be > requested. > > If there are more serious bugs, I am happy to respin and include this > change (hopefully with an explicit test added). But we need to work > towards minimizing the effort needed for a release, and part of that > is setting the bar high for bugs which warrant a respin. > > Thanks > Ryan > > On Sat, Aug 23, 2014 at 7:45 AM, Joel Bernstein <[email protected]> > wrote: > > Hi Ryan, > > > > I just committed and backported a fix for SOLR-6407 which cropped up > > yesterday. It would be great if you could respin so this fix could make > it > > into 4.10. > > > > Thanks, > > Joel > > > > Joel Bernstein > > Search Engineer at Heliosearch > > > > > > On Fri, Aug 22, 2014 at 8:08 PM, Ryan Ernst <[email protected]> wrote: > >> > >> Please vote for the first release candidate for Lucene/Solr 4.10.0. > >> > >> The artifacts can be downloaded here: > >> > >> > http://people.apache.org/~rjernst/staging_area/lucene-solr-4.10.0-RC0-rev1619858 > >> > >> Or you can run the smoker tester directly with this command line > >> (assuming you have JAVA7_HOME set): > >> python3.2 -u dev-tools/scripts/smokeTestRelease.py > >> > >> > http://people.apache.org/~rjernst/staging_area/lucene-solr-4.10.0-RC0-rev1619858 > >> 1619858 4.10.0 /tmp/smoke_test_4_10 > >> > >> Please note, the RC number is starting at 0 because I used the sample > >> command line in buildAndPushRelease.py. If there is another release, > >> I will jump to RC2 to avoid confusion (thus it would be the "second" > >> RC). I also plan to open an issue to clean up some things about > >> buildAndPushRelease.py help (or lack there of). > >> > >> > SUCCESS! [0:35:20.208893] > >> Here is my +1 > >> > >> Thanks, > >> Ryan > >> > >> --------------------------------------------------------------------- > >> 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] > >
