lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erick Erickson <erickerick...@gmail.com>
Subject Hudson "failures"
Date Fri, 03 Dec 2010 21:56:02 GMT
Is there any story on figuring out why there are so many false "failures"
on
the nightly builds?  We try to tell people that trunk is "pretty darn good
code,
better tested than 1.4 ever was", which is true.

But putting on my skeptical hat and pretending that I was getting that
statement and trying to decide whether to go with a release or trunk, I'd
take one look at the failure rate of trunk and say "You've got to be
kidding".

I've dug a bit and seen "failures" with the notation "all tests passed".
Huh???

Don't misunderstand me, I think this is really a perception issue not a code
quality issue. But I also think that it's a hurdle to overcome, and a
showstopper
for some that there are so many "red balls" on the Hudson trunk solr
dashboard.

Best
Erick

Mime
View raw message