Hi,

As I said on the issue: No, there is really no reason to run nightly-smoke 
unless you change the build system itsself. In addition, you cannot even run it 
on Windows by default, it always fails 😊

You did nothing wrong! The hiccups happened for other reasons, too. Also 
Jenkins failed. The problem here was that you renamed a file in the root 
directory of Lucene/Solr, this can easily lead to problem, if you don't know 
the whole story. That was not your fault. I just did not see the change - I was 
aware of the problems, but just because I am one of the Jnekins guys... 😊

If Jenkins fails it did a good job. We do not need every commit works 100%. A 
lot of commits (including many of mines) broke Jenkins. Jenkins finds a bug and 
reports it. I generally commit stuff not in the evening before I go to bed, I 
do it in the morning and then I watch my mail folder! Then you can react fast 
and all is fine. No need for perfection! You don't need to wait 90 mins and 
burn CPU power!

Uwe

-----
Uwe Schindler
Achterdiek 19, D-28357 Bremen
http://www.thetaphi.de
eMail: [email protected]

> -----Original Message-----
> From: Jason Gerlowski [mailto:[email protected]]
> Sent: Sunday, July 2, 2017 11:30 PM
> To: [email protected]
> Subject: Recommend nightly-smoke run before uploading patch
> 
> Hi all,
> 
> Mike Drob, Steve Rowe, and myself recently ran into a small hiccup on
> LUCENE-5822 that was caught by the "nightly-smoke" ant task, but not
> by the more standard "precommit" or "test" targets that we recommend
> contributors run before uploading a patch
> (https://wiki.apache.org/solr/HowToContribute#Generating_a_patch).
> 
> With the goal of making less work for people who are nice enough to
> review my patches, I was wondering whether this page should be amended
> to also recommend running the "nightly-smoke" target.  This would have
> a potential cost; the target takes an hour or two to run.  But I
> figured I'd raise the pros/cons here to see if anyone had any strong
> opinions.
> 
> Best,
> 
> Jason
> 
> ---------------------------------------------------------------------
> 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]

Reply via email to