couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Noah Slater <nsla...@apache.org>
Subject Re: [VOTE] Release Apache CouchDB 1.3.0-rc.1
Date Mon, 11 Mar 2013 23:41:09 GMT
Our problem is not that we ship builds with failing tests. (We have never
done this!) It is that we don't ship... Like ever.

You have already confirmed for me that you were able to get every test
passing. The tests are brittle. We're aware of the problem, and we are
working on it. If you were unable to get a specific test passing, then that
would be very concerning, and would likely point to a bug. But if you are
able to get a test passing with a re-run, then is that really a blocker?

I am not prepared to hold up this release until we have tests that never
ever fail. Because we will be waiting too long. We don't need to be
perfect. As long as we are reasonably sure that the tests DO pass then I
think we're fine.

I understand your concern. I really do. I feel the pain. And have felt the
pain for a while now. And we are making inroads in this area. But the tests
do occasionally fail, and we can't just stop shipping until that is fixed.
These are bugs in the test suite. They are not bugs in CouchDB itself.

I will let someone else comment on the segfault.



On 11 March 2013 22:42, Wendall Cada <wendallc@83864.com> wrote:

> On 03/11/2013 03:05 PM, Noah Slater wrote:
>
>> Flakey tests are frustrating, but I don't think they will block the
>> release.
>>
> I think they are blockers. CouchDB already has a problem with downstream
> maintainers keeping packages current. This effectively pushes these
> problems onto the packagers. IMO, this is a critical PR issue as well. Want
> to improve the project image? Quit shipping builds that require hours of
> work for packagers. Additionally, failing tests block any production
> packages for our company. I set these standards, and I don't think this is
> unreasonable, nor different from what you'd find with other companies using
> couchdb seriously. Encouraging packagers to disable the test suite isn't
> good either. I still maintain a -1 and think this build is broken.
>
>
>> Can you provide details of your segfault?
>>
> I made note of it in this post http://mail-archives.apache.**
> org/mod_mbox/couchdb-dev/**201303.mbox/%3C513E493B.80707%**4083864.com%3E<http://mail-archives.apache.org/mod_mbox/couchdb-dev/201303.mbox/%3C513E493B.80707%4083864.com%3E>
>
> Wendall
>
>>
>>
>> On 11 March 2013 21:22, Wendall Cada <wendallc@83864.com> wrote:
>>
>>  -1
>>>
>>> Tests fail or unable to be run at all  on all my target platforms. Fedora
>>> 18/Centos 6/Amazon Linux 2012.09
>>>
>>> Once the suite finally runs all the way through, I can manually get the
>>> failing test to run on Fedora 18. Centos 6/Amazon Linux just segfaults.
>>> While others may not think this is a -1, I think the tests should ship
>>> in a
>>> condition that they are able to be run. Building should be getting better
>>> for distros, not worse, this is a major regression from my perspective.
>>>
>>> Wendall
>>>
>>>
>>> On 03/10/2013 02:24 PM, Noah Slater wrote:
>>>
>>>  Dear community,
>>>>
>>>> I would like to call a vote on Apache CouchDB 1.3.0-rc.1.
>>>>
>>>> We encourage the whole community to download and test these release
>>>> artefacts so that any critical issues can be resolved before the release
>>>> is
>>>> made. Everyone is free to vote on this release, so get stuck in!
>>>>
>>>> The release artefacts we are voting on are available here:
>>>>
>>>>       wget
>>>> https://dist.apache.org/repos/****dist/dev/couchdb/source/1.3.**0/**<https://dist.apache.org/repos/**dist/dev/couchdb/source/1.3.0/**>
>>>> rc.1/apache-couchdb-1.3.0.tar.****gz<https://dist.apache.org/**
>>>> repos/dist/dev/couchdb/source/**1.3.0/rc.1/apache-couchdb-1.3.**
>>>> 0.tar.gz<https://dist.apache.org/repos/dist/dev/couchdb/source/1.3.0/rc.1/apache-couchdb-1.3.0.tar.gz>
>>>> >
>>>>       wget
>>>> https://dist.apache.org/repos/****dist/dev/couchdb/source/1.3.**0/**<https://dist.apache.org/repos/**dist/dev/couchdb/source/1.3.0/**>
>>>> rc.1/apache-couchdb-1.3.0.tar.****gz.asc<https://dist.apache.**
>>>> org/repos/dist/dev/couchdb/**source/1.3.0/rc.1/apache-**
>>>> couchdb-1.3.0.tar.gz.asc<https://dist.apache.org/repos/dist/dev/couchdb/source/1.3.0/rc.1/apache-couchdb-1.3.0.tar.gz.asc>
>>>> >
>>>>       wget
>>>> https://dist.apache.org/repos/****dist/dev/couchdb/source/1.3.**0/**<https://dist.apache.org/repos/**dist/dev/couchdb/source/1.3.0/**>
>>>> rc.1/apache-couchdb-1.3.0.tar.****gz.ish<https://dist.apache.**
>>>> org/repos/dist/dev/couchdb/**source/1.3.0/rc.1/apache-**
>>>> couchdb-1.3.0.tar.gz.ish<https://dist.apache.org/repos/dist/dev/couchdb/source/1.3.0/rc.1/apache-couchdb-1.3.0.tar.gz.ish>
>>>> >
>>>>       wget
>>>> https://dist.apache.org/repos/****dist/dev/couchdb/source/1.3.**0/**<https://dist.apache.org/repos/**dist/dev/couchdb/source/1.3.0/**>
>>>> rc.1/apache-couchdb-1.3.0.tar.****gz.md5<https://dist.apache.**
>>>> org/repos/dist/dev/couchdb/**source/1.3.0/rc.1/apache-**
>>>> couchdb-1.3.0.tar.gz.md5<https://dist.apache.org/repos/dist/dev/couchdb/source/1.3.0/rc.1/apache-couchdb-1.3.0.tar.gz.md5>
>>>> >
>>>>       wget
>>>> https://dist.apache.org/repos/****dist/dev/couchdb/source/1.3.**0/**<https://dist.apache.org/repos/**dist/dev/couchdb/source/1.3.0/**>
>>>> rc.1/apache-couchdb-1.3.0.tar.****gz.sha<https://dist.apache.**
>>>> org/repos/dist/dev/couchdb/**source/1.3.0/rc.1/apache-**
>>>> couchdb-1.3.0.tar.gz.sha<https://dist.apache.org/repos/dist/dev/couchdb/source/1.3.0/rc.1/apache-couchdb-1.3.0.tar.gz.sha>
>>>> >
>>>>
>>>>
>>>> Please follow the test procedure here:
>>>>
>>>>       http://wiki.apache.org/****couchdb/Test_procedure<http://wiki.apache.org/**couchdb/Test_procedure>
>>>> <http://**wiki.apache.org/couchdb/Test_**procedure<http://wiki.apache.org/couchdb/Test_procedure>
>>>> >
>>>>
>>>>
>>>> Please remember that "rc.1" is an annotation. If the vote passes, these
>>>> artefacts will be released as Apache CouchDB 1.3.0.
>>>>
>>>> Thanks,
>>>>
>>>>
>>>>
>>
>


-- 
NS

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message