couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Davis <paul.joseph.da...@gmail.com>
Subject Re: Problems cutting 0.10
Date Sat, 26 Sep 2009 02:12:52 GMT
On Fri, Sep 25, 2009 at 10:03 PM, Noah Slater <nslater@tumbolia.org> wrote:
>> Noah Slater,
>
> Paul Joseph Davis, if you're going to call me by my full name, at
> least get it right. I am called Noah Thomas Benjamin Slater. However,
> you should know, I do not like being called by my full name. You may
> call me Noah "Awesome" Slater on occasion, but please keep it to a
> minimum. I generally discourage public adulation from my fans.
>

You're only allowed one middle name. Them's the rules so Awesome it is.

>> I believe the units are in milliseconds. So 5 seconds would still
>> violate the 3 second limit.
>
> Without looking at the test, I'm going to fire one of the starboard
> bow. This won't do. How on earth am I mean to use the test suite for
> what I need to use it for, as the release manager, when one of the
> tests only works on the same machine as your user agent. Is there any
> way this test can be made more sane? The last thing I want to do is
> "learn to ignore" various errors.
>

The point of the test is to make sure that the response is 'recent'.
If we define recent to be 1 year, the test loses value. I've thought
the same thought you have also thought in that I think the thought of
recent is gray. But I've yet to come up with a better way to test that
we're not caching responses somewhere. Maybe something along the lines
of adding an etag and testing that it changes? That thought doesn't
seem too bad at first blush.

Though I think really we're just saying that you should learn to
synchronize your system clocks as a release manager. XD

> Thanks,
>
> --
> Noah Slater, http://tumbolia.org/nslater
>

HTH,
Paul Davis

Mime
View raw message