Open Stack

On 04/24/2015 08:30 AM, Sylvain Bauza wrote:
> 
> 
> Le 24/04/2015 14:11, Russell Bryant a écrit :
>> On 04/24/2015 07:21 AM, Amrith Kumar wrote:
>>> We had a hypothesis about why +0 was rarely used (never conclusively
>>> proved). Our hypothesis was that since Stackalytics didn't count +0's
>>> it led to an increased propensity to -1 something. It would be
>>> wonderful if we could try the experiment of giving credit for 0's and
>>> seeing if it changes behavior.
>> I think this makes a lot of sense.  These stats really do drive
>> behavior.  I'd certainly be open to a patch to reviewstats [1] to count
>> +0 comments and I think it would be good for stackalytics to consider
>> the same.
>>
>> [1] http://git.openstack.org/cgit/openstack-infra/reviewstats
>>
> 
> Just a question I have since a while... I know that Stackalytics and
> reviewstats are not counting the same things for reviews (reviewstats is
> definitely better because it counts all the comments, and not just if
> I'm happy or not with a specific patchset). Could Stackalytics modify
> its behaviour to mimic reviewstats ?
> 
> Ideally, Stackalytics should call the reviewstats API and reviewstats
> should be hosted by infra IMHO (it"s already in the -infra namespace...)
> so we could prevent duplicates.

There has been talk of moving stackalytics into infra somewhere.

They're implemented *very* differently, so it's not really practical for
stackalytics to use reviewstats code.  Maybe it's good to have 2 anyway,
as it helps point out subtle differences that implementations can make,
like you've pointed out here.

-- 
Russell Bryant