Re: [Boost-testing] Meaning of the date in detail regression report?

From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2006-11-24 14:25:18

AlisdairM writes:
> Aleksey Gurtovoy wrote:
> 
> > Roland Schwarz writes:
> > > and the list head for RudbekAssociates-V2 reads
> > >       Mon, 20 Nov 2006 14:45:00 +0000
> > 
> > That's the time when the regression run that produced the results
> > displayed under that column has started getting the sources.
> 
> I don't believe this is the case.

Well, yet it is :). See 'regression.py' source.

> Regularly watching the regression pages, this date looks more like
> the timestamp when the results were uploaded to metacomm.  For
> instance, I believe the metacomm test cycle takes a day or two to
> run,

Nope, just a bit over 8 hours.

> but when it posts its update, the timestamp will be within an hour
> or two of the page-generation timestamp.

The length of the report generation cycle depends on the number and
volume of the newly uploaded results. At the moment, in the worst
case, it can take up to 6 hours, so the situation you observed is not
entirely unlikely.

-- 
Aleksey Gurtovoy
MetaCommunications Engineering