On Sun, Nov 15, 2015, at 11:58 PM, Sergio Fernández wrote:
> Marko, the metrics approach has been discussed in the past, for instance
> http://markmail.org/message/ubx3utli3bnltv75 So far my feeling is that
> the ASF prefer to deliver of people to build an opinion of projects rather
> than based them on pure statistical metrics. But I'd be happy to see something
> like that.
Metrics can help people form opinions ("huh, there's been fewer than 5
emails per month on Project Y's dev list ... that sounds like we should
look more closely at that project") but "pure statistical metrics" are a
*horrible* way to decide whether a project is "healthy."
A project could have a very active set of mailing lists, lots of
commits, push out releases regularly - and still be unhealthy in any
number of ways. (e.g. - all of the work is coming from people paid by
one company, or that community is refusing to accept patches from anyone
employed by another company, or any number of other situations that
don't indicate a project that's "healthy" as we understand health.)
Conversely, a project may have minimal traffic, very few commits, and so
forth - but still be "healthy" because it's a mature project that needs
very little development to stay useful and relevant to its community.
Best,
jzb
--
Joe Brockmeier
j...@zonker.net
Twitter: @jzb
http://www.dissociatedpress.net/
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org
For additional commands, e-mail: general-h...@incubator.apache.org
