Michael DeHaan napisal(a):
Hi Michael,
>> Core member posts, regular replies, core member replies in
return.
>>
> In such case there are two incrementations going on:
> - for 'Core member posts, regular replies': CRS+=1
> - for 'regular replies, core member replies in return': CRS+=1
>
> core - regular relation may be reversed as well: regular - core.
> Single counter does the job here.
>
>
I'm not sure I like this idea. Ultimately it seems to be about scoring
"points" for users, and we don't really want to do that... I think a
project focus is a lot more important, and looking at data in aggregate
is key.
Is it one who sends in very big high quality patches (even if they are
small), or one that just asks lots of questions? Must one change lots of
code or send in lots of commits? What about those that speak about
topics for you and help market things?
It's impossible to specify this, which is why I think at most you can
identify "top mailing list posters" (which v2 did) and "top
committeers"
... and it probably
doesn't make sense to pigeonhole users into different classes. As we
attempt to define what "core" means we inject our own biases...
We don't. The definition of the "core" is simple: More than 75% of the
average of replies sent and more than 75% of the average of connections
with other users. No more than that.
When you look at the results you should be aware of how things are
defined and I think that these definitions are simple enough to be
mentioned at the top of every report.
I also think that the way I called this group at the beginning is bad.
It should rather be: core mailing-list members, regular and temporary.
Temporary might not be the best name too. I am not sure what's better
though, 'passive', 'low participation members'?
Different people are great for different things.
This is why EKG shouldn't try to assign meaning to the data -- it should
be about gathering and presenting the data and let people draw their own
interpretations from it.
Fully agreed. And I think that my proposition is not assigning more
meaning to the data than any other. What I wanted to present for end
users is:
mail-list name | exclusiveness level | responsiveness level | temporary
users participation level | pie of core members by domains | pie of
regular members by domains | pie of temporary members by domain
And the line charts for more than one slice to show the dynamics.
But let's assume that we got scores for three mailing lists:
A | 0.67 | 0.43 | 0.12
B | 0.34 | 0.54 | 0.32
C | 0.54 | 0.32 | 0.19
Does it really tell you what to think? I think that it's only about
some general pointers. You look at the scores and you might know more
than before about a situation on given mailing list. But you can't
tell: 'A is better than B' for two reasons - this stat shows
dynamics. Things might change in next slice and moreover without
a context of what is going on on given mailing list you can't say that
the situation is good or bad.
I only think that you may look at the scores, find corner cases - the
situations where the scores are dramatically low or high - and take
a look and try to understand why was it like this. The aim of this
tool is to point you to the mailing lists where the situation is
somewhat different but it doesn't mean that it's good or bad in there.
We definitely don't want to turn this to any kind of Score Card
where
people are trying to get "points" or see who is "more core", or even
assigning score cards to projects.
And we couldn't. At least not in terms of my proposal. Without our
brains and contextual knowledge we cannot score the results like
'++' for this group and '--' for that.
Small projects with niche interests are just as important as some of
the
larger ones... gather the data, make it available, and let people decide
what they want to do with it outside the context of the EKG project.
Agreed. But my proposal is not about scoring or giving ready
interpreatations and it does not discriminating small projects
in any way.
--
Regards,
Wojtek Walczak,
http://tosh.pl/gminick/