[Foundation-l] We have the problem

[Milos Rancic]

On Wed, Oct 29, 2008 at 4:48 AM, Michael Snow <wikipedia at verizon.net> wrote:
> Pedro Sanchez wrote:
>> Alternatively, we've learn not to go into long long nitpicking threads.
>> So this is better, less volume, but signal/noise increases.
>>
>> I'm even reading some threads now instead of automatically canning them.
>>
>> Beware of the emailpostcountitis!
>>
> I agree with this, especially as it relates to recent activity. My sense
> is that recent discussions on this list have been very useful (including
> this thread), quite a bit more than they were roughly a year ago.
> Anyway, using my rough impressions as a guide, I find basically no
> correlation between the volume of messages and the quality.
>
> This is not to say that activity levels aren't something to be concerned
> about, although I might want to start with a focus on something other
> than this mailing list. But sometimes lots of visible activity reflects
> (or promotes) lots of productive work; sometimes lots of activity simply
> means we're bogged down in things that are keeping us from productivity;
> sometimes visible activity is lower because we're busy being productive
> in other ways; and sometimes activity is lower because we're not as
> involved in the projects. Using a two-directional measurement to
> describe phenomena with at least four possible compass points will only
> be a very small start, from an analytical perspective.

Pedro's email is very subjective and as such is a part of widespread
problem (just one digression: I was the part of that problem, too;
especially during the previous years, so, this is not a personal
attack): we tend to lie ourselves and to interpret data as everything
is going on fine -- while it is far from truth. The most "objective"
part was not proven: signal/noise ratio may or may not increase with
the less volume of emails. It is related just to ability of some
persons to handle such amount of emails (which is other type of
problems, while it is still a problem).

At the other side, (I have to say) after the second reading, I
realized that Michael's second paragraph is, at least, in a good
direction. While I would like to see suggestions what to measure,
generally asking for deeper analysis is a step forward.

There are a number of possible indicators which may be measured very
easy. This time, I made a statistics of new persons on foundation-l
list per month. It is worst than the previous results: October has two
times less new persons at the list than the next worst October (2004)
had. Actually, it is the worst month *at all* -- after August of 2004
(6) and July of this year (7). September is the worst September for
all years. August is slightly better than August 2004, July is the
worst July ever, May and June are somewhat better than the the worst
May and June (2005).

And, again, numbers of new participants of this list tend to be higher
in the second part of the year, while this October has the lowest
number of new users for this year (and not just for this year, as I
said before).

2004	xxx	xxx	xxx	13	31	25	13	6	22	9	8	9	
2005	17	21	10	8	9	9	8	18	15	10	15	16	
2006	21	12	18	16	18	20	15	28	25	17	20	28	
2007	19	26	13	21	22	18	18	19	14	15	21	19	
2008	23	11	24	15	12	11	7	8	12	4	xxx	xxx