On 8/29/15 8:39 AM, Rich Bowen wrote:
I'd love to see a comparison with a half dozen other projects.
I would discourage any reasoning based on aggregate message counts.
Every +1 on a PMC member VOTE counts as a message, for example. The
thing to look carefully at is what is being discussed on the private
list. A lot of discussion bearing on topics important to the
direction of the project is bad, bad, bad. Healthy projects have
quiet private@ lists because pretty much everything they need to
talk about they can and do talk about on the public lists. But
committer / PMC votes, security issues and occasional random legal
or must-be-private people-related things pop up and cause traffic
spikes when they do. So I would not draw conclusions or do
comparisons based on message counts. Better to compare what is
actually being discussed.
On Aug 29, 2015 02:42, Marcus marcus.m...@wtnet.de wrote:
In any case this is too much traffic on the private mailing list.
I would understand Dennis' mail as a wake-up call how much it is currently
and that there is an urgent need to turn down the number of mails.
Am 08/28/2015 11:58 PM, schrieb Phillip Rhodes:
So what, if anything, should we take away from this? My (completely
superficial, naive and uninformed) feeling is that that is a LOT of
on the private list. But maybe not. Anyway, is the idea here that
should be less traffic on that list? More? The same?
I have to admit, I've been pretty dormant for a long-time, so I'm a little
out of touch with what's going on (gone on) here, but you have me
This message optimized for indexing by NSA PRISM
On Fri, Aug 28, 2015 at 3:09 PM, Dennis E. Hamiltonorc...@apache.org
From an AOO PMC Member,
I have compiled a high-level traffic analysis of discussion activity on
the OpenOffice PMC private@ oo.a.o list. These are *statistics* and
noisy ones at that. I am looking for trends that are good-enough at this
level of precision. It is in the nature of private@ that message
and even the topics must be held in confidence.
This report of gross metrics is for the community's appraisal of current
state and later progress. The movement of discussions to the community
when the confidentiality requirements for PMC discussion do not apply
should be seen in movements at this level. Further reports over the
of the year may provide an useful indicator.
OVERALL PRIVATE MESSAGE TRAFFIC
This is a breakdown of the traffic in the 212 days from January through
July, 2015, by role of the sender.
2015 | Private List Messages
thru July | PMC ASF Other All
Totals 1145 182 31 1358
Senders22 23 2368
Per sender 52.0 7.9 1.3 20.0
Per day 5.4 0.90.1 6.4
Of all the messages sent,
84% are by members of the PMC,
16% are by other ASF participants, and
17% are by others.
The ASF participants include members of Apache Infrastructure, Officers
the ASF, and other ASF Members and staff who make posts to the private
list. The Other senders are members of the public and non-PMC Apache
OpenOffice contributors that raise questions or provide information to
PMC via private@.
For the 1145 messages from the 22 PMC members who posted to the list so
far this year,
49% of the messages are from the three
PMC members who were the most vocal
in the studied period.
75% of the messages are from the seven
91% were from the most vocal 11 of the
22 PMC members that posted.
I confess to being one of those top three posters.
NUMBER OF SUBJECTS AND AMOUNT OF DISCUSSION
A review of the same message archives, for January - July, 2015, tallied
168 subjects discussed across 1341 posts,
about 0.8 new topics per day.
The variance of 17 from the first tally
is negligible and will not be corrected.
The raw data is available for auditing
by the PMC.
8.0 is the average number of messages on a
5% is the portion of the overall messages
used in the longest thread, one with
50% of the messages are on the 20 longest
discussion threads. The shortest thread
in that group has 18 messages.
75% of the messages are on the 50 longest
discussions. The shortest threads in
that group have 8 messages.
90% of the messages are on the 84 longest
discussions (i.e., half of the
threads). The shortest threads in
that group have 4 messages each.
The remaining 10% consists of 84 threads
having 3, 2, and 1 messages each.
This does not speak to the quality or the necessity of these messages and
any particular thread. The PMC has detailed