[Foundation-l] Mail Stats

Michael R. Irwin michael_irwin at verizon.net
Fri Jun 16 23:49:11 UTC 2006


Erik Zachte wrote:

>In the category 'nice to know':
>
>I made a script to measure who is most involved in public mailing lists
>discussions, and on which lists.
>
>You'll find a complete list of mailing lists and how their activy changed
>over time
>http://www.infodisiac.com/Wikipedia/ScanMail/
>
>For each list you'll find a breakdown of posts by author, with board
>activity added separately, e.g. for the foundation list
>http://www.infodisiac.com/Wikipedia/ScanMail/foundation-l.html
>  
>
snip rest

Eric,

There may be a minor error in one of your loops or pattern matching 
comparisons, or the explanation above is slightly off. 

 The appears enwiki-l stats appear to shut off at five posts.   Perhaps 
the graphics is truncated for ease of use and the totals are accurate?

There was a single post by me to enwiki-l.

http://mail.wikimedia.org/pipermail/wikien-l/2002-November/000003.html

The header was:

*Michael R. Irwin* mri_icboise at surfbest.net 
<mailto:wikien-l%40Wikipedia.org?Subject=%5BWikien-l%5D%20Getting%20started&In-Reply-To=>
/Fri Nov 15 03:32:18 UTC 2002

/The other two places I expected to see data on my activity it was there 
under Michael R. Irwin.   Some of it originated from the above email 
address and some from my current address so I have concluded your 
processing routines were focused on the name Michael R. Irwin which is 
what is shown in the y-axis of the summary stats.

It seemed to accurately pick up posts signed as both mirwin and as my 
later handle lazyquasar but they were all annotated as from Michael R. 
Irwin in the header so that was not surprising.

Forgive my pickiness but these stats look incredibly useful for some 
kinds project management activities such as proposal preparation or 
analsysis of participation trends but it is easy to draw wrong 
conclusions if the basis for the data presented and being used is 
misunderstood.

Obviously any subtle error off by one on my data may be way off for 
someone else.

Any further explanation to clarify exactly what is counted and how it is 
summarized if the data is not exactly accurate in the counts of all the 
emails in the archives would be appreciated by me.

Thanks!

regards,
Michael R. Irwin
lazyquasar




More information about the wikimedia-l mailing list