[Fwd: Re: Results of Fedora Participation Survey?]

[Paul W. Frields]

Hi team,

The Duke researchers working on the Fedora community survey (on which
you might recall Greg DeKoenigsberg consulted) have sent me a list of
some additional information that will help them in their continued
work analyzing the survey results.  Basically, they need someone to
run a few SQL queries via JSON:

(1) The count of unique people, determined by full names and/or email
addresses, active in at least one mailing list/group, which would
provide an estimate of the number of active people in the Fedora
Accounts System.  The count won't include actual names or addresses.

(2) The distribution of mailing lists/groups per person (e.g., X
people are in 1 mailing list/group, Y people are in 2 mailing
lists/groups, Z people in a 3 mailing lists/groups, etc), which would
provide an estimate of how many mailing lists/groups active people are
typically part of in the Fedora Accounts System.  Again, these counts
won't include actual names or addresses

In addition, there are some counts they need for specific lists/groups
based on actual responses, and I can work with someone from
Infrastructure to get those.

Is there anyone who could assist me in pulling these numbers?  We can
do this after Fedora 13 GA, if people are hosed with workload right
now.  The researchers are aware people are always very busy around
release time.

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
          Where open source multiplies: http://opensource.com

[Seth Vidal]

On Mon, 17 May 2010, Paul W. Frields wrote:

>  work analyzing the survey results.  Basically, they need someone to
>  run a few SQL queries via JSON:
>
>  (1) The count of unique people, determined by full names and/or email
>  addresses, active in at least one mailing list/group, which would
>  provide an estimate of the number of active people in the Fedora
>  Accounts System.  The count won't include actual names or addresses. 
So we take the total set of FAS people and then look if that address is 
subscribed to ANY of our mailing lists and if so then count them?

Is that right?

That count won't involve them actually POSTING, right?


>  (2) The distribution of mailing lists/groups per person (e.g., X
>  people are in 1 mailing list/group, Y people are in 2 mailing
>  lists/groups, Z people in a 3 mailing lists/groups, etc), which would
>  provide an estimate of how many mailing lists/groups active people are
>  typically part of in the Fedora Accounts System.  Again, these counts
>  won't include actual names or addresses 

I wonder where mpdehaan is when you need to abuse him.


>  In addition, there are some counts they need for specific
> lists/groups
>  based on actual responses, and I can work with someone from
>  Infrastructure to get those. 
responses? Please tell me we're not talking about going through the 
archives.


>  Is there anyone who could assist me in pulling these numbers?  We
> can
>  do this after Fedora 13 GA, if people are hosed with workload right
>  now.  The researchers are aware people are always very busy around
>  release time. 
I can help. I reserve the right to cry and flagellate myself with some 
sort of hooked whip.

-sv

[Paul W. Frields]

On Mon, May 17, 2010 at 04:23:06PM -0400, Seth Vidal wrote:
>  
>  
>  On Mon, 17 May 2010, Paul W. Frields wrote:
>  
>  >work analyzing the survey results.  Basically, they need someone to
>  >run a few SQL queries via JSON:
>  >
>  >(1) The count of unique people, determined by full names and/or email
>  >addresses, active in at least one mailing list/group, which would
>  >provide an estimate of the number of active people in the Fedora
>  >Accounts System.  The count won't include actual names or addresses.
>  
>  So we take the total set of FAS people and then look if that address
>  is subscribed to ANY of our mailing lists and if so then count them?
>  
>  Is that right?
>  
>  That count won't involve them actually POSTING, right? 
Well, they're looking for "active" in that group, which says posting
to me, but I'll check that with the researchers.  I don't think we can
delineate in that way for FAS groups, though; approved memberships is
the extent of what we can track there.

>  >(2) The distribution of mailing lists/groups per person (e.g., X
>  >people are in 1 mailing list/group, Y people are in 2 mailing
>  >lists/groups, Z people in a 3 mailing lists/groups, etc), which would
>  >provide an estimate of how many mailing lists/groups active people are
>  >typically part of in the Fedora Accounts System.  Again, these counts
>  >won't include actual names or addresses
>  
>  I wonder where mpdehaan is when you need to abuse him. 
Actually, you may want mpdehaan-jr, a/k/a ianweller. ;-)

>  >In addition, there are some counts they need for specific
> lists/groups
>  >based on actual responses, and I can work with someone from
>  >Infrastructure to get those.
>  
>  responses? Please tell me we're not talking about going through the
>  archives. 
No -- counts of group/list membership specifically for respondents to
the survey.  It just wasn't appropriate to dump that list of names
here.

I'll get that list to you, along with the specific requests to apply
to it, and some input from the researchers about their definition of
"active" per above.

>  >Is there anyone who could assist me in pulling these numbers?  We
> can
>  >do this after Fedora 13 GA, if people are hosed with workload right
>  >now.  The researchers are aware people are always very busy around
>  >release time.
>  
>  I can help. I reserve the right to cry and flagellate myself with
>  some sort of hooked whip. 
Someone needs a hug! ;-)

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
          Where open source multiplies: http://opensource.com