Garrett D'Amore wrote:
> I'm not sure if this is expected or not... as I'm not a member of the
> Advocacy group.  But the number of Core Contributors in that group is
> double that of its nearest competitor (ON, 48 CCs) and *quadruple* that
> of its next nearest (Tools, 22 CCs.)

In the last election, it was the User Groups community that was huge - that
merged with Marketing to form Advocacy, and inherited the huge number of
contributors.

> The other wrinkle in all this is that some communities have considerable
> overlap in their CC membership.  E.g. many of the ON CC's are probably
> also CC's in other groups.  It would be an interesting data point to
> measure "voting loyalty" for each CG, where each CC gets one vote, which
> is divided equally amongst all the CGs in which they participate.  CCs
> that belong only to one CG contribute 1 to that CG.  CCs that belong to
> two contribute 0.5 to each of those two CGs, and so forth.
> 
> The resulting graph may yield some surprising data about how fairly (or
> otherwise) weighted the election is likely to be.

If you'd like to make charts or different stats, the raw data I used was
from http://poll.opensolaris.org/grants.csv - note that there's a small
bug in that file, in that the entry for user "roamer" has a comma in the
real name field, so doesn't parse correctly when you're counting commas.
After fixing that with a quick vi, my method of counting core contributors
per community was simply:
        awk -F, '$3 == 30 {print $4}' grants.csv | sort | uniq -c

The fields in that file are:
        username, name, rank, community, expiration

Where the ranks are:
        30: Core Contributor
        20: Contributor
        10: Contributor Emeritus

(I cheated and got this all by viewing the JavaScript sources on
poll.opensolaris.org.)

-- 
        -Alan Coopersmith-           alan.coopersmith at sun.com
         Sun Microsystems, Inc. - X Window System Engineering


Reply via email to