Too much cross-posting?

[David H. Brierley]

In article <159 at zorch.UU.NET> scott at zorch.UU.NET (Scott Hazen Mueller) writes:
>
>Okay, some hard facts:
...
>112 14000   670   95%   132  243.9    39%  0.03    2.7%  comp.sys.att
>423   590    29    3%   202  321.0    59%  0.03    0.1%  unix-pc.general
>427   480    24    2%    16  240.7    67%  0.03    0.1%  unix-pc.sources
>

The readership of the unix-pc.* groups is a lot higher than the arbitron data
suggests because there is a bug in every version of arbitron that I have ever
seen which causes it to exclude these groups from the data.  Somewhere between
lines 200 and 225 (your line numbers are most assuredly different than mine),
you will find two egrep patterns that look like this: '^[a-z]*\.'.  In case you
are not familiar with reading egrep patterns, this says: beginning of line
followed by any number of alphabetics followed by a period.  Now take a look
at the name "unix-pc.general" and you will notice that it contains a
non-alphabetic before the period!  If you receive the unix-pc groups and you
submit arbitron data, please change the egrep patterns to be '^[a-z0-9-]*\.'
and then re-run the arbitron script to submit new data.  Make sure you change
BOTH of the egrep patterns because I don't remember which one is used under
what circumstances.  If everyone on the net does this we should see a huge
jump in the readership of the unix-pc groups.

By the way, just because you are a small site does not mean that Brian does
not want you to submit arbitron data. If only the big sites submit data then
the statistics will be incorrectly skewed in the direction of the big sites.
The more sites of all sizes that submit arbitron data the more accurate the
results will be.  I think we should all submit arbitron data so that the
unix-pc groups get the recognition they deserve.
-- 
David H. Brierley
Home: dave at galaxia.Newport.RI.US   {rayssd,xanth,lazlo,mirror,att}!galaxia!dave
Work: dhb at rayssd.ray.com           {sun,uunet,gatech,necntc,ukma}!rayssd!dhb