On Thursday, October 29, 2015 at 2:19:08 PM UTC-4, badgolferman wrote:
DerbyDad03 wrote:

On Thursday, October 29, 2015 at 8:49:45 AM UTC-4, badgolferman wrote:
badgolferman wrote:

Perhaps you should access Google Groups and inspect the
newsgroup's statistics there. Maybe it will suit your purposes
better than what I post to the group. There used to be a way to
view that data in the older version of Google Groups.

According to Google Groups, alt.home.repair had the following
statistics for 09/2015:
Total Posts: 5486
Total Topics: 354

My newsreader tabulated:
Number of threads ................... 430
Number of articles .................. 5594

One thing to consider is the x:no-archive bit which some people
set. Since I download the headers daily to my newsreader that
setting doesn't apply. It looks like Google Groups honors that
setting, thus the potential difference in our numbers.

I'm no expert in this stats thing, so I have a question related to
the x:no-archive bit. My question is stated in the form of a
"conclusion", but it really is a question.

Google Groups: Total Topics: 354
Your newsreader: Number of threads: 430

Assuming Topics and Threads are equivalent, that must mean that 76
Topics/Threads had one of two things associated with them:

1 - They were single post threads where the author had the
x:no-archive bit set, thus the entire thread (which was comprised of
a single post) was not counted by Google.

2 - Every person that was involved with that thread had the
x:no-archive bit set, thus the entire thread (which was comprised of
multiple posts) was not counted by Google.

Both of those situations seem odd. ~18% of all threads fit one of
those two criteria? That seems high.

I don't know how Google Groups compiles their statistics. What I post
is a function of my newsreader and merely informational, not
authoritative. If you like it great, if you don't then just ignore it.

I wasn't being personal, I was just curious. When you said...

"It looks like Google Groups honors that setting, thus the potential difference in our numbers."

....I though it was open for discussion.

I guess not.

On 10/29/2015 4:43 AM, badgolferman wrote:
Don Y wrote:

Hi,

On 10/1/2015 6:19 PM, badgolferman wrote:
XanaNews Statistic for alt.home.repair. 10/1/2015 9:18:32 PM

From article 700122 (9/1/2015 12:00:42 AM) to article 705696
(9/30/2015 11:44:32 PM)

It looks like this is something tabulated client-side? So,
just reflects your particular news server connection(s)?
I.e., is not really "authoritative" -- if your news server
never sees a post (or a newsgroup!), that wouldn't be
reflected in these figures?

I'm looking for a (somewhat authoritative) source of these
sorts of figures so I can determine how many posts my
"news agent" is opting not to show to me.

It appears that your totals are reflecting what you are
seeing from the Eternal September server?

[I guess I could just install XanaNews and see what it tells me]

Yes, it is a report generated by my newsclient Xananews. Generally I
use Albasani as my newsserver, but sometimes I use Eternal-September as
well. Regardless, there are very few messages that may get filtered
out at the server level. Usually they are obvious spam. And if they
are being filtered by the server then you wouldn't see them either, so
what's the difference?

My question has to do with where the data originated and how it was
collected. E.g., if I look at *my* server, the numbers WILL be
very different -- because my server intentionally filters what it
chooses to expose to me (if it knows I have no desire to hear
posts filled with pejoratives, it won't even show me that the post
exists!). I've known other operators who have imposed other or
similar criteria on what they pass through for particular groups.

[N.B. I think comms that are distributed via the airwaves (ham)
are thusly filtered.]

Perhaps you should access Google Groups and inspect the newsgroup's
statistics there. Maybe it will suit your purposes better than what I
post to the group. There used to be a way to view that data in the
older version of Google Groups.

No need. I can use XanaNews to see what my server tells "me" as
a user and what it would tell a "generic" user (which wouldn't
impose the same filtering constraints). Knowing that you (XanaNews)
are counting posts at the client means I can exploit that mechanism
to do the counting for me (instead of having to grep(1) my server's
logs)

Thanks!