[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: The number of popcon.debian.org-submissions is falling



Russ Allbery wrote:

> Raphael Geissert <geissert@debian.org> writes:
> 
>> Something similar can be seen from
>> http://lists.debian.org/stats/debian-news.png
> 
>> But it still can't be compared to:
>> http://lists.debian.org/stats/debian-announce.png
> 
> Those graphs look suspiciously like we weren't purging bouncing addresses
> from the mailing lists and then started doing so.  The sudden drop at the
> start of the flattening of the curve in particular looks a lot like a
> one-time address purge followed by ongoing application of a stricter
> bounce policy.

Yes, the exponential growth is suspicious.

> That doesn't explain all of the change, but it does make it seem less
> dire.

I've been trying to find a relationship with other lists, but it is rather 
difficult (without proper statistical analysis and some adjustments to the 
POV.)

The only other announcements lists I think I missed are:
http://lists.debian.org/stats/debian-devel-announce.png
http://lists.debian.org/stats/debian-security-announce.png

They could in fact demonstrate that it was just a cleanup (actually, it 
would be easier if somebody from the listsmasters team shed some light 
here.) It is common for moderated, announce-only, lists to accumulate 
bouncing addresses because often those sending the announcements are only 
those who receive the bounces.

: http://lists.debian.org/stats/debian-devel.png

Number of subscribers didn't drop, much, but the traffic _did_ drop (less 
flamewars? :)

: http://lists.debian.org/stats/debian-security.png

Except for the two big events, the traffic and subscribers did drop; but the 
number of subscribers increased again.

: http://lists.debian.org/stats/debian-user.png

This one is very interesting, as it is the main contact point from our 
users. The number of subscribers did not drop, but the traffic did (non 
surprisingly, the number of subscribers increases when the traffic 
decreases, and vice-versa.)

I'm not drawing a conclusion, that's left to the reader.

Cheers,
-- 
Raphael Geissert - Debian Developer
www.debian.org - get.debian.net


Reply to: