develooper Front page | perl.perl5.porters | Postings from April 2008

Re: A non-scientific survey Re: 5.6.x

Thread Previous | Thread Next
From:
Tels
Date:
April 28, 2008 11:51
Subject:
Re: A non-scientific survey Re: 5.6.x
Message ID:
200804282050.55397@bloodgate.com
On Monday 28 April 2008 19:52:28 you wrote:
> On Mon, 28 Apr 2008, Tels wrote:
[snip]
> It is my understanding that these numbers represent downloads from
> unique IP addresses per day. This was done primarily to disregard
> download "accelerators" that open multiple parallel connections and
> just request a chunk of the file (I have no idea if these
> applications exists to overcome high latency, but they seem to be
> used mostly from connections originating from Asia). Their presence
> was significantly distorting the statistics).
>
> So unless these bots use lots of different IP addresses they should
> only be counted once. This is also compensated somewhat by not
> counting multiple downloads from the same institution on the same day
> multiple time (for example someone downloading and installing
> ActivePerl on 20 machines within the same company on the same day).
>
> Finally, the download numbers for e.g. the AIX builds were
> comparatively really small (I can't remember the number, but maybe
> just 1% of the Windows downloads). I would have expected any bots
> from downloading all directories indiscriminately, which again points
> to bots being just a small percentage of the overall numbers.

Ah okay thanx for the explanation :) With bots, you never know[0].

All the best,

Tels

0: I had some crawl my entire photo collection multiple times. Not fun.

-- 
 Signed on Mon Apr 28 20:48:26 2008 with key 0x93B84C15.
 Get one of my photo posters: http://bloodgate.com/posters
 PGP key on http://bloodgate.com/tels.asc or per email.

 "I am soo clumsy today." *crash*


Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About