Date: Wed, 02 Aug 2006 13:49:10 +0100 From: Alex Zbyslaw <xfb52@dial.pipex.com> To: freebsd-questions@freebsd.org Subject: Re: Gotta start somewhere ... how many of us are really out there? Message-ID: <44D09F46.6020300@dial.pipex.com> In-Reply-To: <df9ac37c0608012122q196a6434jf849cc7bd8c1156@mail.gmail.com> References: <20060728164526.E27679@ganymede.hub.org> <871ws1v261.fsf@photon.homelinux.org> <20060731220830.B27679@ganymede.hub.org> <ef10de9a0608010121j154c7ael7ece0997a479572e@mail.gmail.com> <20060801120058.O27679@ganymede.hub.org> <17615.30414.314802.792740@jerusalem.litteratus.org> <ef10de9a0608011037w3609b5a6k1709aea61d43ed0f@mail.gmail.com> <20060801223754.U27679@ganymede.hub.org> <ef10de9a0608011859q45bdd636o757fb4aba2d3404d@mail.gmail.com> <20060801230301.Q27679@ganymede.hub.org> <df9ac37c0608012122q196a6434jf849cc7bd8c1156@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Atom Powers wrote: > > It's still going to take you at least a release to get it into the > base install. But if you can find a way to use the portsnap data and > get useful information out of the cvsup data you can probably get > numbers now with an error margin as low as 8% to 15%. Hey, I said that a week ago! Guess I agree with you :-) Not quite convinced by the error margin, but as long as you count too low then I see no problem. If, as Nikolas pointed out, a URL-based reporting scheme can be bombarded with fakes, as a vendor I would not want to listen to any numbers it produced. But the question then goes back to: can you make any kind of count out of cvsup servers? Someone already said they thought you couldn't. At the end of the day, I think that unique IP address is as close as it's possible to get to host count. It will undercount NATed hosts and networks with single cvsup/portsnap distribution points, and will overcount variable IP addresses. The latter, I think matters the least as long as you do your stats over a short enough period (e.g. 1 month). That wouldn't overcount much and deliberate faking would be hard and limited (how many IP addresses can one faker get access to?). Then, as long as the methodology is clearly explained along with any stats, you'd have the ammunition to persuade vendors (we hope). --Alex
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?44D09F46.6020300>