On Mon, 2007-07-30 at 16:33 +0100, Josh Blacker wrote: > On 7/30/07, Darren Mansell <[EMAIL PROTECTED]> wrote: > > >From my own experience of talking to others and generally noticing > > trends I would say that Ubuntu gets more home usage than Mac OS in the > > UK. > > > > Does anyone have any facts or figures to confirm or deny this? Can > > anyone suggest any ways to get an idea of how true this is? > > > > Mac OS doesn't seem to be very successful in the UK but Ubuntu is > > getting very popular. > > > > Thanks. > > Find a site that's popular within the UK and ask very very nicely for > (some of) their visitor stats? I think websites can 'ask' your > computer what it's running (I've seen ads that normally say 'Protect > your *Windows* PC' telling me to 'Protect your *Linux* PC'. You should > be able to see breakdown of OSs that way, but I think it only records > Linux, not specifically Ubuntu. Something to look into, anyway. > -- > Josh Blacker >
Good idea. It would have to be a very popular site that would be completely neutral as to the type of people visiting it. Something like bbc.co.uk. But then some of the type of people running Ubuntu would probably be more likely to visit tech sites and niche subject sites. Something thats visited a lot by people in the IT industry would probably have a disproportionate number of Windows hits because of Windows dominance in the enterprise. On Mon, 2007-07-30 at 20:49 +0100, alan c wrote: > > > When my UK machine get security updates, what does the repos. server > record? > -- > alan cocks > Kubuntu user#10391 Exactly. The only people I can of who could have the required information are Canonical. I know they dont like to release any kind of figures due to fears of information mining but surely a number of unique IP hits would be fine? Not the IP's themselves, just the number of different ones. Then its relatively trivial to script a location lookup. That would give a concentration per area graph. I assume they already have this information from when they went to Dell etc. On Mon, 2007-07-30 at 21:13 +0000, Alan Pope wrote: > Hi Alan, > As I understand it this is indeed one of the ways that Ubuntu > sysadmins can > roughly determine the number of users. Of course this isnt perfect. > There's > the problem that more than one person might use one computer, some > people > cache their updates, and many requests will come through an ISP proxy > and > will thus all come from the same/similar IP. > > I understand they also use the time server requests that Ubuntu makes > after > an install to determine rough number of users. Each machine (if net > connected) checks in with ntp.ubuntulinux.org to sync the time. Also > suceptable to error of course as with checking web requests. > > Then there is the "popularity contest" package, which allows the > ubuntu guys > to know what packages are being installed. Problem with this is that > it's an > optional feature, so many people will not enable it. > > I recently heard from a Canonical person that this whole subject of > measuring numbers of users is one that comes up frequently - > especially > within marketing circles. It's very difficult to say what market > penetration > you have when you really don't know. :) > > Cheers, > Al. > >From those techniques they should at least be able to get a conservative estimate. There will be non-connected installs, multiple machines behind NATs etc. Hmmm I'll try getting on to Canonical. Darren -- ubuntu-uk@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-uk https://wiki.kubuntu.org/UKTeam/
