Also breakdowns for each country/city - many images, bit slow to load
Hello folks - I've been crunching a bunch more numbers with some more information about various different aspects of Plex usage. These figures will get better over time, and, they're supposed to be taken as indicative rather than definitive, for a variety of reasons, not least of which is the bunch of scripts I churned out to do this.
I've also got a request - if anyone out there is feeling clever and can find a prettier way to autogenerate graphs from a mysql database (or a CSV file - I can easily dump the mysql data that way) - perhaps with an applescript talking to Numbers or something, or maybe with another perl module I'm not aware of (I'm using GD::Graph at the moment) - I'd be most interested.
Some of these could probably be better as bar graphs, but I'm a little PERL weary at the moment, so I'll post them later.
OK - here we go.
First off, a new one, what sort of machines you are all on (and, while it isn't in the graph, I'm terribly jealous of the 20 or so of you that have 32GB of RAM!).
Please have a look, and I'd be very interested to hear any comments you may have - especially for the smaller countries, or those with not many Plex users.
Now - similar to the last bunch I produced, here’s per country and per capita graphs. You can see the Americans are still out in front for raw numbers, but noone (apart from maybe other Scandinavians) can even come close to the Swedes with number of users per head of population.
OK… now graphs for cities by country (this is going to take some uploading). I’ve only taken the top 10 for each country, otherwise it gets unreadable.
How did you get the statistics on machine type? I can understand getting the location data from IP addresses during the download, but the http headers won't have any machine data in there. Is Plex sending home system information?
the sparkle frameworks sends some anonymous statistical data when it checks for updates (every launch unless you’ve turned it off), and these numbers are more accurate, since it’s based off machines where the sparkle updater has checked in within the last last month or something like that, meaning the statistics are about active plex users, not just the number of people that tried it once ( coughBoxee)
Hmmmm... actually, I generated the model stats from all the logs I had, not just the last month... it doesn't matter too much as we aren't interested in the actual numbers there, just the ratios. I will start doing them as a month by month thing because it'll be interesting to see whether there's shifts to and from different types of hardware as features and hardware evolve.
Plex will be the first thing I load on my iTablet that has a 720p capable screen, full phone functionality and will be replacing my iPhone.
sparkle cant really tell what the machine actually is, just what it claims it is, from the model identifier string i think. that 2.5% hackintosh number is likely much lower than the real ammount of hackintosh users because like your machine, a large number of hackintoshes are set to report themselves as a genuine apple model. (so your hackintosh likely is on the graph as a mac pro)
Actually, I’ve been coming up with some ideas to combat exactly that problem… they’ll be controllable by the user of course, but it’d be nice to get an idea of how many hackintoshes we have - it’ll give (or take away from) some strength to the argument that some hackintosh specific issues need to be addressed.
We’ve got a couple of ZX 80’s out there.
One thing I missed off the list is Xserves. There’s one!
