When Max writes about metrics, it’s important to remember that he’s not picking on Ubuntu. We firmly believe that Linux success is roughly proportional forward and backward with Fedora success; my personal feeling is that when Linux does better, any hand-waving and blustering to thte contrary by people with prior agendas, Fedora does better at a greater rate since our product has less suckage overall.
Our numbers are, as far as I know, the first hard evidence, backed up by a known, transparent methodology, of the true magnitude of Linux usage. Everything else up to now? Press regurgitation of numbers spouted by half-informed tech consultants, analysts, and other journalists.
Certainly Red Hat has some idea of the number of RHEL boxen out there — at least the ones used by sysadmins who know what they’re doing and why. There’s also CentOS to consider, and that’s before you reach out for Debian, Ubuntu, Slackware, and everything else. The numbers start to add up easily into the tens of millions.
Then think of the legacy versions in place, not just of Fedora but all other Linux distros. It wasn’t too long ago that I met an administrator running a Red Hat 4.2 box for his company’s 1,000+ seat email server. I naively asked, “why not upgrade?” to which his response was an understandable “if it ain’t broke…”. Fat chance counting those boxes.
The code for smolt is out there, free to use. I would encourage other distros to start putting it to work as a way to benefit all of FLOSS. With real numbers in our pockets we can succeed in new and undreamed-of ways.
4.2 – wow. The first Linux systems I administered were 4.2 (in ’98).