What metric of “most popular” is the author using?
Ubuntu claimed be the most popular Linux distro on their website, backed by hot air. People who didn’t know any better took that at face value, including the author of this shoddy article, perhaps.
They do have statistics about how many systems send upgrade pings. There are some caveats to that, but I believe the difference with other distros is significant enough for that not to matter.
What other desktop Linux would be more popular? Fedora? Arch?
Ubuntu chooses to log upgrade pings to create such statistics. Contrary to Ubuntu, others respect your privacy, and don’t log upgrade pings. Hypothetically, if Ubuntu is the only distro that logs upgrade pings even though everyone uses Linux Mint in practice as an example, they can’t claim to be the most popular distro as for a matter of fact, that reality has more people that use Linux Mint than Ubuntu.
Linux Mint is even more troublesome, because I believe it uses Ubuntu’s repos as well - meaning that they’d be counted in Ubuntu’s logs?
That said, with the exception of probably Distrowatch pageviews (which of course have very little relation to actual usage), AFAIK all proxy metrics we have do point to Ubuntu’s dominance, as well as anecdotal evidence such as the distros you see people using at e.g. FOSDEM. I’d be interested to see any data that might show otherwise though, but until then, my working hypothesis is that Ubuntu is still the most popular desktop Linux.
