Two Small iOS 26 Usage Share Updates ⇥ fosstodon.org

On Friday, I received an email from Aodhan Cullen, CEO of StatCounter, confirming iOS 26 users had been incorrectly counted as iOS 18.x in its analytics software and, accordingly, in its public trends. Cullen said the company was working on a patch. According to a note pinned today to the top of its iOS version chart, corrected reporting only began rolling out yesterday. However, because this chart represents a version share breakdown for a month that is mostly behind us, more accurate figures will start becoming noticeable in February.

That is the first update. The second is by way of Timo Tijhof, principal engineer at Wikimedia, who points me to Wikimedia’s network-wide stats showing, as of 11 January, around 50% of “Mobile Safari” visitors were using iOS 26, compared to 41% using iOS 18. (Also, 2.8% using “Mobile Safari 19”, and I suppose that can be added to the ’26 total.) Not bad — until you start poking around the figures from the same time in prior years. In the week of 12 January 2025, for example, nearly 72% of visitors were using some version of iOS 18, then the most recent. The week of 14 January 2024, over 65% were using iOS 17. iOS 26 adoption is fifteen to twenty points behind the uptake rate seen before. Not good.

I am irritated at myself for not thinking of using Wikimedia’s figures, which represent users across all versions of Wikipedia, Wikiquote, the Commons media library, and plenty of other widely used websites. I have relied on them plenty of times before for similar projects since they represent such a gigantic and general-purpose sample. Thank you to Tijhof.

Fine, have a third bonus update: I have been trying to get Chris Taylor at Mashable to correct his assertion — attributed to me — that the frozen version number in the user agent string of Safari in iOS 26 is a “bug”. It is a claim that appears in the dek (“a bug Apple won’t squash”), and a few times in the text (“there’s actually a bug in the reporting system, and it’s Apple’s fault”; “a tiny bug in Safari”). I told Taylor about the error, and he updated the article way down in the fifth paragraph, of seven total, to claim “it isn’t a bug, exactly”, which is a long and misleading way of saying it is not a bug at all. Taylor writes “for obscure techie reasons, as far as Apple is concerned, it’s a feature”, but does not elaborate or explain, which is kind of Taylor’s job as a “veteran tech … journalist”. I guess I am a little peeved to be cited for Taylor’s own error.