Hacker News Reader

I guess the revised explanation would be "random regression in ios 15 safari that affected AMP"

Why is this a regression in iOS? Changing the user-agent string is not a regression.

As in, a change/feature added to the rendering/javascript engine caused an regression.

Except AMP still loads in Safari 15 if you spoof the User Agent, per TFA.

"regression" doesn't have to mean "totally broken". maybe it's a little laggy. Maybe it breaks in certain edge cases.

I don't know that I would classify "rendering someone else's feature" a responsibility of iOS 15 or Safari.

Perhaps, although "it was just random" is arguably an even worse explanation.