Mastodon Feed
slightlyoff@toot.cafe ("Alex Russell") wrote:
What sort of data would have cleared the bar? And how should an exception-based policy read? I might write a blog post on this, but the high points would be:
- observable (by measurement or proxy) interaction depth; i.e., how frequently do users or servers update the *same* data across a session?
- falsifiable claims about research into user devices and networks
- P90+ gates for latency across all critical user journeys
- ...which implies having defined what those journeys are