Swatch internet time, measured in ".beats", was an idea for a common and practical way to define time in useful units independent of time zones. It never really took off.
tomasino recently mentioned it in a
which again got me thinking why I didn't like ".beats".
There are two basic flaws with it, in my opinion:
".beats" are not based on UTC but on UTC+1, or even "Biel mean time" as the (IMHO) slightly megalomaniac Swatch PR guys decided to call it. As UTC is already the time base for most interesting stuff (nautics, telecom), it's totally stupid to not use it. For me, that was the killer.
If already it was intended to be used all over the planet, one should have thought of the difficulty concerning the day: ".beats" only cover 24 hours, and therefore it's not clear which day is meant, if I tell somebody in the Pacific ocean I'd like to meet online next Tuesday @654. "Their" Tuesday or "mine"? One could easily have prevented this ambiguity by prepending ".beats" with the day of week or day of month also in UTC. "Next Tuesday @654" would then be written "2@654" if we set Sunday=0, Monday=1, Tuesday=2, etc. (Just noting day of week would keep it short by only adding one character to the string.)
With these modifications, I'd immediately accept Swatch internet time and advertise it, and I'd even use it for my amateur radio logs and other private stuff, because time resolution of 1/1000 day = 86.4 sec (or about 1.4 min) is way sufficient for most interactions.
I'll probably add the modified version to my CGI time script,
if I may steal tomasino's idea!