subtoot ranting about random thread i stumbled across
Dummies who think that because adding two numbers that are both between 1 and 24 is too hard and because programming time zones is too hard that we should forcibly change the entire world to some random time scheme (but not UTC because BRI'ISH IMPERIALISM, despite UTC being the closest we have to a universal time).
You would think that they'd realize that saying “let's have lunch on the 26th next month at uhh… gotta do calculations to figure out when midday is” is just a minuscule bit harder than “let's have lunch at 12:00 on the 26th next month”
While “let's set the international meeting at XX:XX DAT (dumbass time)” is no different from saying “let's set the international meeting at YY:YY UTC” or “person has set a meeting for ZZ:ZZ” (automatically converted to local time by the computer)
subtoot ranting about random thread i stumbled across
@nytpu
Timezones in principle are fine. What complicates things and what should be abolished are:
1. Multiple observed offsets, which include that of DST (daylglight saving time). Certain jumps in offset result in seemingly valid but actually ambiguous or nonexistent datetimes. For example, in "Europe/Amsterdam" on 1995 March 26th, the local time 02:30 was never observed, whilst on September 24th, 02:00 was observed twice (02:00 CET and 02:00 CEST).
2. Certain offset changes result in some timezones skipping entire days! In Samoa "Pacific/Apia", on December 29th at late midnight, their offset changed from -10 to +14, skipping December 30th entirely.
3. Speaking of midnight, some politicians are dumb enough to change offsets on midnight, which causes some midnights to be observed twice. The question is, how do you calculate the start of a day? Bugs bugs bugs.
I have previously worked on a datetime library. There's many more cases to cover. Leap seconds between UTC and TAI (International Atomic Time) complicate things even more, which means certain UTC datetimes aren't even valid.
