I'm really confused by the common complaint in this thread about the arrogance of programmers wanting to change time for everyone else to make it simpler for th

I'm really confused by the common complaint in this thread about the arrogance of programmers wanting to change time for everyone else to make it simpler for themselves. No, in today's world time is a programmers' problem. Nobody except programmers (and some very technical software product managers, e.g. astronomers) needs to deal with time issues like these. Who builds systems that need to be synchronized to the microsecond level or more? Programmers. Who makes calendar software that should support timezones and DST correctly? Programmers. Who runs astronomy calculations that really care about UT1? Programmers working for astronomers. Every time someone has an issue that requires very delicate handling of time, there is a programmer implementing the solution. So of course it would be a group of the largest software companies, responsible for software like Outlook and GCal and public clouds, who would get together to say "listen, world, this leap second thing is more trouble than it's worth, how about we just stop fiddling with it?" They're the people responsible for implementing everyone else's time handling requirements. I think we should listen to them. Another confusing argument here is "let them just move from UTC to TAI, and thus move leap seconds to be a display issue, like timezone best practice". In a vacuum, this could have been a great solution. But we live in a world where almost all computer software already written is based on UNIX Epoch Time, which is defined as tweakable by the leap second committee, and the breakage usually occurs either in existing software or in sync issues between new software written at Meta or MS or wherever and existing software, so moving all their centrally managed systems to TAI while leaving us programmers of the less hierarchical outside world to keep suffering wouldn't solve much, and also wouldn't be very nice. Kind of an https://xkcd.com/927 https://xkcd.com/927 situation. Instead, they're writing a letter saying "dear leap second committee, we are your largest users and our use case is representative of everyone else's, can you please stop fiddling with the standard agreed time, it's causing problems for everyone. You can get the same effect by twiddling with display time (timezones), which we already have to support anyway and is much easier to test." My favorite design principle is the Representable/Valid principle: your representation/model of the world should have one way to represent each valid state, and should not have ways to represent invalid states. Lets get together as a species and refactor our model of time to have one less case of two states having the same representation! I for one welcome our new monotonic timelords.