Hacker News Reader

Comment by lmm | Hacker News Reader

> It has to be more often than leap seconds to really get those gears oiled. Unless this is a proposal to do leap deciseconds, I don't think this method reduces the pain.

Yeah, if we're not going to abolish leap everything in UTC completely (which I think we should) then leaps need to happen at most quarterly. When the last leap second came through it was long enough since the one before that the bugs in Linux had been fixed and then unfixed.

View on HN

Do we introduce an unnecessary leap second and then negative leap second every year just to keep the gears oiled? And then if we need one for reals we can increase it to two leap seconds added and only one removed. Though of course that will cause bugs in any code that isn't expecting two leap seconds to be added and assumed that there would only ever be one. sigh

We could build it into daylight savings time. Instead of jumping an hour, we announce what the jump will be every year.

Not everyone's jumps happen at the same time (or ever).

We already have non-integer timezone differences. We could allow for leap seconds in that. Europe's timezone could be UTC+2:00:01 while the UK is still UTC+00:00:00

That has all the same legal difficulties as removing leap seconds from UTC (indeed it sounds like you would remove leap seconds from UTC).

Just accept that time will very slowly drift over the period of many centuries. It's really not a big deal. It's a complete non-problem that doesn't need any solving. Especially since coming up with a robust scheme that will work over the period of many centuries is quite hard (the current leap second system doesn't suffice).

> It's really not a big deal.

Unless, of course, you are a satellite.

I don't know what you're referring to, but as far as I know satellites can work perfectly fine without a leap second. I'm not aware of any practical advantage of leap seconds anywhere. The only difficulty in abolishing the leap second is that some systems have it incorporated in their protocol, but that's a protocol issue and not an intrinsic time drift issue.

Time is attached to the orbit of the earth (very precisely) so when time goes out of sync of the earth, it won't be a good day for anyone.

Most satellite systems don't use leap seconds; you don't need leap seconds to have satellites. And leap seconds don't really "sync time with the earth"; that's not how any of this works. Whether it's now 13:04:00 or 13:04:27 is somewhat arbitrary and all leap seconds do is maintain a certain arbitrary definition.

Orbiting around earth is going to be based on sidereal time, which drifts 4 minutes per day.

Converting into normal units has no reason to care about leap seconds when it can use the actual offset and be a few orders of magnitude more precise.

It vaguely feels like it ought to happen yearly. Yearly, but no option of a shift of exactly zero, so everybody knows it is coming every year no matter what.

Reply on news.ycombinator.com