3 ms·
DST doesn't exist in UTC, so that's irrelevant. A one-hour UTC shift would totally, utterly, screw stuff up. But, at the current rate, a one-hour "leap" would h
by MichaelGG 14y ago
DST doesn't exist in UTC, so that's irrelevant. A one-hour UTC shift would totally, utterly, screw stuff up. But, at the current rate, a one-hour "leap" would happen in thousands of years, so maybe it's not such a bad idea after all. But I think the reason for leap seconds has to do with keeping UTC in sync with other clock systems, and that probably overrides any inconvenience to software.
- moe 14y agoI wonder why they don't implement the google solution on pool.ntp.org. I.e. gradually slow/accelerate time over the course of a day, rather than stepping it hard at once. I'd say this approach would be vastly preferable for about 100% of the systems relying on pool.ntp.org. The remaining 0%, e.g. scientific applications that absolutely need the leap second to appear at exactly the right moment, most likely don't use pool.ntp.org anyways. And for those who do they could create a second pool with the old behavior. Maybe call it science.ntp.org.
- mikeash 14y agoA one-second UTC shift seems to be pretty good at totally screwing stuff up already. At least a one-hour shift would happen once every few centuries instead of once every few years.
- Negitivefrags 14y agoThe solution is not to change UTC, but just to rotate the time zones of each country every now and then. DST has proven that a country is able to change time zones twice a year. This would happen far less often.