Hacker News Reader

I'm also curious if anyone has a good article to read on how this is managed. My guess would be that you have reference points and sets of nearby things that are recorded relative to the closest reference point. As noted in the sibling, your plat map is likely set against a local landmark. (Though, even there, they often also include the lat/lng of where it is expected to be found.)

View on HN

This is a good example of why capturing coordinates in a local spatial reference system may be better than capturing coordinates in a global spatial reference system.

Global coordinates of a place of interest will drift relative to a global coordinate system but remain stable relative to a local coordinate system. This article illustrates that "stable" is not the same thing as unchanging in the presence of 7.5 earthquakes.

It’s similar to capturing future dates in local time including timezone. There’s no guarantee that 1300 london time on 20-Dec-2029 will be 1300UatC, as the U.K. may shift its daylight rules before then.

That depends entirely ok whether the event being reflected is local or global. If its the time of an solar eclipse, for example, then its unaffected by daylight rules - unlike business hours and such. Working with dates/timestamps is a nightmare, but appears simple at first blush.

Reply on news.ycombinator.com