Hacker News Reader

Comment by exabrial | Hacker News Reader

Yeah, that's my thoughts as well. The answer is literally in the article:

> Mercator useful for sailing

This is the only correct answer, and everything else is unscientific nonsense. It hasn't been replaced because... it just hasn't. Mercator + knots + nautical miles is a system that is extremely useful for navigation. Since we aren't sailing ships around a whole lot anymore, a lot of teachers forget to mention the relationship between the three. I imagine that was more common knowledge back in the day.

That being said, I do like the new proposed projection for how we use maps today. I think another useful tool would simply be overlay maps.

View on HN

> Mercator useful for sailing+

It is also used by the military (or at least NATO). They use https://en.wikipedia.org/wiki/Universal_Transverse_Mercator_...

While most of that system is a way to tag coordinates that are "simpler" than just latitude and longitude to a position, it's underlying system is also the Mercator projection. And some properties of Mercator, e.g. that it is angle-correct. So you can use UTM e.g. in Elint for direction finding.

It's not distance or area correct, but for normal infantry this is meaningless, since this only shows on large scales, but not the slow scales of infantry.

The transverse Mercator projection is completely different than what people are complaining about. The UTM system is roughly what you get if you only take a narrow band of the ordinary Mercator projection near the equator, but turn the globe 90°, and use the projection for a narrow strip of a specific longitude, with a different projection for each strip, overlapping at the seams.

It was adopted by the US military between WWI and WWII, and similar projections were adopted by other militaries around the same time, because earlier (non-conformal) map projections led to dramatic errors when they were used for artillery targeting.

Maybe to be more precise, conformal projections like the transverse Mercator projection and the conformal conic projection were adopted by militaries after WWI, and the UTM system per se was created during WWII.

Mercator is also useful for tiled maps, which is why web Mercator is a standard. The tiles always have nice north/south and east/west bounds and there is no difference in rendering depending on longitude or latitude. For software maps the better approach, as done by Google, is just to postprocess the Mercator tiles into a sphere at rendering time if the user is zoomed out enough.

Why not just use KM? We should reduce the number of units of measure in 2026.

Because nautical miles is an angle measurement which then happens to be 1852 meters when you break it down.

And only 1852m at sea level; since it's an angular distance it's a bit longer in the air and a bit shorter underwater (~ 29 cm per km of elevation change).

A nautical mile is roughly one minute of latitude. Every blue-water sailor has heard "a minute's a mile" -- although that's hardly even half-correct.

If you do positioning in degrees, minutes, and seconds (or decimal minutes which I guess is more common nowadays), a 'natural' length unit for navigation is a minute of longitude, which is the original definition of the nautical mile. And then we get velocity as nautical miles per hour, or knots.

Blame the French for reinventing the wheel instead of, say, using a 1/1000th of a nautical mile as the definition of the metre. Alas, that's all water under the bridge by now.

If you want to do navigation in SI units, the UTM projections (see sibling thread) typically do positioning in metres inside each grid square.

Funnily enough the metre is defined by the globe as well: the arc from the equator to North pole is defined as 10,000 km. I think this might be where the alternative angle metrics came from, which went with the metre:

full circle is 400 grad, so a right angle (10,000 km) is 100 grad.

Then minutes and seconds are factors of 100, so a centesimal minute is 1 km and 1 centesimal second is 10 m.

Originally it was defined like that, but it was fairly quickly redefined as the distance between marks on a prototype platinum bar. It's had a couple of other definitions since then, but it's currently defined "by taking the fixed numerical value of the speed of light in vacuum c to be 299 792 458 when expressed in the unit ms⁻¹, where the second is defined in terms of the caesium frequency ΔνCs".

https://www.bipm.org/en/si-base-units/metre

But that isn't a change to the definition of the meter. What do you think the motivation for the constant 299,792,458 is?

> that isn't a change to the definition of the meter

How is it not? it's not a change to the actual distance (well, it is for a sufficient precision), but a change to the definition of the metre is exactly what it is. It's defined relative to the speed of light rather than a random platinum bar length, with the constant picked so that the actual measurement stay the same (up to a certain level of precision).

Yes technically speaking there is a different physical reference now but that's where it came from. A bit like how 1 ft is now 304.8mm but that's because it's the size of a human foot (with some historic details no doubt).

Etymology vs dictionary, I suppose.

Reply on news.ycombinator.com