Hacker News Reader

Comment by boraoztunc | Hacker News Reader

>> The Good Design Grand Award, arguably Japan’s most prestigious award for design, went out to architect Hajime Narukawa’s completely groundbreaking design for a map of the world. His map, the AuthaGraph World Map, isn’t your average map of the globe: It’s a near-perfect representation of the continents and oceans as they exist on our spherical planet, all laid out on a two-dimensional surface.

https://totravelistolive.co/authagraph-world-map/

View on HN

I’m not a fan of AuthaGraph at all. It has horrible distortion look at Brazil or Alaska - and it’s proprietary (creator never released the actual map details, so you can’t make your own...) The worst thing, in my eyes, is that the distortion is unpredictable. Looking at a map where Canada is too big, you can mentally imagine Canada as smaller. But you can’t mentally undo the multi-direction transformation of Alaska into not-Alaska. The distortion is so bad that real world things like a road trip through South America would end up looking completely different from this map to another. You better hope you never get on a boat or a plane using this map - try tracing a straight line if you want a laugh. (The unfortunate distortion-hacking done by its creator effectively dumped the distortion in the oceans, so good luck if you ever go there.)

Design features like “you can do nifty origami with it” shouldn’t factor into if a map is used or not.

In my opinion, if you want something modern and cool, and you really care about pure accuracy, go with Waterman Butterfly. If you want something four sided and traditional, and relative sizes are slightly less important (but still important), go with Peirce Quincuncial - which, I might add, has far better looking Tissot’s indicatrices.

If one really cares about accuracy, why not getting a globe.

We all do : Google Earth. Is there still a point to 2D maps?

> Is there still a point to 2D maps?

Hiking away from cell towers and wifi

Historical maps

Maps that aren't owned by Google

> Hiking away from cell towers and wifi

Just download the map in advance.

> Historical maps

Agreed on this.

> Maps that aren't owned by Google

I agree with this in principle, but there are alternatives to Google Maps and some of them are even gaining popularity because developers in general are terrified of ever relying on Google for anything besides ads.

> Just download the map in advance

Okay, hiking away from cell towers and wifi and power supplies

My phone gets about 4 days of use currently before recharging, and I could pretty easily find ways to recharge it even using solar power if I wanted to.

If you don't want to use your phone as a map that's fine, but it's incredibly convenient for (I'd wager) 99% of use cases.

What about redundancy. You drop your phone and you are lost in the wilderness. You drop the map, you pick it up again.

I actually just came home from a hiking trip hours ago. I was supposed to meet up with a friend along the way. I had a proper map. He only had a phone. Some rainfall and a little bit of bad luck and he could no longer use his phone. Luckily, everything worked out in the end, after some unneeded trouble. But I wouldn't rely on a phone like that.

At least here hiking maps are much better than Google maps. (trails are marked and there are smaller paths visible)

Your criticisms are way too strong. Every 2D projection map of a 3D globe is going to make some compromises. So you can make the similar kinds of arguments for Mercator projection maps, which preserve shapes, but not relative sizes.

>You better hope you never get on a boat or a plane using this map - try tracing a straight line if you want a laugh.

Then don't use it for that!

>Design features like “you can do nifty origami with it” shouldn’t factor into if a map is used or not.

Maps are made for functional or aesthetic reasons or both. Why shouldn't that be a consideration? The pertinent question is: what do you want to use your map for?

>In my opinion, if you want something modern and cool

Again, what are you trying to optimize for?

"try tracing a straight line if you want a laugh" And on the Mercator projection, it too is also going to turn a straight line into a curved line. Draw a long enough line, and in a specific direction, it'll make an 'S' curve.

Personally, my favorite map to look at is the Goode homolosine projection. Also difficult for the usecase of navigation across oceans, but among continental comparisons, general shape and size, it's useful. Not all maps need to necessarily be used for navigation aeronautics, or seafaring.

It’s a near-perfect representation of the continents and oceans as they exist on our spherical planet, all laid out on a two-dimensional surface.

This is a bullshit statement. You cannot project the surface of the earth onto a flat plane without distortion. AuthaGraph favors the preservation of size at the cost of extreme distortion of shape.

Here is an interactive tool, that lets you center other projections on different countries, to see the effect of the distortions move about:

https://engaging-data.com/country-centered-map-projections/

From playing around with that, it would seem the design process for the map was to "center" japan, then fiddle around with the polar regions to make them more recognisable.

This projection seems to really skew shapes. For example, Alaska looks skewed diagonally, Brazil looks really stretched east-to-west, and the west half of Australia looks too big compared to the east half.

Mercator is what you get when you preserve local angles (ie shapes). So it's only natural that moving away from Mercator would distort shapes.

I guess you could try to concentrate your angle distortion into the ocean away from the land masses that people care about?

I prefer the (related) Dymaxion approach myself, but I'll admit it's not quite as accessible.

I really like the Dymaxion projection as well, especially the fact that one can easily see that all major land masses are separated by only small distances of water.

https://en.wikipedia.org/wiki/Dymaxion_map

Dymaxion map is neat, but I don't think human brains are very good at handling interrupted maps and Dymaxion is particularly challenging. It is just pretty difficult to mentally intuit how the edges connect, and especially judge distances that cross the edges. Trying to eyeball something like Sydney-Lima great circle path on Dymaxion is unintuitive.

Of similar interrupted style, I think butterfly projections are even better. They might distort land masses slightly more, but are immensely more readable. And they have very pretty symmetric outline.

Even better, a close relative of it: https://en.wikipedia.org/wiki/Cahill%E2%80%93Keyes_projectio...

Very neat. And he didn't forget New Zealand[1] =)

I wonder how much the Mercator projection influenced the recent flat earth movement, specifically the Antarctic ice wall.

[1]https://www.reddit.com/r/MapsWithoutNZ/

I far prefer https://en.wikipedia.org/wiki/Cahill%E2%80%93Keyes_projectio...

Minimal distortion, preserves east west alignment as much as possible.

That’s a fascinatingly done projection. It really demonstrates how it’s possible for Maine to be the closest US state to Africa.

Really cool. The shapes of the contents and countries seem like what I am used to. Except for Brazil.

Brazil really looks like it bulges out from South America compared to what I have seen on globes and other projections.

Does anyone know why that happens?

Any comparison of map projections should include Tissot's indicatrix [0] as a visualization of how distortion varies across the map. I find that it's easier to understand than looking at stretched and bent lines of latitude and longitude.

For AuthaGraph: https://kunimune.home.blog/2017/11/23/the-secrets-of-the-aut...

[0] https://en.wikipedia.org/wiki/Tissot%27s_indicatrix

The map also follows the tradition of putting one's own country at the center

Reply on news.ycombinator.com