Hacker News Reader

Comment by brey | Hacker News Reader

brilliant concept ...

... but this could probably be the poster child of where not to use the Mercator projection.

yeah, Greenland overlaps five time zones. but they're tiny at that latitude! this makes it seem like places as far apart as New York & Los Angeles are in the same time zone.

edit: Miller not Mercator - but the point stands that it's distorting the size of countries at high latitude, significantly exaggerating the issue.

View on HN

It's not Mercator and I feel personally offended (it's Miller). I decided to use a equirectangular just because otherwise drawing the gradients would have been a hell.

ok, it's not Mercator, but it has the same high-latitude distortions

any projection where Greenland is the same size as South America is IMO the wrong choice :)

Relevant http://i.imgur.com/10sfTYY.png

http://mapfight.appspot.com/in-vs-gl/india-greenland-size-co...

^ companion knowledge

Unless the data you are conveying has nothing to do with relative area of different regions, like this map.

> I decided to use a equirectangular just because otherwise drawing the gradients would have been a hell.

If you've got a function that maps between map-position and globe-position, couldn't you just compute the gradient color for each pixel one by one? That doesn't sound particularly difficult... :]

I edited an existing SVG with emacs.

Miller projection is in the class of ideal projections for this data.

We're comparing time zones here. Instead of miles, this is a map of latitude against longitude. Using Miller, the longitudes are vertical, and that's the data being visualized.

(edit) I would have enjoyed seeing the Antarctic time zones as well. I suspect they're weird.

not sure I agree - a non-continuous projection such the Dymaxion / Fuller projection could indicate the same timezone errors without size distortion.

the Miller used here gives the impression that an utterly huge physical area - as with Greenland, appearing to be several thousands of miles from edge to edge - has just one timezone. this detracts from the point it's trying to make.

Antarctica would have been cool to see, yeah. I love the concept :)

How does this make it seem like places as far apart as New York and Los Angeles are in the same time zone? It's pretty clear that they're in different time zones.

the distance between the western and eastern tip of Greenland (which is one time zone) appears to be the same width as the continental US, but it's far, far smaller.

I don't think that's confusing anyone. Greenland appears to span 5 (hour-wide) timezones, because it does, and NY and LA are almost as far apart and span four (almost 5). This is exactly what you'd want to see in a map about the widths of timezones.

But Greenland is 1000km wide at its widest (east-west), which is 1/4 the distance from NY to LA.

We're talking about timezones. The whole article is about timezones. The map has no purpose except to show timezones. I cannot emphasize enough the point that timezones are the relevant unit of measurement in this article.

I think the point is that although it's only 1000km, the difference when expressed as a proportion of one day is the same. So noon in LA corresponds to late afternoon in NY, noon on Greenland's west coast corresponds to late afternoon on Greenland's east coast.

Greenland's span of longitude in degrees is greater than that of the continental US. That is what this map projection is displaying. Even though Greenland's distance is smaller, it's more longitude and time zones because the meridians are very closely packed near the pole.

How come? It seems clear they are in different timezones. The colors represent the difference between local solar time and local standard time.

Reply on news.ycombinator.com