Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mapcode – A short address for any location on Earth

mapcode.com

Post reply on HN
Print view View original

111–120 of 124 posts

Re: Mapcode – A short address for any location on Earth

#111

Post by jxf »

jofer wrote: post #109
jxf wrote: post #22

I'm a little confused about what this replaces, but it's definitely not postcodes. * Postcodes have a specific meaning: they're both a geographical area and a mail delivery zone (at least in the US). Replacing one short code with another, which requires a different and less reliable third-party service to correctly decode, doesn't seem like a good idea. * Latitudes and longitudes are unambiguous, well-understood, and…

> "Latitudes and longitudes are unambiguous..." This is a tangent, but actually, they're not. There have been (oil) wells fail because someone gave a lat/long without specifying the datum, and someone else assumed it was in a different datum. Without a datum defined, latitude/longitude (no matter how precise) only gets you to within ~1km of an actual location. The Earth is not a sphere (or even an ellipsoid). Therefo…

You're absolutely right, and I should have said that!

I was trying to contrast to the case of just typing it into some kind of map service, i.e. MapCodes. A lat/long pair unambiguously resolves to a single point in that service.

Re: Mapcode – A short address for any location on Earth

#113

Post by mseebach »

Shish2k wrote: post #46

> The mapcode foundation reserves the right to change the country list and population density maps, at which point existing codes will point to different places or be invalid Nope.jpg

From the docs:

"[This] issue should be prevented at all costs. The second [adding new countries] may be necessitated by changes to the world order, i.e. new countries coming into being. But unless changes are made by a single central authority, the first issue will quickly come to be."

Re: Mapcode – A short address for any location on Earth

#115

Post by mxfh »

jisaacso wrote: post #29
grey-area wrote: post #14

Objections: 1. This mapcode is country specific (!) 2. Many locations need altitude or fine lat,lon differences to differentiate them (e.g. blocks of flats, offices). 3. What's wrong with altitude, lat and lon on a given projection which can be as precise/imprecise as you want? IMO it'd be better to have a hash of lat,lon,altitude which was globally unique for each level of precision. A system like this needs to be g…

There exists a hash of lat, long globally unique for each level of precision, a geohash [1]. One common application utilizes a space filling Hilbert curve to define an arbitrary precision bounding box around a set of coordinates. You can see an example hash [2]. [1] http://en.wikipedia.org/wiki/Geohash [2] http://geohash.org/9q8yyvfekk8h

I'm pretty certain one could make a geohash compatible partly human readable hybrid that performs way better than this proprietary obscurity.

Something like DE.BERLIN.2.1H

First, human readable part defines a look-up to defined geohash of center of region on a given level.

So DE.BERLIN could resolve to u33, u33d or u33db (length/level 5) the last one seems reasonable, since a few multiples of the area cover the whole city of Berlin

http://geohash.2ch.to/u33db

Second, optional, part encodes offset of Z-Curve index of location at same zoom level.

Z-Curve on globe: http://www.bigdatamodeling.org/2013/01/intuitive-geohash.htm...

This enables global coverage at any level if needed, and makes the code robust to administrative changes that affect the area of the given region! The offset part could start at 1 and alternate the sign with each increment.

isOdd(offsetPart) ? -(offsetPart-1)/2 : offsetPart/2;

1 → 0, 2 → 1, 3 → -1, 4 → 2

An offset part of 1 equals an offset of 0 and could be omitted.

The third part adds to the precision of that given cell.

Brandenburg Gate, within area of origin cell of region part. could be written as DE.BERLIN.1.2M3 or short DE.BERLIN.2M3

http://geohash.2ch.to/u33db2m3

DE.Berlin.2.1H would be the area and center point of the Berliner Fernsehturm with sufficient accuracy u33db + 1 = u33dc

http://geohash.2ch.to/u33dc1h

This example uses base 32 to comply with geohash but other more error robust encodings and variations (e.g. second part as base ten numerals) could be used as well.

Add optional checksum part based on unshortened geohash if needed to verify integrity.

Variants: since geohash, covers a varying amount of area depending on the latitude. A similiar approach could be realized on a UTM based grid like MGRS if a precise area is required (https://en.wikipedia.org/wiki/MGRS)

Re: Mapcode – A short address for any location on Earth

#116

Post by jimktrains2 »

There is still a special place in my heart for UTM [0].

MapCodes seem to be based on a similar concept (grid, sub grid, sub-subgrid), but with a bit shorter representation. The thing that worries me about using mapcodes, however, is what seem to be the arbitrary borders, order of girds, and the up-keep required for usage. [1]

There is nothing to stop us from creating shorter UTM Codes, though.

Assume 0->33 represented by

    0         0         0         0
    ABCDEFGHJKLMNPQRSTUVWXYZ0123456789
Same as Map Codes [2]

The CN Tower is at

* WGS84: 43°38′33.24″N 79°23′13.7″W, (26 chars)

* UTM: 17N 630084 4833438 (18 chars)

* "Shortened" UTM: T.N.SBB6.DW9F8 (13 chars)

* Map Code: CAN PYLK.XY8P [3] (13 chars)

So I guess the utility of such a complex system is lost on me when something like UTM, that can be decoded with minimal effort, is just as good.

[0]: http://en.wikipedia.org/wiki/Universal_Transverse_Mercator_c...

[1]: http://www.mapcode.com/mapcode_documentation.doc

[2]: http://www.mapcode.com/alphabets.html

[3]: http://www.mapcode.com/getcoords.html?iso3=112&ifrom=aboutmc

Re: Mapcode – A short address for any location on Earth

#117

Post by Yardlink »

jimktrains2 wrote: post #116

There is still a special place in my heart for UTM [0]. MapCodes seem to be based on a similar concept (grid, sub grid, sub-subgrid), but with a bit shorter representation. The thing that worries me about using mapcodes, however, is what seem to be the arbitrary borders, order of girds, and the up-keep required for usage. [1] There is nothing to stop us from creating shorter UTM Codes, though. Assume 0->33 represente…

If you use CAN-ON it's only 12 chars. But if you already know you're in Ontario it's only 5 including the dot.

So that makes the difference between saying it to someone and having to write it down. Mapcode is massively shorter for local locations which is exactly what sat-nav is for.

The "grids" for mapcode are human-understandable - country and state.

I think a lot of HNers are thinking too much like programmers and forgetting that people need to tell each other locations and write them down. That's where mapcode is good.

Re: Mapcode – A short address for any location on Earth

#118

Post by blahedo »

gkoz wrote: post #72

The choice of Cyrillic translations seems arbitrary to me. There is a well known Morse code convention (Q=Щ, V=Ж, W=В, X=Ь, Y=Ы).

The problem with e.g. W=В or X=Ь is that they want glyph matching; so the roman alphabet B and the cyrillic alphabet В (even though they are different letters) map to each other, to avoid having same-looking codes map to different locations. They're solving a slightly different problem than the Morse convention was trying to solve.

Re: Mapcode – A short address for any location on Earth

#119

Post by jarek »

jimktrains2 wrote: post #116

There is still a special place in my heart for UTM [0]. MapCodes seem to be based on a similar concept (grid, sub grid, sub-subgrid), but with a bit shorter representation. The thing that worries me about using mapcodes, however, is what seem to be the arbitrary borders, order of girds, and the up-keep required for usage. [1] There is nothing to stop us from creating shorter UTM Codes, though. Assume 0->33 represente…

The CN Tower is at M5V2T6 - 6 chars

This is unique within Canada and the physical area is small enough to be sufficiently useful for navigation. Add country code and it gets you to 9 or 10 chars.

UK has similarly useful alphanumeric postcodes and specifying them for satnav is the norm.

Re: Mapcode – A short address for any location on Earth

#120

Post by chippy »

fixermark wrote: post #76
chippy wrote: post #68

Earlier quoted context omitted.

OpenStreetMap's better in that you can correct it yourself or leave a note for someone else to fix

That's possible with Google Maps also ( http://www.google.com/mapmaker ) But thank you for noting that I can do the same with OSM. After making an account and logging in, I've created a feature-point for my address, but I don't see a way to correct or suppress the incorrect information from OpenStreetMap Nominatim. shrug Problem for another day.

Hmm, I think that Nominatim gets updated with a bit of lag (minutes to an hour), so the address should be there now? Assuming you changed the information from the map, it should update it within Nominatim, as it get's it's info from the map

If not, you can leave a note http://www.openstreetmap.org/note/new and someone can take a look at it. I've been contributing for a while now and still like using Notes to let people with more techy experience or local knowledge help.

Post reply on HN