Hacker News Reader

Comment by fs111 | Hacker News Reader

It's a business, so they update the maps, where it makes the most sense from a business point of view. Making quality maps is a very labor intense task and looks simpler from the outside, than it actually is. google for instance has more people working on their maps than TomTom has employees in total (which is betweem 3k and 4k). That's how intense it is...

(full disclosure: I worked in TomToms maps division for 7 years).

View on HN

So I think you're actually agreeing with the OP. "where it makes the most sense from a business point of view" is a subset of "global".

Your point about the amount of effort to curate map data is well taken, too. But doesn't that mean that a crowdsourced approach makes sense? It might be possible to build a data set that no one company could afford to pay to build.

It makes sense, where the data is available to you, but modern maps contain attribution that has to be licensed from 3rd parties(RDS/TMC codes for instance). You cannot include those, since you don't own the copyright.

Another example was the 7-digit postal codes in the UK: Before the UK did this big open data push, those had to be licensed as well.

I am all for open maps, but some things are just impossible to do legally with OSM.

Every map is a subset of global - it's inevitable that there are newly built roads, private roads that mapping companies can't reach, dirt tracks that no-one has reported as missing from maps, temporary roads, roads that can't be seen in aerial photographs and so on. OSM also has omissions in areas where there aren't any OSM volunteers nearby to do the mapping.

The question is what distribution of omissions is acceptable before a map ceases to be global :)

Reply on news.ycombinator.com