[OSM-talk] Macromapping problems

Daniel Koć daniel at xn--ko-wla.pl
Tue Aug 22 21:17:41 UTC 2017

W dniu 22.08.2017 o 22:29, Martin Koppenhoefer pisze:
> it's not easy for forests or similar things either, because they might be split (for good reason) into several smaller objects. What you need for good macro maps is generalization, i.e. merging together of similar things, reduction of unwanted detail while keeping the characteristics and significant features and shapes. Nothing we should encourage our mappers to do in the main db (at least as long as we have one scale fits it all data), it must be done parallel or with locally processed data.

In Europe (but also in Japan, India and in Canada) there are many 
smaller forests, not just one big woodland, and you can still see them 
as a big area (see for example French style), so I don't think we would 
need generalization in other parts of the world too.

Sahara not only doesn't have a name in OSM, but it's also evident for me 
that most of its parts are missing. Even if French style shows 
natural=desert and not natural=sand (on the osm-carto we do the other 
way around), both types combined are far from covering the whole area.

"Probably it's an eternal problem - too many chiefs, too few Indians" [O. Muzalyev]

More information about the talk mailing list