[OSM-talk] "The Future of Free and Open-Source Maps" Slashdot.org , Saturday February 17, 2018

Jakob Mühldorfer mail at jmuehldorfer.de
Sat Feb 17 21:02:10 UTC 2018


Thanks for pointing it out to us!

I too have some thoughts on points in the article.
One I agree with, one not

Let me start with this one:
"No Support For Observational, or Other Datasets"
This is the point I agree with.
OSM is missing out on some valuable information due to this strict 
"verifiable on the ground" policy.
There is much more to a location, that the things that are currently 
there to see and easy to check.
I spent a long time browsing "atomic bomb text explosion sites" that are 
currently in OSM, a few people not interested in OSM were also really 
excited to see where they happened. But you would hardly be able or want 
to verify these?
Same goes for shipwrecks, they are accepted in the database, but how 
many people can first hand confirm the Titanic is really at this location?
I myself would also have liked to add a few locations or buildings used 
in movies, whoever dislikes such information could just ignore it.
Generally people who are experts in certain fields or have certain 
interest could generate a lot of highly interesting information, that 
others could learn from and that could also amaze "non OSM" people and 
get their interest in the project.

The thing I disagree with is
"OSM's Lack of Layers"
First concern is, that it would really restrict us.
How are you going to connect a highway to a building entrance? Have the 
entrance node on one layer, the building layer on another, the highway 
on a third, and then have some "relation style" super link across 
layers? Same for a ford on a highway over a river, and many more.
Second concern is, that it does not seem to be working in other maps 
that have it. Sure, it might more simple to write a few tools, but for 
example the map of the big G has illogical collisions of these layers in 
almost all parts of the world. Human editors would still need to work in 
the context of all layers to avoid these illogical "two features in once 
place" errors, and on a software level you would need a whole new set of 
complex validation software checking across all layers.

My thoughts on the issues



Am 17.02.2018 um 10:56 schrieb Oleksiy Muzalyev:
> This article is on the front page of the Slashdot today:
>
> Fri 16 February 2018 "Why OpenStreetMap is in Serious Trouble"
>
> https://blog.emacsen.net/blog/2018/02/16/osm-is-in-trouble/
>
>




More information about the talk mailing list