[OSM-talk] Attribution guideline status update
nunocapelocaldeira at gmail.com
Sat Aug 10 17:33:20 UTC 2019
> For another perspective, imagine someone making a world map with 85%
> OpenStreetMap data and 15% XY inc. data, if someone looks on a part of
> this map which is fed by these 15% XY data, you would not want to have
> it incorrectly attributed to OpenStreetMap (although we are generally
> the principal data provider).
Well, the example i gave previously
https://janaodaparaabastecer.vost.pt/ is a good example of what you are
saying. What do you do to fix it? Mapbox will say nothing or "believe
this is the common, VOST won't say anything. Meanwhile 99.9% of that map
is OSM a the gas station status update is provided by Waze. Sounds fair
> I believe the 50% rule is ok, if it refers to the displayed objects on
> the screen (although this can also be arbitrary, since you can always
> split a way, or interpolate nodes to get more of them).
> Imagine a map which chooses a different data provider per country. For
> zoomed in maps (you only see data from one provider) you would want
> this one provider prominently attributed. If you attribute to someone
> else more prominently and show the actual data provider only in
> „others“, you will inevitably create a wrong impression about the
> source, and if it’s us who miss out on visible attribution, we should
Good that you mention this. On my email from 10th of October 2018 to
facebook and Mapbox (both stopped replying), i pointed out these
examples which have zero issues about having multiple sources being
attributed visibly and not hiding them:
> Microsoft - Uses HERE and OSM and attributes both visibly on the
> ARCGIS Web - Uses OSM and ESRI data, credits both
> European Commission - credits OSM and other sources
Sadly, some say this is hard to implement. The above sites, must have a
hell of a research UX dept to make it possible and others just say it's
hard. Google does the same on "dynamic attribution". It's not rocket
science, especially when it's for desktop use, there's plenty of space
to attribute visibly. It's just excuses.
> What about maps that display an overlay over a basemap? This would
> lead to the overlay data provider mostly being pushed in the second
> row because it is quantitatively less, but the overlay data might be
> the rare unique data that is interesting. In case someone displayed an
> OpenStreetMap based overlay over a different background, why would we
> deliberately renounce from attribution in these cases?
We shouldn't as it would violate the license.
> It is crucial that the 50% relate to the actually visible map
> features, and not to the total map. If the latter was possible, you
> could just fill your db with random crap in the middle of the ocean
> and distort the proportion.
Obviously, we know those dirty tricks. Fatmap is a perfect example of
and there's is plenty of room to add the attribution visibly.
To be honest i'm kinda fed up of all of this, nothing happens. And it's
a shame stating "the license doesn't say this or that", it neither says
you must attribute with the exact text “© OpenStreetMap contributors”,
must be unreasonable calculated to acknowledge. Common sense and
fairness is all needed, not crappy legal interpretations and placing
fear for legal actions from corporate interests. Sadly i'm starting to
believe the concerns that some have shared on the list that OSMF is
being "controlled" by corporate interests and not by the spirit that it
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk