Flaw: Name of major street will be suppressed by the one of a alongside minor street


#1

Please take a look at this cutout:
https://maps.openrouteservice.org/directions?n1=50.721351&n2=7.063156&n3=15&b=0&c=0&k1=en-US&k2=km

About in the middle there is a street called “Schleidener Straße” (50.720923, 7.06371). If you zoom in, it is apparently that this is a rather minor street compared to the alongside “Provinzialstraße”. I recommend to adjust algorithms in a way that in doubt a street like “Provinzialstraße” supersedes “Schleidener Straße” – but not, like currently, the other way around.

Thanks, Peter


#2

Sure, you’re right. But you should consider how AMAZINGLY complex mapping is. The rules involved for what to draw on top of each other at what zoom levels is a whole science. Taking one tiny example might strike you, but in the grander scheme it’s likely to be an exception. But we’ll let the developer know, maybe it really could be smth easy to fix.

Just to brag (with other people’s work;)): the mapsurfer algorithm is hands down (one of) the most sophisticated in the world (far better than any ESRI, Google or Apple crap). To get half an idea, consider the labeling in the Caribbean and compare to other map providers:
https://maps.openrouteservice.org/directions?n1=16.161921&n2=-64.182129&n3=6&b=0&c=0&k1=en-US&k2=km

BTW, also Germany’s official ministry for mapping makes the same “mistake”:
https://maps.openrouteservice.org/directions?n1=50.721406&n2=7.064638&n3=16&b=0&c=0&k1=en-US&k2=km


#3

I just thought that the developers are collecting known problematic spots – similar to audio-codec developers, who have a collection of known problematic audio samples, known for being brainteasers for the coder.

Regards, Peter


#4

I can be a little over-protective over our “products”. You’re definitely right. Just wanted to emphasize the complex nature of generating smth as ubiquitous as maps.