Rapid’s Data Integration Capabilities

In reply to a split conversation:

This is really getting off-topic. Microsoft building detections aren’t what people normally mean when they say, “Don’t copy from other maps.” We could be talking about one of the high-quality, LiDAR-derived building datasets also listed in Rapid and it would still be off-topic.

The primary reason for the rule is the risk of violating intellectual property rights. For example, you could violate the rule by:

  • Scanning in a copyrighted topo map, running a feature detection algorithm over it to detect buildings, and piping all 10,000 detections into upload.py (clearly an import, clearly problematic)
  • Scanning in a copyrighted topo map, setting up a WMS, loading it as an overlay, and meticulously tracing a single mile-long abandoned railway from it at zoom level 21 (not quite an import, but still problematic)

The original thread and the wiki page shouldn’t turn into a general discussion of OSM threat vectors, because novice mappers will have even less clarity about what they should and shouldn’t do.

4 Likes