Proposal talk:Tag:highway=unknown
Difference between highway=unknown and highway=yes?
I think *=unknown is generally bad practice, and I'd suggest that highway=yes (which currently has more uses, 222, and mirrors other uses of *=yes) is a better alternative as it clearly defines a way as a highway open to iterative refinement.
To me, highway=yes reads as "This is a highway." while highway=unknown reads as "Is this a highway?" Lumikeiju (talk) 18:26, 22 September 2024 (UTC)
- If we don't know if it's really a highway, I think it's better to use fixme=* or not map it at all.
highway=yes seems confusing, highway=yes can mean that we know it's a highway but we don't know what road classification to use. — Koreller (talk) 20:10, 22 September 2024 (UTC)
Open to better names, but as @Koreller pointed out, highway=yes confirms that this is a highway while It may not be (e.g mapped from imagery or GPS) julcnx (talk) 01:03, 23 September 2024 (UTC)
How many renderers don't handle this?
I've seen mention in passing that many renderers treat unknown "highway=" values as "highway=unclassified", meaning this proposal will have little to no benefit. Has this been confirmed not to be the case? --Carnildo (talk) 19:22, 22 September 2024 (UTC)
Check for yourself the Rendering Support section in the highway=road wiki page. Any Mapbox based providers and many others which are missing from that list. julcnx (talk) 01:05, 23 September 2024 (UTC)
2 Contras
I have 2 observations in oppossitionto this proposal. 1) Don't map for the render. If a render can't handle some tags in a correct way, must fix the render, not the DB. It's a very bad habit to force the tagging to get nice renders. 2} highway=unclassified appears to cover perfectly the issue raised, and getting more tags for similar situations, is bad for the ecosystem, due confussions in what tag must be used and dispersing the data coherency (fragmentación) Camello AR (talk) 20:10, 22 September 2024 (UTC)
- See highway=unclassified for meaning of this tag (yes, it is really unfortunate value name) @Camello AR: Mateusz Konieczny (talk) 20:39, 7 October 2024 (UTC)
renderers
Breaking rendering reasonably displaying OSM data because some renderers display is badly is really going into wrong direction Mateusz Konieczny (talk) 20:44, 7 October 2024 (UTC)