iD editor validation rules in Norway

eisa01 · 2024-04-14T10:54:49.852Z

The iD editor has a lot of validator rules, but not all of them may be applicable to Norway as we saw with cycleway. Those warnings will now be disabled for Norway

Are there other warnings we should turn off because we don’t intend to fix them?
Crossing has incomplete tags: Using crossing:markings=* schema?
Bus stop has incomplete tags: Adding bus=yes and public_transport=platform (also applies to other forms of transport)

It is currently quite hard to see the real warnings in the map that may be present, e.g., disconnected foot paths from the road network.

If we don’t turn them off, we could consider mass edits to upgrade the tags? This could be the case with e.g., the missing wikidata brand tags (with appropriate upgrade of the import scripts as well)

NKA · 2024-04-14T16:11:33.728Z

Those two + bicycle=designated for highway=cycleway are the most frustrating cases. Would be great to have them removed from validation in iD. We already have had community discussions on them.

Wulfmorn · 2024-04-14T17:35:48.516Z

I want to keep the crossing:markings. I think they’ll likely matter for accessibility (mainly for reduced vision) and its a better schema than the alternative.

harahu · 2024-04-15T14:26:54.151Z

Agreed. I also make use of the crossing:markings=* warnings and would like them to stay.

NKA · 2024-04-15T14:55:11.822Z

The problem here is that if you open “Issues” in iD over a town, you will be met with hundreds of error messages for Marked Crossings and Unmarked Crossings. It is difficult for a user to discover any other real issues, so this feature in iD becomes almost unusable. If we keep Crossing validation activated for Norway it will stay that way, and a lot of needed fixes will not be done.

If we deactivate validation, crossing:marked=* will still be a highly visible attributes for anyone editing those features in iD.

And keep in mind that we have no consensus on this tag in the community. It does not add any information on top of crossing=*.

eisa01 · 2024-04-15T15:22:28.200Z

Let me first check if there’s a distinction between a preset and validation. If we theoretically keep the preset for when you select e.g., a zebra crossing, but not have the validator spam warnings

edit: The presets and validation can be independent

I have no clue what’s best with crossing:markings, but noticed a lot of them is that unmarked crossing need crossing:markings=no added.

Is this something we can do a mass edit on?

Wulfmorn · 2024-04-15T16:23:41.539Z

I want to build a consensus for crossings, but I’m not ready to begin yet. I have a significant project that could impact and be impacted. If we can just skip all decisions on crossing for now We can get back on the topic at a later date.

I fully agree about the bus_stop and cycleway issues. For iD and StreetComplete.

NKA · 2024-04-15T16:29:04.033Z

eisa01:

I have no clue what’s best with crossing:markings, but noticed a lot of them is that unmarked crossing need crossing:markings=no added.

It simply does not make any sense to add crossing:markings=no to crossing=unmarked …

harahu · 2024-04-15T20:34:27.370Z

One can view the crossing:* tags as supplementary to the crossing tag, or one can view them as a competing schema that can eventually replace the latter. If you think of it as a competing schema, adding crossing:markings=no to crossing=unmarked clearly makes sense, even though they represent the same piece of information.