iD users adding crossing:markings=yes to crossing=marked?

I have noticed a few instances of iD users adding crossing:markings=yes to crossing=marked nodes.

The problem with this is that users are effectively adding a tag to confirm that they didn’t bother to look at the aerial imagery to determine what sort of crossing is present here and just ticked a (low) quality assurance box. Knowing whether a marked crossing is a zebra crossing or one with traffic signals is useful for data consumers, knowing that a marked crossing may be marked somewhat less so.

If iD suggested or only allowed adding values other than “yes” based on actually looking at the available imagery, which was presumably the intent, this might have been quite helpful.

2 Likes

I also find this a strange warning. Now when I want to contribute in crossing markings tagging I have to search not only for !crossing:markings=*, but also for crossing:markings!=yes.

1 Like

crossing=zebra? :popcorn:

2 Likes

Crossing:markings says nothing about whether there are signals. It is all about the kind of markings. So the value of Yes adds nothing over a crossing=marked – Mapper time is for free, keep them busy?

Presence of signals not always easily observed from aerials. Crossing:signals in need of a kind person to draft and go through all the OSM bureaucracy?

For Bing imagery in the UK at least, the stop lines of traffic signals are usually easy to spot, then crossing=marked can be replaced with a much more useful crossing=traffic_signals. If they’re visible in the imagery, crossing:markings=dots are usually present. I generally add crossing:signals=yes just in case iD’s developers decide to “upgrade” crossing=traffic_signals to something unhelpful like crossing=crossing at some point in the future :slight_smile:

1 Like

A bit unlikely, traffic_signals only got in very recently, Invisible crossing=traffic_signals · openstreetmap/iD · Discussion #8866 · GitHub

You might post your issue in their tracker?

I was being a little facetious, as iD also facilitates tagging highway=traffic_signals with traffic_signals=signal i.e. “this traffic signal is a traffic signal”, providing no useful additional information whatsoever. We’ve now got several of the crossing:= tags (and crossing_ref=* in the UK) in part because crossing=* cannot always be trusted to have a sane or consistent value.