In addition,

• sidewalk=no is the tagging used by Vespucci
• sidewalk=no is the tagging used by StreetComplete (which was recently updated in Change no-sidewalk tagging to =no streetcomplete/StreetComplete#3194). Note that the prior use of sidewalk=none in StreetComplete was likely the cause of the recent 100K tag increase of sidewalk=no; the tagging schema change here will help maintain consistency across the community of OSM editing software.

@ZeLonewolf Have you opened or found equivalent JOSM issue already? (at https://josm.openstreetmap.de/)

@ZeLonewolf Have you opened or found equivalent JOSM issue already? (at https://josm.openstreetmap.de/)

@matkoniecz it appears that JOSM is already using sidewalk=no, which is consistent will all the other editors. So it looks like no change is needed in JOSM. Here's a screenshot I took of the the street dialog with the sidewalk pull-down menu opened:

I was thinking about equivalent of this one - validator asking to replace none by no

Good point! I made an issue in JOSM's issue tracker:
https://josm.openstreetmap.de/ticket/21242

The wiki now https://wiki.openstreetmap.org/wiki/Key:sidewalk#Deprecation_of_sidewalk.3Dnone has this to say:

[…] the deprecation of sidewalk=none in favour of sidewalk=no, though retagging is not always supported. […] Please note that updating none to no without manual review should be done in compliance with the Automated Edits code of conduct.

I think that this means that we should still keep this on hold for now.

Btw, here's a recent usage comparison:

It seems clear that the growth spike of sidewalk=none was entirely fueled by SC mistake and once it went away it went to being dead again

http://taghistory.raifer.tech/#***/sidewalk/no&***/sidewalk/none

@tyrasd What else do we need here? Are you expecting formal deprecation proposal before adding this deprecation rule?

In my opinion tagging situation is clear enough in this case

sidewalk=no has gotten approximately 100k more instances over the last 4 months while sidewalk=none is pretty much stagnant. I don't think that there is any reason to delay adding this deprecation rule.

Going by #401 (comment) - though it is a bit different one, as it is about no longer promoting tag rather than active deprecation

until we have at least one of the following:

the tag's status is deprecated on the wiki
there is significant drop in usage compared to the current numbers and negative trend
usage of the tag remain stagnant for a longer time (~1 year)

Stably listed as deprecated at https://wiki.openstreetmap.org/wiki/Key:sidewalk#Values

http://taghistory.raifer.tech/#***/sidewalk/no&***/sidewalk/none shows completely flat organic use (while sidewalk=no use increased by 500 000 and doubled), with some drops from bot retagging

new organic usage is not there at all

I also considered proposing worldwide bot edit retagging remaining cases.

I have actually unwittingly submitted this change as part of a larger PR with typos for sidewalk tags: #1278. I'd be happy to remove the none→no deprecation from my PR in favor of this one.

In any case, I thought I'd add my voice to the comments above and the general sentiment in the other discussions linked from the wiki, which IIUC pretty much all agree with deprecating none as long as it's explicitly documented as such (which has been the case for almost 3 years now).

As an update, here's what the tag history graph looks like today, also including the :left and :right subtags (note how even sidewalk:left=no has overpassed sidewalk=none by now 🙂)

@waldyrious no pride in ownership here. If your PR is merged, I'll happily close this one as complete.

I agree, we should merge this. It is on the edge of of what we call "clear cut" PRs in our WIP process guidlines. But given the consensus here and overall, I thing this is fine.

A bit of documentation:

• The wiki still says the same. Which I consider a good enough +1.
  Side note: It would be great if we where used clear message in the wiki like "X is deprecated, use Y instead"
• The usage stats show a downward trend and the other tag has now significant more usage. This trend is what Martin was waiting for last time, so also a +1
  Side note: It will be interesting to see if this change will reduce the numbers.
• The JOSM ticket has no progress for the last 2 years and and the discussion there read similar to the one here. Maybe the merge here will trigger activity there as well. +0

Other notes:

• I extended the WIP guidelines with the "indicators for consensus" that you linked, @matkoniecz, thanks for that. 6776de0
• We really should add proper sidewalk fields to make all the existing data visible. I hope that this will be easier once Proposal to add field groups in presets (How to start iterating on more detailed OSM data) #1202 continues.
• PS: Lets not forget the :both variant, @waldyrious :-)