We have difficulties in our work because wiki is contradict.
Can anybody confirm the following:
Does motorway impily oneway tag or not? If you go to OSM wiki page, you can see that it says highway=motorway implies oneway=yes, however, two rows later it says Useful combination is oneway=yes. https://prnt.sc/ma3gqa
Also, if you open motorway_link wiki page which is available here, it is said that useful combination is oneway=*.
We got confused when we saw that the user MapAnalyser465 deleted a oneway tag from this way.
Can anyone confirm if oneway=yes is implied on both motorway AND motorway_link, or it isn’t?
As per the highway=motorway article’s “How to map” section, it seems to say that oneway=yes is the default and to use oneway=no if it’s a special case of a single-way, two-direction motorway. Therefore, both highway=motorway and highway=motorway_link are implied oneway=yes.
That being said, there’s no need to remove oneway=yes if it’s been explicitly tagged. It never hurts to have attributes like this explicitly specified, because it allows data consumers to properly use the data even if they aren’t following “defaults” and “implications” that are specified in the wiki.
Can you quote the source for this? My reading of the wiki is that motorway indeed defaults to oneway=yes (but it is preferred to explicitly oneway=yes nevertheless: “These ways should all point direction of travel and be tagged with oneway=yes”), but that motorway_link has no reliably established default for the oneway value.
I could have sworn it was established several years ago that motorway_link was treated as default oneway=yes unless otherwise specified, but the wiki article definitely doesn’t say this. It contains similar wording to highway=motorway, effectively saying that it’s typically oneway=yes, but it could be either yes or no and should be tagged explicitly. It does mention that some tools may make the assumption that the links are oneway=yes unless otherwise specified, but that won’t necessarily apply to all data consumers.
I guess we can then say that highway=motorway is default oneway=yes, whereas highway=motorway_link has no default and the oneway value should be specified in every case. I’ll have to check my local motorway_links for this.
Many motorway_links around here have two way traffic. At some point they split of course but for the two-way section there is no physical barrier between the lanes that would warrant mapping separate ways by the usual criteria to do so. I quickly checked, and they all seem to be mapped with oneway=no around here. Probably a precaution for oneway=yes being treated as a default. But I don’t see any need to set a default here at all.
Most of the motorway_links in North America are one-way, so that may be where some of the thoughts about making that a default came from. I just did an Overpass Turbo query for my local area, and all highway=motorway_links have been tagged with a oneway tag. Across all of Canada, there are a little over 30 that don’t have that tag. Spot-checking a few of them, I think there’s a mix of some that should have oneway=yes and it was just missed, and some that are oneway=no and that hasn’t been tagged explicitly.