I noticed that there are about 200 paths in Germany (but it is an international issue) with cycleway/sidewal:left/right trying to modele the side of the cycle path or footpath at segregated cycle-footpaths.
something like highway=footway + cycleway:left=lane bicycle=designated
instead of highway=path + segregated=yes foot=designated bicycle=designated
From my point of view, a cycleway=yes/track/lane does not fit to path types (footway, cycleway, path).
In the case of a segregated path (or possibly cycleway), I always expect the cycle path to be on the left (in countries where driving on the right is mandatory), so simply draw the path in the normal direction of travel.
Given the relatively infrequent use, I don’t expect applications to understand that either.
How do you see it? Is it necessary, ok, superfluous or wrong to use cycleway:left/right on a path?
Difficult. At a highway=footway I do not expect information about a bike path. Would feel it there therefore as wrong.
With path/cycleway + segrerated I know enough places, where also in Germany the cycle track is on the right side of the footpath.
Therefore, I think it makes sense that you record this info if necessary. The way it is recorded, however, I find unfortunate. However, I would not think of a better variant ad hoc.
But that is also another issue.
Footway + bicycle=yes defines primarily a footpath, which can be used by bicycles (depending on legal requirements; e.g. in Germany only walking speed).
Footway + bicycle=designated contradict each other. For this there is highway=path!
To me, those define different sorts of objects (albeit with the same legal access rights).
The “highway=footway” one says “this was built for use by pedestrians, and someone has added a segregated bicycle lane on the left (in the direction of travel of the way), which cyclists have a legal right to use”, but may be inferior to other “proper” cycle paths elsewhere.
The “highway=path” one initially says nothing about whether it’s likely to be appropriate for cyclists, but the " bicycle=designated" modifier changes that to say “actually, this is a highway=cycleway” (i.e. not inferior to other proper cycle paths). It doesn’t say which side pedestrians or cyclists go.
Not really, or more specifically it would be meaningless. It would nor change that existing data was tagged without being aware of such supposed default and cannot be expected to match it, except by accident.
The only thing that feels wrong to me, is the usage of highway=footway instead of highway=path, but I know of sidewalks with signs that would justify even this strange tagging (signs say footway with bicycles allowed but markings are clearly a separated cycleway). So while it looks odd, I actually don’t think it’s wrong. At least with the appropriate bicycle=* tag.
I haven’t felt the need yet, but ff there is a requirement to map the side of a cycleway on a segregated path, how else would you tag it?
But to answer the actual question: For me, it feels superfluous, but that’s in a country where the cycleway is always on the left. In countries with a different legislation, this might well be different, though.
If it was cycleway:<side>=lane, then not. But I agree that bicycle:lanes = yes|no seems the more elegant solution, because it’s gradually refining, so even routers that don’t look at :lanes can make use of the information. I have the feeling that a lot of people shy away from :lanes-tagging, though.
A path, to me, is basically a way for pedestrians and cyclists (and horse riders, but let’s ignore trhem) and both are equal users of the way. A footway or cycleway, to me, conveys a ranking of one above the other, which in the majority of cases isn’t given on segregated ways.
But the question was whether and how to tag the side where cyclists are supposed to ride on segregated ways. Not which highway=*-value to use.
cylceway bicycle=designated foot=designated
is all the same because the designated says that it is the intended type of traffic.
The German community mostly uses path in this case because it is in fact a multi-purpose path. ID prefers cycleway (maybe because cycling ist the faster travel mode). But yes this is controversial since the introduction of path and I do not want to discuss it here.