[Tagging] Comments on documenting winter speed limits tagging
jyri-petteri.paloposki at iki.fi
Thu Apr 4 11:33:49 UTC 2019
On 3.4.2019 19.40, Tobias Knerr wrote:
> Despite the feedback that maxspeed:conditional would be better suited
> for this use case, this key has now been documented on the wiki:
> I find that rather disappointing because the key is a poor fit with
> several existing conventions. Even if we keep "winter" in the key, the
> ":seasonal" should definitely be dropped. After all, we use
> maxspeed:forward (not maxspeed:directional:forward) and maxspeed:hgv
> (not maxspeed:vehicular:hgv).
> Conditional restrictions still strike me as the best solution, though,
> and there are a few hundred examples in the database already.
The feedback was quite limited and you or anyone else expressed no
definite stance against documenting the current practice. As the use
case currently has no documented solutions, I considered it appropriate
to document at least the most common one. This is considering there was
no clear opposition for documenting it, although some alternatives were
suggested. We have also discussed this on #osm-fi at freenode and I
considered also that input in the decision to write the documentation.
Points raised raised in the discussions on #osm-fi at freenode were that
a) the winter speed limit is a relatively long-term condition, being
applied up to half of the year,
b) having this kind of long-term condition applied in the same scheme as
short-term conditions may make it more difficult to apply and understand
ways that have both a winter speed limit and other conditional speed limits,
c) many users also thought that the conditional scheme overall is
difficult for data consumers to use, and unlike short-term conditions,
it's important that the winter speed limit is easy to parse and use to
make sure the data is accurate at least most of the whole year.
I think it's most important that we at least have one scheme documented,
otherwise people won't create the tag at all. maxspeed:seasonal:winter
is the only one that is IMO used enough to warrant documentation without
proposal, and therefore I decided to document it based on the input.
More information about the Tagging