[Tagging] maxspeed:signed=no - new proposed tag

Minh Nguyen minh at nguyen.cincinnati.oh.us
Thu Jan 7 23:40:07 UTC 2021


Vào lúc 05:28 2021-01-07, Martin Koppenhoefer đã viết:
> Am Di., 5. Jan. 2021 um 02:48 Uhr schrieb Minh Nguyen via Tagging 
> <tagging at openstreetmap.org 
> <mailto:tagging at openstreetmap.org>>:
> 
>     source:maxspeed=* seems capable of indicating the presence of a default
>     speed limit, since apparently the "<country_code>:<context>" syntax is
>     only for unsignposted speed limits. [1] But in the scenario Mateusz
>     describes -- an intrepid surveyor tags a street beforehand to flag
>     it to
>     armchair legal analysts -- how would the surveyor know which context to
>     assign? Also, wouldn't it be strange to see source:maxspeed=* on a
>     feature without maxspeed=*?
> 
> 
> 
> they would not know which context applies and would set the tag. From 
> the missing maxspeed and source:maxspeed, someone else can see that 
> information is missing and could check it on the ground. Maxspeed is not 
> something that should be done by armchair mappers (save the case that 
> there is sufficent and recent coverage of (non-google) street view data, 
> mapillary etc.).

What would a mapper set source:maxspeed=* to if they know a default 
speed limit would apply but don't know which context applies? Something 
ambiguous like source:maxspeed=US:no_sign?

>     The "<country_code>:<context>" syntax is also overly simplistic. It's
>     telling that no one has attempted to extend the table at [2] with
>     values
>     for the U.S. Each state has plenty of default speed limit contexts,
>     some
>     of which defy the OSM tagging model. [3]
> 
> 
> 
> this has been brought up several times, but someone would have to 
> actually add it to the wiki ;-)
> I have no idea, and have limited my actions to places where I am well 
> aware of the situation.

I wasn't simply lamenting that no one had rolled up their sleeves and 
added the information. My point is that the scheme is ill-fitting, 
bordering on unusable with respect to the United States (if not other 
countries as well). As a matter of fact, in the U.S., the 
source:maxspeed=* key is predominantly set to a value reminiscent of 
inscription=*, related_law=*, or source=*, not a regular syntax that a 
data consumer would find readily consumable, and not a value that would 
realistically fit in the documented syntax. [1]

>     Not everyone who surveys for OSM does so behind the wheel or even has a
>     driver's license. Not many data consumers can consult a local licensed
>     driver for the correct answer on demand. ;-)
> 
> 
> 
> they can simply omit the information. If someone has no clue about the 
> legal situation, they should avoid adding tags that describe this aspect 
> of the situation.

The proposed maxspeed:unsigned=* key is essentially a standardized form 
of fixme=* calling for the addition of maxspeed=* and source:maxspeed=*. 
Or are you advocating for the use of lots and lots of notes for this 
purpose instead? Alternatively, you could view it as tagging the absence 
of a sign, but I guess that's the topic of another ongoing thread about 
mapping things that don't exist. ;-)

>     Even someone who does drive may not have a full grasp of default speed
>     limits in their jurisdiction.
> 
> 
> 
> it still boils down to you cannot add what you do not know. Adding that 
> you do not know it on the other hand, does not seem helpful either. If 
> you do not know the limits, they are maybe not so important in your 
> area? ;-)

This proposal is about tagging the fact that one cannot ascertain the 
speed limit by a specific method, one among several. On the contrary, 
one cannot tag a source:maxspeed=<country>:<context> that one doesn't know.

The way the law works in the places I map, there are some things that 
are undetermined until there's a dispute. Speed limits here are 
ultimately determined by statute, but the statute is written to be 
subject to signage or, in the absence of signage, somewhat open to 
interpretation. You would want to avoid violating a default speed limit 
if you want to have a good day, so if you don't know it in a given 
situation, you have to make an educated guess or err on the side of caution.

An explicitly posted speed limit here isn't absolute or even 
deterministic: in driving school, I was taught that you may exceed the 
posted speed limit by a certain amount without any penalty, but 
conversely the basic speed rule means you may be cited for going the 
posted speed limit in inclement weather. If you rely on the ability to 
"go five over", as most drivers here do, then you need to be prepared 
for the one fine day when the state trooper on patrol decides to enforce 
the speed limit more strictly at their discretion.

The point of tagging speed limits is to give the user a best possible 
general answer. Recording the first step in a multi-step process of 
determining that answer, however unusable on its own, doesn't seem like 
such a bad idea.

> If you are able to recognize a maxspeed sign, you could add this 
> information, although I would say its usefulness is limited if you 
> cannot tell as well, what the implications are. E.g. maxspeed:signed=no 
> (I would not encourage maxspeed:signed=yes, as it is already tagged with 
> source:maxspeed=sign).

Right, as I see it, the proposal calls for maxspeed:signed=no but 
maxspeed:signed=yes would be an error or at least eyebrow-raising. We 
have other tags like that, such as noname=no and noexit=no.

[1] 
http://taginfo.geofabrik.de/north-america/us/keys/source%3Amaxspeed#values

-- 
minh at nguyen.cincinnati.oh.us




More information about the Tagging mailing list