[Tagging] Primary or Trunk?

Fernando Trebien fernando.trebien at gmail.com
Sun Nov 3 19:05:17 UTC 2013


If we use a new key, no apps would support it, until the Brazilian
community grows enough to start building its own apps.

Indeed, as discussions progressed, we began associating "importance"
primarily with traffic intensity and less with city connectivity. But
again, we would have to measure traffic objectively to classify like
that, so we picked observable structural characteristics that
correlate with traffic intensity. Importance was also associated with
safety, where lacking infrastructure (such as hard shoulders)
implicitly represents the (lack of) importance assigned by the state.
This is reinforced by the fact that the state itself publishes a map
where roads are classified by several structural characteristics (most
of which are part of the decision flow now). The only published
alternative classifies by administration level - which, as I said, we
found almost completely useless.

How many vehicles do you assume for each type of road? And how do you
measure it?

Note: I've felt the need for quaternaries and then thought that
ideally we should simply change the idea of road hierarchy into a
numeric tag, like admin_level. But given the length of the current
tradition, it seemed absurd to propose this.

On Sun, Nov 3, 2013 at 4:12 PM, fly <lowflight66 at googlemail.com> wrote:
> Hi Fernando
>
> Still think you should not change the meaning of a key but use an
> new/own key or even only a subkey for the other existing values
> considering living_street. Actually, I am in favour of deprecating
> living_street as it is already used for too many different cases and
> wrong in concept (at least for the original intention).
>
> Comming back to highway=* we do not attribute who is maintaining the
> road nor if it is a called and signed as national, state and municipal
> roads but we use it to tell about the importance of the road within the
> traffic network and about the number of motor_vehicles using the road.
>
> cheers fly
>
> Am 03.11.2013 18:35, schrieb Fernando Trebien:
>> Hello fly,
>>
>> Thank you for your concern. I'm glad we agree that classification is a
>> hard topic sometimes. First, the graph is intended for mapping within
>> Brazil and not overseas. I don't wanna delve too deeply in the details
>> of our discussion (it would probably be long, boring and maybe only
>> interesting for Brazilians themselves), but I can assure you that when
>> we established those definitions we were trying to make our mapping
>> style as compatible with the rest of the world as possible. And by
>> "we" I mean I was not alone - many active users participated, the
>> graph had many "candidate" forms before we agreed this was probably
>> the best we could do for now. Best in terms of balancing clarity
>> (objectivity), complexity (how hard is the decision-making process)
>> and quality of the end result (the map). The graph hasn't changed in
>> the last 6 months and we only have two Brazilian-borne complaints in
>> queue since then, one of which is minor, the other a little harder
>> (but already being addressed) and none related to living streets.
>>
>> So, you're right, living streets do not exist in Brazil according to
>> the strictest definitions of OSM (a legally designated road where
>> pedestrians have right of way over cars). However, I disagree that
>> they should not be used here at all. Brazil has no "primaries",
>> "secondaries" or "tertiaries", it has arteries, collectors and local
>> streets, national, state and municipal roads, and so on. These
>> concepts are mapped to OSM, with local adaptations, right? Many
>> definitions are left open in OSM; for instance, we had heated debates
>> on how to distinguish track, path and footway. When we reached out for
>> the international community, we discovered that each country has its
>> own definition, often more or less incompatible with that of other
>> countries.
>>
>> Then, there are a few people who think it is useful to map the concept
>> of "living streets" to some sort of warning system for drivers. Those
>> who would benefit most from this would be tourists, certainly. There
>> are places with poor infrastructure that simply force pedestrians to
>> share space with vehicles. There are places where this sharing is
>> simply part of the local culture. For all practical matters, these
>> places function pretty much like living streets (heavy pedestrian
>> traffic sharing space with vehicles, which are then forced to slow
>> down considerably), just without the legal requirement. A distracted
>> driver that does not know these facts has a higher chance of causing
>> an accident. And if the driver knew about these streets beforehand,
>> he/she would usually prefer to avoid them, thus affecting routing
>> decisions. All that is valuable local knowledge.
>>
>> On Sun, Nov 3, 2013 at 12:10 PM, fly <lowflight66 at googlemail.com> wrote:
>>> Am 02.11.2013 20:43, schrieb Fernando Trebien:
>>>> I know that in Germany and in Argentina roads are being classified
>>>> based primarily on administration level (national, regional, city,
>>>> etc.). Classifying like this probably works well when the entire road
>>>> system is well maintained.
>>>>
>>>> In Brazil, however, we had tons of discussions on how to do it and
>>>> ended up deciding (though reluctantly) to classify based on several
>>>> objective structural characteristics that seemed closely related to
>>>> "importance". That is mostly because many regional/municipal roads are
>>>> definitely more important (thus, preferable) than other, smaller
>>>> national roads. Here's what we ended up with:
>>>> http://i.imgur.com/YH8azIA.png
>>>
>>> Please, do not use living_street the way you have it in your picture.
>>>
>>> Living_street is a special case and has to be signed individually.
>>>
>>> It is one major mistake in OSM-Historie and should be changed to
>>> highway=*, living_street=yes like motorroad and bicycle_road.
>>>
>>> As already said, if one classification does not exist in a country just
>>> do not use it.
>>>
>>> On the other hand classification is not always that easy and if politics
>>> get involved it gets even more tricky.
>>>
>>> In my area traffic was split on two roads running almost parallel but
>>> one has a restriction on weight. I ended up with two primaries as one
>>> has much traffic but the otherone is the route for hgv.
>>>
>>> Cheers
>>> fly
>>>
>>>
>>> _______________________________________________
>>> Tagging mailing list
>>> Tagging at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/tagging
>>
>>
>>
>
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging



-- 
Fernando Trebien
+55 (51) 9962-5409

"The speed of computer chips doubles every 18 months." (Moore's law)
"The speed of software halves every 18 months." (Gates' law)



More information about the Tagging mailing list