[Tagging] relations & paths

[Mateusz Konieczny]

May 15, 2020, 04:05 by bradhaack at fastmail.com:

>
>
> On 5/14/20 5:53 PM, Mateusz Konieczny      via Tagging wrote:
>
>>
>>
>>
>> May 15, 2020, 01:36 by >> jmapb at gmx.com>> :
>>
>>> On 5/14/2020 12:07 PM, Mateusz Konieczny via          Tagging wrote:
>>>
>>>> May 14, 2020, 16:40 by >>>> jmapb at gmx.com>>>> :
>>>>
>>>>> On 5/14/2020 10:01 AM, Paul Johnson wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> On Thu, May 14, 2020 at 5:48                    AM Steve Doerr <>>>>>> doerr.stephen at gmail.com>>>>>> >                    wrote:
>>>>>>
>>>>>>> On 14/05/2020 09:31, Jo wrote:
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Wed, May 13,                                2020, 17:44 Jmapb <>>>>>>>> jmapb at gmx.com>>>>>>>> >                                wrote:
>>>>>>>>
>>>>>>>>> Regarding the original question --                                  in what circumstances are                                  single-member walking/hiking/biking                                  route relations a good mapping                                  practice -- what would be your answer?
>>>>>>>>>
>>>>>>>>
>>>>>>>> Always
>>>>>>>>
>>>>>>>
>>>>>>> Doesn't that violate>>>>>>> https://wiki.openstreetmap.org/wiki/One_feature,_one_OSM_element>>>>>>>  ?
>>>>>>>
>>>>>>
>>>>>> No.  The route traverses the way, it's not the                    way. 
>>>>>>
>>>>>
>>>>> Okay. But surely this doesn't mean that every named              footway or path should be part of a route relation. 
>>>>>
>>>>>
>>>>> The bike trail that brad linked to, >>>>> https://www.openstreetmap.org/relation/6632400>>>>>  -- I've never been there but I don't offhand see any              reason to call it a route. (Brad has been there, I assume,              because it looks like he updated it 2 days ago.) There's              no information in the relation tags that isn't also on the              way itself. Is there any benefit to creating a route              relation in cases like this?
>>>>>
>>>>>
>>>> Better handling of future way splits, consistency.
>>>>
>>>
>>> I can see the advantage of using a route relation as a          somewhat future-proof persistent identity -- a relation URL          that will show the whole trail even if the way is split to add          a bridge, specify surface, etc. At the same time, though, it          feels like a bit of a stretch to declare any named trail of          any length as a route, 
>>>
>>>
>> Named way is not enough to be a route.
>>
>> Named path across forest is just a path. Route would be a        signed path through a forest,
>> with two objects:
>>
>> - path across forest (with or without name)
>> - signed route (that has some topology, signs, maybe also a        name)
>>
>>
> So you're saying any path with a sign should be a route.   Should    that extend to all tracks, and roads of all varieties also?    I    assume you are not limiting this to 'path across forest', it could    be path across desert,  or prairie, or town park?
>  
>
Any signed route may be mapped as a route relation.

But in many cases there will be a sign, without a route. "Beware of a dog" sign
does not mean that there is a route there.

And sometimes signed route will be signed with paint markings on trees,
or by piles of rocks or by some other method rather than be a sign.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200515/305eeb5a/attachment.htm>