[Tagging] Proposed rewrite Of highway=track wiki page
Bert -Araali- Van Opstal
bert.araali.afritastic at gmail.com
Mon Mar 8 23:06:29 UTC 2021
When I talk to my 8 year old daughter, who is very interested in
becoming an OSM mapper, or try to engage people from the community who
was not so fortunate to have higher education, I always use the example
of a tree to explain the different classification mechanisms.
Every linear part of the tree transports sap, all roads that take living
beings from A to B is a road. That's our basic classification.
Trunk roads, as fortunately the word already says, is like the trunk of
a tree. It's strategic, it takes the sap, the minerals, the sugars from
one important place to another, from the roots to the crown and vice versa.
Some trees have wide trunks, other narrow. Some might have a lot of sap
running through it, some not. Some have short trunks, others are high
and long. That doesn't however change the fact that all of them are trunks.
Their are trees who have multiple trunks, but they are the same
important, but you might call them primary, like primary roads. They
are still strategic but provide roads to move from one strategic place
to multiple other strategic places, multiple crowns, or one crown fed by
So Primary roads are strategic roads, you might also call them trunk
roads, although we prefer trunk to identify a single strategic road.
Secondary, or if you want collector roads, collect the sap traffic from
the major branches to direct it to the trunk.
Tertiary, or if you want feeder roads, collect the sap from minor
branches to direct it to the collector branches.
Residential roads, or if you want local roads, are the stems that lead
to the leafs, or the smallest branches with many leaves. Like the roads
that lead to one or more houses. So what about industrial roads ? Yep,
they are the same, they are also local roads, like the stems or small
branches that carry the fruits or the flowers. The stems have exactly
the same function, they carry the sap locally. We have chosen the word
residential, because most of them lead to houses where people live, work
etc... Like mostly there are more leaves then factories or fruits.
Why service roads ? Well, some leaves grow on one single stem, sometimes
leaves look different, and you have multiple leaves on one stem. If the
stem belongs to one leave, it serves 1 particular leave only, we call it
a stem serving 1 purpose, it's a service road. The same like some houses
are build along a road, a local road, don't have driveways because they
are close and along the road. Others are more far, and have their own
driveway. So those driveways are service roads.
A satelite image of a road network looks much the same like a tree.
Why do we need unclassified ? Look at it from the point of view from an
ant, walking on the tree. The ant only knows that it walks on a road,
it can't see that far to see if that road is a minor branch or a major
branch, or even a shallow trunk. It does however know that it is
something that belongs to one of these, because sap runs through it.
It's like a mapper in the field without a satellite imagery. The ant can
only find out what the road exactly is, by walking all the way from one
end to the other, or by asking other ants who already acquired that
knowledge. In case they are not there, it should call it unclassified,
to indicate it needs more information to assign a final judgement.
So where does the classification end up with track ? Well, all branches
that are in the early stages of development, that have not led to the
development of fruits or leaves or flowers, all of those transport sap
but have no clear determination yet. All those remain tracks.
What about motorway, express-way, path, the classification based on
throughput. Also there is a very good similarity with our tree. To
increase the throughput of sap, a tree may grow a wider trunk. To
increase the flow of sap a human may trim some branches on the trunk.
Cut or trim some minor branches from major branches etc...
Motorways, express-ways and paths are tracks where specific access
AND/OR physical measures are have evolved spontaneously or taken
purposely to improve the throughput.
For motorway and express-way: higher speed desired, so access
restrictions for slower then average speed.
For path: slower speed desired: so access restrictions for higher then
So the most important criteria here is separation based on speed. Other
measures taken are there only to further support the throughput. More
lanes, width, lane width etc... are some of these measures but not the
prime and most important: separation.
So I might have wide motorways, narrow motorways, narrow paths, wide paths.
Path comes here as a difficult case here to find similarity. The best
comparison could be the nerves in a leaf, they are still transport ways
for sap, but slow moving. These nerves can be wide or narrow, but all
are typical because of slow moving sap.
Remains, what are tracks: all roads that have no specific socio-economic
purpose and where no intervention has taken place to separate traffic
that moves at significant different speeds.
Hope this helps too,
On 08/03/2021 20:39, Bert -Araali- Van Opstal wrote:
> Every road is a track, every modification of a natural environment by
> living creatures to move from one place to another is a track.
> So start your mental journey with highway=track.
> Now you identified a track. First question: Can I add a social or
> economic importance to it (for humans but might as well work for animals).
> If yes, find a suitable socio-economic tag, you end up with
> highway=trunk(= all strategic tracks), primary(strategic bt not
> crossing international borders), secondary (=collector),
> tertiary(=feeder), residential (=local, any local importance, can be
> industrial) ,service or unclassified if you do mean it has a
> socio-economic importance but you couldn't determine which
> specifically or a single key, then highway=unclassified (room for
> considering combining multiple values).
> You can't or don't want to add a socio-economic importance, you remain
> with highway=track.
> Can I use a tag to specify if it is modified by humans to improve the
> throughput of the road. (by access restrictions, physical
> modifications like solting it from crossing roads, making it wider or
> narrower). If yes, use highway=motorway, path. Possible additional
> tags like express can be used.
> Ask yourself the same question if you used one of the socio-economic
> keys, so any highway from the above socio-economic classes can be a
> motorway or a path. (also room for improvement, our current tagging
> guidelines don't allow multiple values in the highway key), but as far
> as I am concerned might be highway=primary;motorway or might as well
> be highway=primary;path).
> Possible additional values: alley etc...
> No clear answer ? You still are with highway=track.
> You remain with highway=track.
> Missing in OSM: highway top level values to indicate how it is
> managed, however these should not be on the top level, because the
> tracks from the above can be managed by humans, taking in
> consideration values for names given worldwide.
> So we need a subkey for the highway tagging to describe it's
> management, often reflected in specific names.
> This could be highway:name_managed=highway, interstate, freeway,
> motorway, trunk, bypass, ringway... with additional tags for the
> operator=* and or translation so it can be understood worldwide: f.i.
> highway:name_managed:nl= snelweg, express weg, ringweg ... or
> highway:name_managed:de= Autobahn or highway:name_managed:fr=
> route_nationale, autoroute ... or highway:name_managed:hi=हाईवे... or
> highway:name_managed:zh-hant= 高速公路... or
> highway:name_managed:ru=автомагистраль... ) etc...? Add an admin_level.
> Finally, ask yourself if you would like to add tagging to describe
> physical restrictions or improvements that support one of the above
> classifications, like paved/unpaved or specific surface, smoothness,
> seasonal or weather condition aspects ?
> But it remains a highway=track or one of the highway=* tags above.
> This doesn't change the top-level highway key.
> In the end it remains a track, like all roads do, like all highway=*
> do. You remain with highway=track through the elimination process
> described above.
> Works everywhere, globally, in every language, simple and shows how
> the mess is actually not a mess if we all think in the same way....
> which we do, we are all living beings.
> Bert Araali
> On 08/03/2021 19:09, Kevin Broderick wrote:
>> I think it's worth noting that, at least as currently tagged, there
>> are more than a few tracks in rural portions of the U.S. that, while
>> not maintained for travel by sedan, are more a part of the road
>> network than the dead-end agricultural/forestry access roads that
>> have been discussed thus far.
>> In New England, a lot of them are old parts of the road network that
>> are not maintained to an auto-friendly level (if at all), yet may
>> remain travelable and in some cases are still public right-of-ways.
>> E.g. https://www.openstreetmap.org/way/19723462
>> <https://www.openstreetmap.org/way/19723462> — if you're on foot and
>> wearing muck boots, it's probably quicker to walk the track than to
>> follow the residential road out to the state highway and go all the
>> way around to the other side. Or
>> <https://www.openstreetmap.org/way/19724346>, similar story, although
>> it's actually parallel to a well-maintained state highway. Both are
>> public right-of-ways, but neither is maintained at a level that is
>> going to make taking your rental car through a particularly great idea.
>> <https://www.openstreetmap.org/way/288861133> is another good
>> example; it's decidedly a shorter route (in distance, not time)
>> between Waitsfield Common and Northfield than any of the modern
>> roads, but it is no longer maintained for automotive traffic nor a
>> public right of way. I'm pretty sure that, historically, it was a
>> higher-level member of the road network.
>> Out west, a couple of examples where tracks are decidedly part of the
>> road network, but don't easily fit into other classifications:
>> <https://www.openstreetmap.org/way/5424246> — largely a recreational
>> route for public-land access via ATV and 4x4 (or dirt bike), but a
>> through route nonetheless.
>> <https://www.openstreetmap.org/way/5425643> — roughly parallel to a
>> state highway, not particularly well-maintained (IIRC, this section
>> isn't sedan-friendly) but provides access to public (USFS) land as
>> well as some remote residential structures (possibly seasonal, not
>> 100% sure if that particular segment is plowed in winter or not, but
>> I don't think it is)
>> The second (5425643 / Meeteetsee Trail) could arguably be a
>> residential road, but that doesn't seem to be its primary function in
>> the road network, plus the physical characteristics of a residential
>> road generally allow for UPS deliveries and such, which I don't think
>> happen there (I believe you *could* get a box van across it in the
>> summer, as long as it hadn't rained recently, but it wouldn't be quick).
>> I'm not sure what the best answer is, but I thought that having some
>> more corner-case examples to look at might inform the discussion.
>> On Mon, Mar 8, 2021 at 9:49 AM Zeke Farwell <ezekielf at gmail.com
>> <mailto:ezekielf at gmail.com>> wrote:
>> Interesting. I am familiar with wide logging roads like the one
>> shown here:
>> I would not expect it to be tagged as 'track' though. I would
>> expect 'unclassified' because the wide logging roads I'm thinking
>> of connect from the general road network to the narrower logging
>> roads that I would expect to be tagged as 'track'. In my mind
>> the classification 'track' is a combination of function/purpose
>> and physical characteristics. A driveway to a house may match
>> the physical description by looking like this tracktype grade3
>> but it does not match the functional description because it
>> provides access to a full time residence. A wide logging road
>> matches the functional description by being primarily for
>> logging, but does not meet the physical description because it is
>> wide enough for two large vehicles to pass. Perhaps my mental
>> model is too narrow though. I will be interested to here more
>> feedback on the subject.
>> On Mon, Mar 8, 2021 at 9:18 AM Mateusz Konieczny via Tagging
>> <tagging at openstreetmap.org <mailto:tagging at openstreetmap.org>> wrote:
>> Mar 8, 2021, 15:00 by dieterdreist at gmail.com
>> <mailto:dieterdreist at gmail.com>:
>> Am Mo., 8. März 2021 um 14:47 Uhr schrieb Mateusz
>> Konieczny via Tagging <tagging at openstreetmap.org
>> <mailto:tagging at openstreetmap.org>>:
>> In this specific case it was clearly highway=track as
>> it was
>> - used by vehicles solely and only for logging forest
>> (and by tourists as annoying connecting segment)
>> not accurate according to your description below (access
>> to sawmill)
>> It was not between general road network and sawmill, it was
>> between forest and sawmill, branching later into smaller and
>> more standard highway=track
>> that branched into even smaller ones and terminated in a forest.
>> And to clarify tourists parts - they were on foot, maybe some
>> using bicycles.
>> Tagging mailing list
>> Tagging at openstreetmap.org <mailto:Tagging at openstreetmap.org>
>> Tagging mailing list
>> Tagging at openstreetmap.org <mailto:Tagging at openstreetmap.org>
>> Kevin Broderick
>> ktb at kevinbroderick.com <mailto:ktb at kevinbroderick.com>
>> Tagging mailing list
>> Tagging at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Tagging