Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

classic Classic list List threaded Threaded
90 messages Options
12345
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
Two routes, currently mapped together in one relation, but also containing subrelations, and not in the right order, and the total of that is part of a higher relation which also has variants. You can’t blame the route designers, you simply have to accommodate the real life variations. For rendering, it’s ok like it is. For routing, not at all. Cycling planners are slightly better than hiking planners, but I would like to see one make a plausible navigation route out of the E2 Yorkshire relation as it is now.

And yes, that’s a challenge.

Fr Gr Peter Elderson

> Op 17 aug. 2019 om 11:20 heeft Andy Townsend <[hidden email]> het volgende geschreven:
>
>> On 17/08/2019 07:28, Peter Elderson wrote:
>>
>> Gpx gaps in some software do show up as straight lines. If it's just a missing piece and the rest is in order, no problem. In the case of the E2 in Yorkshire, lots of straight lines. Feed that to a navigation device and it will have you start in Muston, take you around and across the entire region multiple times, and end up near Barnetby Ie Wold. You wil actually have followed the E2 as well, I'll give you that!
>>
> "Following in the E2 in Yorkshire" would be an odd thing to do as there are two parallel legs of it (see https://hiking.waymarkedtrails.org/#?map=8!54.1195!-1.3988 ). From the Humber bridge one side follows the Wolds Way / Cleveland Way etc. to just west of Darlington, and on the other side of the county it runs from the Tan Hill Inn down the Pennine Way.  The problem here isn't the mapping in OSM, but the decision by whoever created the route to have two parallel routes called the same thing.
>
> What you'd logically actually do on the ground, of course is ignore the E2 altogether (it's not signed here) and either follow the Pennine Way signage or the Wolds Way / Cleveland Way etc. signage.
>
> Best Regards,
>
> Andy
>
>
>
>
> _______________________________________________
> Tagging mailing list
> [hidden email]
> https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Andy Townsend
> but I would like to see one make a plausible navigation route out of the E2 Yorkshire relation as it is now.

Where are you going from and where are you going to?  Without that information "make a plausible navigation route out of the E2 Yorkshire relation" makes no sense. 

You could certainly argue that splitting the legs of the route rather than the "parts geographically in Yorkshire" would make more sense, but making use of whole sections of the Wolds Way, Cleveland Way etc. would be better still.  That's one for you to resolve though, since you're the one who seems to have the issue here.

Best Regards,

Andy


.
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
The issue here is that these relations are there, they conform to current wiki documentation, but you can't simply use them in applications other then for rendering. Some of the issues (sorting, bridge simple gaps) may be solved with software at the data user's side, but on the whole you can't rely on it. Usability depends on mappers applying strict tagging and ordering standards and heavy maintenance to prevent problems. 

If code exists that solves these problems, I would very much welcome that if it could become availabe as e.g. a tool in JOSM. We could test that on the E2 Yorkshire relation.

Fr gr Peter Elderson


Op za 17 aug. 2019 om 12:31 schreef Andy Townsend <[hidden email]>:
> but I would like to see one make a plausible navigation route out of the E2 Yorkshire relation as it is now.

Where are you going from and where are you going to?  Without that information "make a plausible navigation route out of the E2 Yorkshire relation" makes no sense. 

You could certainly argue that splitting the legs of the route rather than the "parts geographically in Yorkshire" would make more sense, but making use of whole sections of the Wolds Way, Cleveland Way etc. would be better still.  That's one for you to resolve though, since you're the one who seems to have the issue here.

Best Regards,

Andy


.
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
In reply to this post by Andy Townsend


Op za 17 aug. 2019 om 12:31 schreef Andy Townsend <[hidden email]>:
> but I would like to see one make a plausible navigation route out of the E2 Yorkshire relation as it is now.

Where are you going from and where are you going to?  Without that information "make a plausible navigation route out of the E2 Yorkshire relation" makes no sense. 

That's exactly the point why you can't use a walking route relation as a navigational route, unless it has been carefully prepared to be an ordered, uninterrupted single-chained A2B route. 

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Andy Townsend
You haven't answered the question - I asked "Where are you going from and where are you going to?" in order to try and understand what real-life problem you're trying to solve. "I would like the ways of a relation to be sorted" is not a real-life problem.  What navigation software are you using and (in this example) where are you going from and where are you going to? 

I'm assuming that (on this imaginary hike through Yorkshire) you're not lugging a PC around running JOSM, trying to shade the screen from the sun on every hilltop so that you can see it, and wrapping it up whenever it starts raining.

Best Regards,

Andy



_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
I would like to use walking route relations as they are in OSM, in apps like OsmAnd, Garmin devices, and planner applications. Currently, you can't. As far as I can see, they all re-route  instead of using an already available route (= chain of ways). I would like to see unbroken elevation profiles for OSM-routes. I would like a gpx from an OSM route to be direcly usable in navigation apps and devices and reasonably adaptible as one route in route editing software (without spikes and ordering problems).

Fr gr Peter Elderson


Op za 17 aug. 2019 om 13:33 schreef Andy Townsend <[hidden email]>:
You haven't answered the question - I asked "Where are you going from and where are you going to?" in order to try and understand what real-life problem you're trying to solve. "I would like the ways of a relation to be sorted" is not a real-life problem.  What navigation software are you using and (in this example) where are you going from and where are you going to? 

I'm assuming that (on this imaginary hike through Yorkshire) you're not lugging a PC around running JOSM, trying to shade the screen from the sun on every hilltop so that you can see it, and wrapping it up whenever it starts raining.

Best Regards,

Andy


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

voschix
In reply to this post by Peter Elderson
The sequence of the component ways in a walking/hiking route relation is irrelevant for a hiker who use a navigation device to walk along the route.
Why?
How do you walk a walking route with a navigation device?
Basically you have two options:
A) you have prepared beforehand a GPX track, typically by using a routing/planning tool on a web site. That site does off-line routing for you. During the walk you follow the GPX track on the display of your GPS device (you keep you position on the track on that you see on the screen)
B) you use an equivalent navigation process as in (A) but you execute it in real time on the map that you have loaded on your navigation device.

In both cases you use some kind of algorithm that finds for you a (nearly) optimal route from waypoint to waypoint. What does the routing algorithm do? It assigns, from a table that defines your means of transportation, "cost" (or "penalty") per meter to each way on the map and basically finds the "cheapest" connection between consecutive waypoints (in the simplest case there are two waypoints: start and destination for the off-line navigation, and actual position and final destination for the on-line version).
The cost for each way type is dependent on your means of transport.
In the case of walking the cost for a motorway will be near infinity, with the effect that a hiker will not be routed via a motorway. A perfectly smooth footway will have a very low cost, a bumpy footway will have a bumpiness penalty (you are forced to walk slower) and so on.
How are hiking route relations handled? The fact that a specific way is part of a hiking route relation means that it gets a cost benefit with respect to an equivalent way that is not part of it. In this way the algorithm gives preference to those ways that are part of a walking route relation. I's up to designer of the routing cost table, how much advantage she will give to a given type of route; she may want to give a national hiking route a bigger cost advantage in comparison to a local hiking route.

This is the basics why the routing  algorithm does not "care" at all about the sequence of the ways in the hiking route relation.


On Sat, 17 Aug 2019 at 13:08, Peter Elderson <[hidden email]> wrote:


Op za 17 aug. 2019 om 12:31 schreef Andy Townsend <[hidden email]>:
> but I would like to see one make a plausible navigation route out of the E2 Yorkshire relation as it is now.

Where are you going from and where are you going to?  Without that information "make a plausible navigation route out of the E2 Yorkshire relation" makes no sense. 

That's exactly the point why you can't use a walking route relation as a navigational route, unless it has been carefully prepared to be an ordered, uninterrupted single-chained A2B route. 
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Andy Townsend
In reply to this post by Peter Elderson

> apps like OsmAnd, Garmin devices, and planner applications

I can answer the Garmin bit of that, as it's something that I do all the time.

Firstly, the ability to "follow a GPX" isn't something that most Garmin devices support.  None of the eTrex or GPSMap (walking) or Nuvi (car) models that I've had support it.  I believe that some models designed for cycling do, but let's be honest - you're not going to be cycling along most of the Wolds Way or Cleveland way.  It not just not allowed in places it'd be close to physically impossible.

What Garmins do offer is navigation by a series of waypoints, and they'll navigate between those either "as the crow flies" or allow a route that is appropriate for your current travel method.  Although the UI for walking- and car-focused devices are different, the principle is the same.  You'd typically add waypoints along the route manually - either on the device or in something like Garmin's PC software before loading to the device. "waypoints" and "routes" can fit into the same XML file as any GPX tracks you move to and from a Garmin device.

When creating maps from OSM data you can also include route membership information on the on-screen map - I usually put the route names in brackets after any way name.  I create different Garmin maps for walking and driving to remove some paths you don't want to drive on or roads you don't want to walk on (or aren't allowed to, based on access tags).

As an aside I do tend to create Garmin routes composed of waypoints for guideposts / route markers where I've got that info - it helps to find missing ones, and having guideposts and route markers in OSM helps other people see which bits of route are well signposted and which not.  You may notice that the Wolds Way has a few of these mapped as relation members ("role=marker"); other nearby routes are more complete.

To summarise then, I've never found a need to sort the ways in a walking route before being able to navigate it using a Garmin handheld.

Best Regards,

Andy



_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
Your viewpoint is different from mine. You want to do the routing. I want to avoid that, because the routing has already been done. The OSM-relation IS a route. It is entered as an exact chain of ways to folllow.

OsmAnd and Garmin should take the route itself, not waypoints to route to. It is odd that OsmAnd cannot navigate me along an ÒSM route that's shown on the map and is readily available from OSM. Instead, I have to translate it into a series of waypoints (gpx), then feed that to the app, and then it recalculates the route (chain of ways) it already had.

Vr gr Peter Elderson


Op za 17 aug. 2019 om 15:08 schreef Andy Townsend <[hidden email]>:

> apps like OsmAnd, Garmin devices, and planner applications

I can answer the Garmin bit of that, as it's something that I do all the time.

Firstly, the ability to "follow a GPX" isn't something that most Garmin devices support.  None of the eTrex or GPSMap (walking) or Nuvi (car) models that I've had support it.  I believe that some models designed for cycling do, but let's be honest - you're not going to be cycling along most of the Wolds Way or Cleveland way.  It not just not allowed in places it'd be close to physically impossible.

What Garmins do offer is navigation by a series of waypoints, and they'll navigate between those either "as the crow flies" or allow a route that is appropriate for your current travel method.  Although the UI for walking- and car-focused devices are different, the principle is the same.  You'd typically add waypoints along the route manually - either on the device or in something like Garmin's PC software before loading to the device. "waypoints" and "routes" can fit into the same XML file as any GPX tracks you move to and from a Garmin device.

When creating maps from OSM data you can also include route membership information on the on-screen map - I usually put the route names in brackets after any way name.  I create different Garmin maps for walking and driving to remove some paths you don't want to drive on or roads you don't want to walk on (or aren't allowed to, based on access tags).

As an aside I do tend to create Garmin routes composed of waypoints for guideposts / route markers where I've got that info - it helps to find missing ones, and having guideposts and route markers in OSM helps other people see which bits of route are well signposted and which not.  You may notice that the Wolds Way has a few of these mapped as relation members ("role=marker"); other nearby routes are more complete.

To summarise then, I've never found a need to sort the ways in a walking route before being able to navigate it using a Garmin handheld.

Best Regards,

Andy


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Andy Townsend

> You want to do the routing. I want to avoid that, because the routing has already been done. 

To be clear, I want to navigate from where I currently am to where I want to go.  If a route is blocked (or, heaven forfend, wrong in OSM) I still want to get where I'm going, even if I am not exactly where I thought I should be.

> OsmAnd and Garmin should take the route itself, not waypoints to route to. It is odd that OsmAnd cannot navigate me along an ÒSM route that's shown on the map and is readily available from OSM. 

I'm not sure what the current situation is with OsmAnd, but do know that it's been discussed within the last few months in the OsmAnd Google Geoup

With regard to Garmin, it sounds like you should be submitting a feature request to them - it is unlikely that they monitor OSM's tagging list for those.

It does sound, however, that you don't have a concrete use-case at all - you have a view of how things "should" be, but this doesn't seem to be driven by a real-world requirement.  That's why I've been asking for specifics throughout this thread (and Richard has too).

Best Regards,

Andy


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Joseph Eisenberg
I agree with Andy Townsend here.

Routes are complicated enough without needing to be always perfectly
sorted. Software developers and database users should make up for
this. Mapping is hard and takes precious human time. Computational
cycles are cheap.

OSM has never been designed to be used directly "as-is" from the raw
data. Even JOSM and ID do all sorts of clever things to show you a
reasonable rendering of the data as you enter it.

Please open issues requests with your favorite software providers,
like Garmin, or switch to one that will support routes in the way that
works for you. Perhaps one of the bike/hike websites like Map My Run,
Ride With GPS, etc already have something like this offered?

Mappers should not to do extra work and extra maintenance of data.
Spend time mapping, not sorting relations.

-Joseph Eisenberg

(PS: I help develop software a little, by contributing at
Openstreetmap-carto to improve and maintain the rendering. I haven't
developed a routing engine or anything so complicated myself).

On 8/17/19, Andy Townsend <[hidden email]> wrote:

>
>> You want to do the routing. I want to avoid that, because the routing has
>> already been done.
>
> To be clear, I want to navigate from where I currently am to where I want to
> go.  If a route is blocked (or, heaven forfend, wrong in OSM) I still want
> to get where I'm going, even if I am not exactly where I thought I should
> be.
>
>> OsmAnd and Garmin should take the route itself, not waypoints to route to.
>> It is odd that OsmAnd cannot navigate me along an ÒSM route that's shown
>> on the map and is readily available from OSM.
>
> I'm not sure what the current situation is with OsmAnd, but do know that
> it's been discussed within the last few months in the OsmAnd Google Geoup
>
> With regard to Garmin, it sounds like you should be submitting a feature
> request to them - it is unlikely that they monitor OSM's tagging list for
> those.
>
> It does sound, however, that you don't have a concrete use-case at all - you
> have a view of how things "should" be, but this doesn't seem to be driven by
> a real-world requirement.  That's why I've been asking for specifics
> throughout this thread (and Richard has too).
>
> Best Regards,
>
> Andy
>
>

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
In reply to this post by Andy Townsend
In this case, I do NOT want to go from A to B. I want to do the hike, that is the route, exactly as it is specified OSM. Those ways, in the exact order. I want my smartphone or garmin to guide me exactly along those ways, which were carefully picked when the route was entered into OSM. 

If that can't be done directly, I want to get an export that I can feed to my device or app, so it can recreate the route exactly, without adding, weighing, guessing or rerouting anything.

When I'm planning a hike, I want the software to start with the exact OSM route, not a rerouted version.

Fr gr Peter Elderson


Op za 17 aug. 2019 om 15:44 schreef Andy Townsend <[hidden email]>:

> You want to do the routing. I want to avoid that, because the routing has already been done. 

To be clear, I want to navigate from where I currently am to where I want to go.  If a route is blocked (or, heaven forfend, wrong in OSM) I still want to get where I'm going, even if I am not exactly where I thought I should be.

> OsmAnd and Garmin should take the route itself, not waypoints to route to. It is odd that OsmAnd cannot navigate me along an ÒSM route that's shown on the map and is readily available from OSM. 

I'm not sure what the current situation is with OsmAnd, but do know that it's been discussed within the last few months in the OsmAnd Google Geoup

With regard to Garmin, it sounds like you should be submitting a feature request to them - it is unlikely that they monitor OSM's tagging list for those.

It does sound, however, that you don't have a concrete use-case at all - you have a view of how things "should" be, but this doesn't seem to be driven by a real-world requirement.  That's why I've been asking for specifics throughout this thread (and Richard has too).

Best Regards,

Andy

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
In reply to this post by Joseph Eisenberg
Again, software can not handle e.g. the E2 relation. Simple sort, fine, bridge a small gap, handle a roundabout, fine, but not the more serious route-breaking issues. You can't expect Garmin to solve that, it's a data issue in OSM. Currently the only way to solve it is making the data flawless: one main route, single chain, no nodes, handle hierarchies, no duplicates, no branches, no areas. All alternatives separate. Then you can use it  as is for further processing. 

The main problem is maintenance against editing software and users breaking the routes. The solution should be protection against breaking, not repairing flaws at the client side.

Maybe the latest improvements in ID help. It's too early to tell. I'm still encountering issues all the time in the routes I check and use.

Fr gr Peter Elderson


Op za 17 aug. 2019 om 15:58 schreef Joseph Eisenberg <[hidden email]>:
I agree with Andy Townsend here.

Routes are complicated enough without needing to be always perfectly
sorted. Software developers and database users should make up for
this. Mapping is hard and takes precious human time. Computational
cycles are cheap.

OSM has never been designed to be used directly "as-is" from the raw
data. Even JOSM and ID do all sorts of clever things to show you a
reasonable rendering of the data as you enter it.

Please open issues requests with your favorite software providers,
like Garmin, or switch to one that will support routes in the way that
works for you. Perhaps one of the bike/hike websites like Map My Run,
Ride With GPS, etc already have something like this offered?

Mappers should not to do extra work and extra maintenance of data.
Spend time mapping, not sorting relations.

-Joseph Eisenberg

(PS: I help develop software a little, by contributing at
Openstreetmap-carto to improve and maintain the rendering. I haven't
developed a routing engine or anything so complicated myself).

On 8/17/19, Andy Townsend <[hidden email]> wrote:
>
>> You want to do the routing. I want to avoid that, because the routing has
>> already been done.
>
> To be clear, I want to navigate from where I currently am to where I want to
> go.  If a route is blocked (or, heaven forfend, wrong in OSM) I still want
> to get where I'm going, even if I am not exactly where I thought I should
> be.
>
>> OsmAnd and Garmin should take the route itself, not waypoints to route to.
>> It is odd that OsmAnd cannot navigate me along an ÒSM route that's shown
>> on the map and is readily available from OSM.
>
> I'm not sure what the current situation is with OsmAnd, but do know that
> it's been discussed within the last few months in the OsmAnd Google Geoup
>
> With regard to Garmin, it sounds like you should be submitting a feature
> request to them - it is unlikely that they monitor OSM's tagging list for
> those.
>
> It does sound, however, that you don't have a concrete use-case at all - you
> have a view of how things "should" be, but this doesn't seem to be driven by
> a real-world requirement.  That's why I've been asking for specifics
> throughout this thread (and Richard has too).
>
> Best Regards,
>
> Andy
>
>

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Kevin Kenny-3
In reply to this post by Andy Townsend
On Sat, Aug 17, 2019 at 9:44 AM Andy Townsend <[hidden email]> wrote:
> It does sound, however, that you don't have a concrete use-case at all - you have a view of how things "should" be, but this doesn't seem to be driven by a real-world requirement.  That's why I've been asking for specifics throughout this thread (and Richard has too).

There are some trails for which I wish eventually to produce map
books: a series of large-scale maps showing the route of the trail,
together with tables of linear distances along the trail (to obtain
distance between waypoints), and elevation profiles.  Of course, this
has the requirement that the "main stem" of the trail (I recognize
that there may be links, alternatives, etc.) needs to be _reducible_
to a coherent linear topology. (Note that I say _reducible_: I can
tolerate its not being actually _reduced_ in the database.) Answering
a question such as, "how far is it along the trail from the Route 28
highway crossing to the campsite at Salmon River?" needs such a
topology.

My operational definition when I'm mapping: "sorted well enough for
JOSM to show route continuity except where discontinuities exist in
the field or the mapping is incomplete, and for Waymarked Trails (or
similar software) to produce the elevation profile."

My operational definition when I'm consuming data: "the route segments
have endpoints that are sufficiently close to contiguous that a
reasonably simple, greedy algorithm can assemble a linear route"
(obviously, if there are route segments with specific roles in the
relation, that fact can imply a different topology). I shouldn't have
to resort to simulated annealing or machine-learning techniques to
assemble a route from a mess. If I get into needing to "approximate
longest Hamiltonian path", I'd be spending my time more effectively in
cleaning up the underlying data.

Ideally, I shouldn't have to resort to very much computation at all to
assemble a linear route, which is why OSRM flags all discontinuities.
But I don't complain when route segments are scramble. I fire up JOSM,
unscramble them, and leave it at that; I recognize that there are
editors that are less capable of maintaining a coherent route
relation, and that there are data consumers that require a stricter
one, so I just go in and repair it. I get rather more cross at way
ends that don't quite meet, because closing the tiny gaps requires
considerably more computation. I grumble considerably more at closing
gaps of a few metres.

I appreciate the work that the iD team has done so that I have to
repair these things much less often.

"Be strict in what you emit and permissive in what you accept" is a
reasonably good principle for data interchange in general.


--
73 de ke9tv/2, Kevin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Sarah Hoffmann
In reply to this post by Richard Fairhurst
On Sat, Aug 17, 2019 at 01:11:17AM -0700, Richard Fairhurst wrote:

> Peter Elderson wrote:
> > I would like to see this software in operation! Could you give me the
> > links of some applications
>
> I use my code in the backend of cycle.travel. It's not open source. I've
> seen code used by one other OSM-based site and there's a further one that's
> clearly using something similar. There are at least two really obvious
> strategies for dealing with relations like this.
>
> > The point is, as it is it's not good enough for data use besides
> > rendering. you can't rely on route relations for anything but rendering
>
> Once again: pretty much every OSM-based bike router uses route relations to
> influence routing. (That might give you a clue to one of the strategies.)

But this is a task which is essentially the same rendering. You only need
to know what routes are on a certain way segment and use that information
to adjust the weights for the road. Even if you do something fancy like
ensuring that you remain of the same cycling route, it still comes down to
using the relation information as a property of the ways. Our route relations
are well suited for that.

The problems come in if you want to go the other way. When you start with the
relation, want to determine where the route goes along. That information is
simply not contained in the route relations as long as you don't impose a
couple of restrictions. Sure, you can apply a couple of heuristics and get
a reasonably good result for most of the routes. But it remains guess-work.

I have no issue if relations require reasonable processing to get to a result
but I would like to see enough information encoded in the route relation that
the processing invariably gets me to the result that the mapper intended.
I consider sorting and the use of roles essential for that.

Sarah

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Paul Allen
On Sun, 18 Aug 2019 at 10:01, Sarah Hoffmann <[hidden email]> wrote:

I have no issue if relations require reasonable processing to get to a result
but I would like to see enough information encoded in the route relation that
the processing invariably gets me to the result that the mapper intended.
I consider sorting and the use of roles essential for that.

So far we've only been talking of hiking/walking routes where some feel sorting is
beneficial and others feel it is unnecessary.  At the risk of derailing the thread,
there are other types of route relation.

Take bus routes.  I know of a very complicated bus route.  At three times along the
route it drives into a cul-de-sac, reverses into a side-road of that cul-de-sac, then
goes back the way it came.  Those three places are not termini, they are merely
points along the route where most of the passengers remain on board.  In
one part of the route it does a loop-the-loop, going around four sides of a square.
It traverses other segments of the route twice, in the same direction, approximately
30 minutes apart.  Oh, and it does the reversing-turn trick in another place that
isn't a cul-de-sac, it just doesn't go any further along that road.

It is perfectly possible to render the route.  Trying to figure out the steps in the route
from the rendered route is pretty much impossible, although detailed inspection of
the one-way markings eliminates several possibilities.  We don't (as far as I know)
have any tool comparable to uMap that would allow ordinary users to step through
the route so they can comprehend the details, although splitting the route
into sub-relations would allow uMap to give a crude approximation of that
capability.  But one day such a tool might appear, at which point having the route
sorted would be beneficial.

Here is the route:
Good luck figuring out where it goes, even though the list of ways is (I think) correctly
sorted.  On the public transport layer you wouldn't even get the list of ways unless you
queried the route.

So maybe sorting isn't absolutely necessary, but it can make life easier.  Maybe
we shouldn't ever insist that mappers sort the routes they add, but I don't think
we should discourage them if they want to put in that effort.  Especially if, one day,
somebody comes up with a step-by-step tool for displaying routes.

--
Paul


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

dieterdreist


sent from a phone

> On 18. Aug 2019, at 14:43, Paul Allen <[hidden email]> wrote:
>
> Maybe
> we shouldn't ever insist that mappers sort the routes they add, but I don't think
> we should discourage them if they want to put in that effort.


+1, this is my opinion as well. Sorted routes make it usually easier for other mappers to do modifications and to see if the route has gaps


Cheers Martin
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Richard Fairhurst
In reply to this post by Sarah Hoffmann
Sarah Hoffman wrote:

> On Sat, Aug 17, 2019 at 01:11:17AM -0700, Richard Fairhurst wrote:
> > Peter Elderson wrote:
> > > The point is, as it is it's not good enough for data use besides
> > > rendering. you can't rely on route relations for anything but
> rendering
> >
> > Once again: pretty much every OSM-based bike router uses route relations
> to
> > influence routing. (That might give you a clue to one of the
> strategies.)
>
> But this is a task which is essentially the same rendering.

I was addressing Peter's point that route relations can only be used for
rendering, which is demonstrably false - they're used for influencing
weighting in routing.

> The problems come in if you want to go the other way. When you start with
> the relation, want to determine where the route goes along. That
> information
> is simply not contained in the route relations as long as you don't impose
> a
> couple of restrictions. [...]
> I consider sorting and the use of roles essential for that.

I'd fully agree on roles. The use of 'alternate' and 'forward'/'reverse'
roles for bike route relations dates back to the earliest days of bike route
mapping in OSM and is well established by now.

But just as long established in OSM is the principle that - since mappers
are our most precious resource - we optimise for the mapper, not for ease of
consumption. Allowing relations to be unsorted is an example of that. If a
relation represents a linear route, it's a SMOP to put the ways in the right
order.

There are two obvious strategies. 1 is: create an empty polyline P with
endpoints P1 and P2; iterate through the relation members; every time you
encounter a way with an endpoint P1 or P2, add it to the polyline
(potentially in reverse order) and remove it from consideration; repeat
until there are no ways left. This is broadly the approach I've used until
now.

2 is more involved but more fault-tolerant and flexible; create a routing
graph, then route from the relation's startpoint to its endpoint using a
very heavy uplift for members of this relation. This is a useful approach
where the route actually _is_ non-contiguous but you nonetheless want to
include connecting routes between the sections. (Quite a lot of American
rail-trails are like this, as are several UK National Cycle Network routes.)
This is an approach I'm investigating at present.

Approach 1 does of course fail if the relation doesn't represent a single
linear route. That would, however, still be true if the route was ordered.
There's probably a little more that editing software can do here, but unless
you want to require people to have 12 months of OSM experience before they
can map a change to their local cycle route, ultimately the solution is to
have better QA tools. Something like OSM Relation Analyser is faultless
algorithmically but the UI is less than immediate. If we were to have an OSM
Inspector-like view of lacunae, spurs and other relation issues, it would be
a whole lot easier to fix them. I occasionally wonder about coding this but
I'd love it if someone were to beat me to it.

cheers
Richard



--
Sent from: http://gis.19327.n8.nabble.com/Tagging-f5258744.html

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Peter Elderson
Richard Fairhurst <[hidden email]>:
Sarah Hoffman wrote:
> On Sat, Aug 17, 2019 at 01:11:17AM -0700, Richard Fairhurst wrote:
> > Peter Elderson wrote:
> > > The point is, as it is it's not good enough for data use besides
> > > rendering. you can't rely on route relations for anything but
> rendering
> >
> > Once again: pretty much every OSM-based bike router uses route relations
> to
> > influence routing. (That might give you a clue to one of the
> strategies.)
>
> But this is a task which is essentially the same rendering.

I was addressing Peter's point that route relations can only be used for
rendering, which is demonstrably false - they're used for influencing
weighting in routing. 

Ok, I agree you can use membership of a route relation as a weight factor, while you do routing. You don't use the relation as a route. The idea of walking routes is that they are the route and you follow the route, partly or with own adaptions, but not for rerouting. 
 
I did not mean that they can only be used for rendering. The point I made is that if there are interruptions, branches, duplicates, stray nodes, nested relations, etc in a route relation, it can still be used for rendering (or, for granting weight to individal ways when routing), but you cannot use it as a route that you follow from A to B by the ways it contains.

> The problems come in if you want to go the other way. When you start with
> the relation, want to determine where the route goes along. That
> information
> is simply not contained in the route relations as long as you don't impose
> a
> couple of restrictions. [...]
> I consider sorting and the use of roles essential for that.

I'd fully agree on roles. The use of 'alternate' and 'forward'/'reverse'
roles for bike route relations dates back to the earliest days of bike route mapping in OSM and is well established by now.

How then do I get it in OsmAnd and have it guide me forward the entire route, then backward the entire route? Or in a Garmin device? Mine does not have the option, as far as I know. I can do a lot, but it requires getting the OSM data organized exactly as required, and involves gpx transfer and route editing software, and in the end, Garmin reroutes the whole thing again, to arrive at exectly the chain of ways I started with in OSM. The main point is, as you say, data quality. QA tools can find problems, but the fixing is still up to the mapper. RA does things, but can simply not handle complex walking route relations.

JOSM does a decent job in detecting problems and it provides decent fixing tools. E.g. sorting. But that often fails when the route is broken. Just like overpass turbo export has a sorting routine, which works for simple ordering tasks but just gives up if it gets complicated. Sam with waymarkedtrails: its fine when simple ordering problems are there, it also handles hierarchies quite nicely, but if it doesn't work out completely it gives up and gives the relation as is. 

I would sure like a better QA tool. I also would like a sorting routine in JOSM that can handle all the complexities that prevent the full sort. If anyone has goed code for that and knows how to make that available as a plugin, that would be a great asset in maintaining OSM walking routes. I'm often sorry I am no good at programming. 


But just as long established in OSM is the principle that - since mappers
are our most precious resource - we optimise for the mapper, not for ease of
consumption. Allowing relations to be unsorted is an example of that. If a
relation represents a linear route, it's a SMOP to put the ways in the right
order.

Sure. The problem is that walking relations often do not add up to a linear route when sorted. If they do, that doesn't last very long. So you can't presume that they are, so it's up to the mapper to do it. 

There are two obvious strategies. 1 is: create an empty polyline P with
endpoints P1 and P2; iterate through the relation members; every time you
encounter a way with an endpoint P1 or P2, add it to the polyline
(potentially in reverse order) and remove it from consideration; repeat
until there are no ways left. This is broadly the approach I've used until
now.

2 is more involved but more fault-tolerant and flexible; create a routing
graph, then route from the relation's startpoint to its endpoint using a
very heavy uplift for members of this relation. This is a useful approach
where the route actually _is_ non-contiguous but you nonetheless want to
include connecting routes between the sections. (Quite a lot of American
rail-trails are like this, as are several UK National Cycle Network routes.)
This is an approach I'm investigating at present.

Approach 1 does of course fail if the relation doesn't represent a single
linear route. That would, however, still be true if the route was ordered.

The problem is, the route relation is often meant to be a single linear route, but in fact it isn't.
 
There's probably a little more that editing software can do here, but unless
you want to require people to have 12 months of OSM experience before they
can map a change to their local cycle route, ultimately the solution is to
have better QA tools. Something like OSM Relation Analyser is faultless
algorithmically but the UI is less than immediate. If we were to have an OSM
Inspector-like view of lacunae, spurs and other relation issues, it would be
a whole lot easier to fix them. I occasionally wonder about coding this but
I'd love it if someone were to beat me to it

I'm afraid testing is all I can offer. I could list problems to detect, but I think I would not be telling you news. Very important: handle nested relations (hierarchies). RA currently does not. 
 
cheers
Richard



--
Sent from: http://gis.19327.n8.nabble.com/Tagging-f5258744.html

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Roles of route members (was: Merging tagging scheme on wiki pages of Hiking, ...)

Kevin Kenny-3
On Sun, Aug 18, 2019 at 3:31 PM Peter Elderson <[hidden email]> wrote:
> I'm afraid testing is all I can offer. I could list problems to detect, but I think I would not be telling you news. Very important: handle nested relations (hierarchies). RA currently does not.

Uhm, yeah, that's a problem. For what it's worth, Waymarked Trails
seems to have a much better route assembler. It copes with
https://hiking.waymarkedtrails.org/#route?id=919642, which definitely
has lacunae (because I never seem to have time to get to Schoharie
County to get the data to close the gaps).

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
12345