Author Topic: GPX editor / route planner wish lists?  (Read 34894 times)

Panoramix

  • 50 61 6E 6F 72 61 6D 69 78
  • Suus cuique crepitus bene olet
    • Some routes
Re: GPX editor / route planner wish lists?
« Reply #25 on: July 13, 2012, 11:43:12 am »
Well whilst I love the nobility of the idea of contributing selflessly to the cause of advancing the sum total of human intellect, I am at least partly doing this for my own personal gains.

Fair enough.

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #26 on: July 13, 2012, 12:11:00 pm »
Have you thought at all about the save/share/library side of things?

I can understand wanting to avoid this, either for the same reasons as bikehike or just to keep the server-side part of things as light as possible, but it is a pretty useful facility.

The current bikehike approach is pretty neat, but I suspect it's quite a high barrier for those who don't already have web hosting stuff set up.  I wonder if something along the lines of Dropbox integration might be a better compromise.
Yes, I have thought about this, it is definitely something I would like to do, but would like to understand exactly what the legal issues are to be honest.

What actually is the reason bikehike took the route database off line?
Is it just being overly paranoid, as bikeroutetoaster, bikely etc. all seem to store routes.
Is it not pointless anyway since you can still display a route with a url just by passing a "lnk" parameter, it's still bikehike that's displaying the route, just not hosting it.

I know it says something like 'to protect against the, however small, possibility of litigation', but is it copyright litigation - like you're displaying my intellectual property and you don't have the right to, or is it of the 'you recommended this route and I crashed so i'm holding you responsible' type?
If it's the latter, I wonder if I could simply store UNrecommended routes, such as dual carriageways, with a warning if you then route down one. You can hardly sue me for advising you NOT to go down a certain road...

I'm thinking I could at least display recommended cafes, etc, with hopefully accurate locations unlike google  ::-) ;)


Re: GPX editor / route planner wish lists?
« Reply #27 on: July 13, 2012, 12:42:37 pm »
Clicking along BikeRouteToaster at small intervals whilst changing the routing prefs in the side panel as appropriate, and using 'delete last' when it goes wrong, is probably the best I've found so far. 
Ditto. Sometimes, one can let it auto-route over fairly big sections, but in general the intervals have to be modest. Also works well in BikeHike, with the option of manually routing on the OS map - until it runs out of tiles.

I've recently found, to my pleasure, that the cycle cut-through on one rat-run blocked road I often use to get out of Reading on CTC rides that I lead is now recognised, thus saving me a little bit of effort.

Quote
But drag-editing afterwards is almost impossible, so it's far from perfect.
Agreed.
"A woman on a bicycle has all the world before her where to choose; she can go where she will, no man hindering." The Type-Writer Girl, 1897

frankly frankie

  • I kid you not
    • Virtual Alps
Re: GPX editor / route planner wish lists?
« Reply #28 on: July 13, 2012, 01:21:13 pm »
oh sorry I thought OSM was OS maps... :-\ I know it stands for openstreetmap and OS stands for ordnance survey, but I thought the map it rendered with its contours, etc was an ordnance survey map... is it not?
It's possible the contours and so on on OSM are from the OS, I'm not 100% sure.

AFAIK the contours that are usually used in conjunction with OSM maps are sourced from NASA (originally, probably via some intermediaries).
"This is a complex subject, with a need for more than one highlighter pen."

Re: GPX editor / route planner wish lists?
« Reply #29 on: August 08, 2012, 12:11:25 pm »
oh sorry I thought OSM was OS maps... :-\ I know it stands for openstreetmap and OS stands for ordnance survey, but I thought the map it rendered with its contours, etc was an ordnance survey map... is it not?
It's possible the contours and so on on OSM are from the OS, I'm not 100% sure.

AFAIK the contours that are usually used in conjunction with OSM maps are sourced from NASA (originally, probably via some intermediaries).

Specifically: http://wiki.openstreetmap.org/wiki/SRTM

Re: GPX editor / route planner wish lists?
« Reply #30 on: August 08, 2012, 12:14:12 pm »
Being dissatisfied with the comprehensiveness of any of the current gpx creation / route planning tools available, I've decided to write my own...

Why not check out / contribute to an existing Open Source project, like QlandkarteGT ?
The basics of what you want are probably there, and additional features are usually welcome !

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #31 on: August 08, 2012, 12:59:12 pm »
Being dissatisfied with the comprehensiveness of any of the current gpx creation / route planning tools available, I've decided to write my own...

Why not check out / contribute to an existing Open Source project, like QlandkarteGT ?
The basics of what you want are probably there, and additional features are usually welcome !

See reply #21 to Panoramix on previous page.

For my own personal usage, the most pressing want is to have a simple editor that uses a google map a bit like bikehike, but can edit multiple tracks / routes at once, which bikehike can't.

MapSource can handle multiple tracks/routes at once, but its rendering is (imho) rubbish.

When this basic functionality of loading/saving multiple tracks/routes is there, I'll publish it, but it'll still be a work in progress and I'll continue to add new features all the time.

contango

  • NB have not grown beard since photo was taken
  • The Fat And The Furious
Re: GPX editor / route planner wish lists?
« Reply #32 on: August 25, 2012, 06:49:09 pm »
Have you thought at all about the save/share/library side of things?

I can understand wanting to avoid this, either for the same reasons as bikehike or just to keep the server-side part of things as light as possible, but it is a pretty useful facility.

The current bikehike approach is pretty neat, but I suspect it's quite a high barrier for those who don't already have web hosting stuff set up.  I wonder if something along the lines of Dropbox integration might be a better compromise.
Yes, I have thought about this, it is definitely something I would like to do, but would like to understand exactly what the legal issues are to be honest.

What actually is the reason bikehike took the route database off line?
Is it just being overly paranoid, as bikeroutetoaster, bikely etc. all seem to store routes.
Is it not pointless anyway since you can still display a route with a url just by passing a "lnk" parameter, it's still bikehike that's displaying the route, just not hosting it.

From what I recall they were concerned that someone might download a route, attempt to ride it, have an accident, and then sue bikehike for presenting it as a bike route. It might be the kind of case that would be destroyed with the application of an ounce of common sense but then people hosting web sites largely as a public service don't want to be tied up with legal threats, lawyers' fees etc as payment for their good deeds.

Quote
I know it says something like 'to protect against the, however small, possibility of litigation', but is it copyright litigation - like you're displaying my intellectual property and you don't have the right to, or is it of the 'you recommended this route and I crashed so i'm holding you responsible' type?
If it's the latter, I wonder if I could simply store UNrecommended routes, such as dual carriageways, with a warning if you then route down one. You can hardly sue me for advising you NOT to go down a certain road...

I may be verging on paranoia here but I'd be inclined to avoid anything relating to a recommendation. If a dual carriageway is "unrecommended" and the adjacent steep rocky bridlepath is not, you could end up with someone who routed down the bridlepath to avoid the dual carriageway and blames you for the damage to their bike or themselves when they considered the dual carriageway to be entirely safe. Especially since "dual carriageway" could mean anything from the A1 to a 30mph urban dual carriageway.

We're talking about the potential for truly silly legal action but with the proliferation of no-win-no-fee ambulance chasers out there my own thoughts would be to stick to the facts and let people decide for themselves which route they would rather take.

Quote
I'm thinking I could at least display recommended cafes, etc, with hopefully accurate locations unlike google  ::-) ;)

Cafes with reviews could be a good thing, recommendations can easily become outdated especially if a cafe changes hands. Accurate locations would be good, as would waypoints for cafes so you could download a route with waypoints for cafes, bike shops, cheap lodging etc along the way. If you could include some basic information, even if just a phone number and an average rating, that would help - if you need a bike shop and there are several to choose from you can choose a good one, if a bike shop is nearby but overwhelmingly badly rated you might stop there to buy a spare tube to get you to the next bike shop, or whatever.
Always carry a small flask of whisky in case of snakebite. And, furthermore, always carry a small snake.

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #33 on: August 27, 2012, 11:40:27 am »
Thanks for the thoughts - exactly what I thought. I'm not sure how the likes of bikeroutetoaster, bikely.com etc get round the threat of legal action - I presume it's just that they're hosted by big enough companies to fight it off, while bikehike isn't?

Toady

Re: GPX editor / route planner wish lists?
« Reply #34 on: September 19, 2012, 04:13:42 pm »
I'm a bit late to this thread.  FWIW here are a couple of thoughts.

Background: I use the Garmin Mapsource tool with downloaded OSM maps and it does what I want pretty much OK.  I click on the roads I want (typically little B roads) and it builds my route between these points.  That way I coerce it away from the main roads.  Very rarely I want to go where it doesn't think a road exists (eg a pedestrian cut-through) and I have problems there (and typically just commit the route to memory for that bit).  Then I view the profile to see where the hills are.

Anyway - the features I don't have, which I would like, are (a) "where is that hill?" - ie correlating between the profile and the route on the map.  I'd like to be able to add a route point by clicking on the profile (typically at the top of a hill) and (b) "how far to here?", ie to be able to show how a cumulative distance to a particular point just by hovering the mouse.

I think (b) has been requested above. 

Agree that the rendering of multiple routes by Mapsource is confusing.  However I typically only work on one route at a time.

Manotea

  • Just 1 sob, Vassily
Re: GPX editor / route planner wish lists?
« Reply #35 on: September 19, 2012, 09:57:17 pm »
A mechanism for plotting the flattest/hilliest route by specified distance threshold, i.e., shortest then +n%.
Anfractuous & Nyctophobic Oct 13       Willy Warmer Jan 19    Winter Boat Ride Feb 10

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #36 on: September 19, 2012, 11:48:42 pm »

Anyway - the features I don't have, which I would like, are (a) "where is that hill?" - ie correlating between the profile and the route on the map.  I'd like to be able to add a route point by clicking on the profile (typically at the top of a hill) and (b) "how far to here?", ie to be able to show how a cumulative distance to a particular point just by hovering the mouse.

I think (b) has been requested above. 

Agree that the rendering of multiple routes by Mapsource is confusing.  However I typically only work on one route at a time.

yes, will be able to do (a) and (b).
will likely be more right click and context menu based for interacting with the profile rather than hovering (I find implementations that auto-pan the map when hovering over the profile annoying), but yes, essentially what you want, will be context menu items like 'Highlight [this summit] on route" and "Routepoint here" etc.

This will be just another way of creating a route based on a track. The other main way will be by simply inputting a distance, so you can create a 'sparse' route with just the routesheet instructions and distances not have to fathom out where the junctions/instructions are, it will figure it out for you based on the track (this is one of the main things I want myself - and what sort of led me to start doing it.)


A mechanism for plotting the flattest/hilliest route by specified distance threshold, i.e., shortest then +n%.

No chance. Sorry. well, certainly not like you want, anyway. Google does offer 'route alternatives' so I could do a quick check and give the profiles of each one I suppose but it wouldn't be a true profile-based routing algorithm like I suspect you want.

What i can do, and will probably build in which i think vorsprung also suggested, is to enable the user to 'chop and change', i.e. change the route and see how it alters the profile, change it back, try a different route, see what that's profile is, etc. but you've got to provide the suggestions in the first place for it to evaluate - having it go away and automatically figure out the least hilly route is a whole different kettle of gibbons i'm afraid!

Manotea

  • Just 1 sob, Vassily
Re: GPX editor / route planner wish lists?
« Reply #37 on: September 20, 2012, 10:40:08 am »
A mechanism for plotting the flattest/hilliest route by specified distance threshold, i.e., shortest then +n%.

No chance. Sorry. well, certainly not like you want, anyway. Google does offer 'route alternatives' so I could do a quick check and give the profiles of each one I suppose but it wouldn't be a true profile-based routing algorithm like I suspect you want.

What i can do, and will probably build in which i think vorsprung also suggested, is to enable the user to 'chop and change', i.e. change the route and see how it alters the profile, change it back, try a different route, see what that's profile is, etc. but you've got to provide the suggestions in the first place for it to evaluate - having it go away and automatically figure out the least hilly route is a whole different kettle of gibbons i'm afraid!


He heh...  how about, assuming one can identify a hill on a map (contours) or be able to identify the hill/road from the profile, having the option to 'exclude' the road/region from the route? I guess that's a re-statement of what you wrote above...
Anfractuous & Nyctophobic Oct 13       Willy Warmer Jan 19    Winter Boat Ride Feb 10

dasmoth

  • Techno-optimist
Re: GPX editor / route planner wish lists?
« Reply #38 on: September 20, 2012, 11:08:29 am »
He heh...  how about, assuming one can identify a hill on a map (contours) or be able to identify the hill/road from the profile, having the option to 'exclude' the road/region from the route? I guess that's a re-statement of what you wrote above...

Pretty sure Google's API won't do this.  Only positive constraints ("pass through this point"), not negative ones ("keep away from this huuuuge mountain").  The best you can do is ask for a few alternative routes and hope to get one that skirts the mountain.

There is always the "write your own routing algorithm" solution...  :demon:
Half term's when the traffic becomes mysteriously less bad for a week.

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #39 on: September 20, 2012, 02:17:38 pm »

He heh...  how about, assuming one can identify a hill on a map (contours) or be able to identify the hill/road from the profile, having the option to 'exclude' the road/region from the route? I guess that's a re-statement of what you wrote above...


Yes, it would be possible given the right data, but you would have to write a new routing algorithm to do it.
I'm not writing my own routing algorithm for behind this site, I'm just using google's. They expose it as a service that you can just call via an API.
You just say 'give me the route from this coordinate to this other coordinate', and it fires back an array of intermediate points representing the route you should take.
Now, once you've got those intermediate points back, for each of them you can then also ask it 'what is the altitude of this point', and it will tell you that, so from that you can draw the profile.

But you don't have a hand in deciding what the intermediate points are based on the altitude. For that I would have to write a whole new router of my own, which is something I'm not going to do (yet). I might eventually, but it is quite a big job, and I want to write this interface first. You would need node data (possibly from OSM) that is allowed to be used in such a way and also altitude data but given that it could be done.

Manotea

  • Just 1 sob, Vassily
Re: GPX editor / route planner wish lists?
« Reply #40 on: September 20, 2012, 03:36:48 pm »
Thanks for the explanation. The option to avoid roads/set exclusion zones is fairly standard for PC based tools (autoroute, mapsource).
Anfractuous & Nyctophobic Oct 13       Willy Warmer Jan 19    Winter Boat Ride Feb 10

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #41 on: September 21, 2012, 02:16:05 pm »
A mechanism for plotting the flattest/hilliest route by specified distance threshold, i.e., shortest then +n%.

No chance. Sorry. well, certainly not like you want, anyway. Google does offer 'route alternatives' so I could do a quick check and give the profiles of each one I suppose but it wouldn't be a true profile-based routing algorithm like I suspect you want.

What i can do, and will probably build in which i think vorsprung also suggested, is to enable the user to 'chop and change', i.e. change the route and see how it alters the profile, change it back, try a different route, see what that's profile is, etc. but you've got to provide the suggestions in the first place for it to evaluate - having it go away and automatically figure out the least hilly route is a whole different kettle of gibbons i'm afraid!


He heh...  how about, assuming one can identify a hill on a map (contours) or be able to identify the hill/road from the profile, having the option to 'exclude' the road/region from the route? I guess that's a re-statement of what you wrote above...

Been having a think about this and how I would do it - it would definitely be possible but the quality of the results you would get would depend on how meaningful the parameters you supply it with are.
For instance, you would have to have a method of associating height gain with 'cost'.
The actual routing algorithm itself doesn't deal in distance, it deals in the more abstract concept of 'cost', in other words the 'effort' of traversing from one node to another. Often on a 2D graph (map) this cost can simply be the distance (for a shortest-distance implementation), or the distance divided by the road speed for a fastest-time implementation.

So what this means is that in order to choose the flattest route, it would have to have some way of munging the height gained (and possibly even also the gradient) into the cost as well as the distance. In other words, you would have to be able to say, for instance, that each 1000m climbed is the equivalent of an extra 5km.

Panoramix

  • 50 61 6E 6F 72 61 6D 69 78
  • Suus cuique crepitus bene olet
    • Some routes
Re: GPX editor / route planner wish lists?
« Reply #42 on: September 21, 2012, 03:23:41 pm »
Wouldn't the best way to measure this simply be spent energy?

I am sure that there are around formula that relate speed, gradient and power.

simonp

  • Omnomnomnipotent.
Re: GPX editor / route planner wish lists?
« Reply #43 on: September 21, 2012, 03:34:45 pm »
Varies per rider and there are a lot of factors. Node to node costs probably need to be pre computed for speed for routes of any significant distance.

Panoramix

  • 50 61 6E 6F 72 61 6D 69 78
  • Suus cuique crepitus bene olet
    • Some routes
Re: GPX editor / route planner wish lists?
« Reply #44 on: September 21, 2012, 05:04:40 pm »
Varies per rider and there are a lot of factors. Node to node costs probably need to be pre computed for speed for routes of any significant distance.

Yes, agreed but you could assume an average rider and replace the node to node distances by node to node calories or kJ. The computed number of calories wouldn't be very accurate but the routing use the data as a relative number and it would favour flat routes.

And you could indeed make it extremely complicated by integrating wind, road surface actual rider weight, Cx (even Cy for side winds)....

Euan Uzami

Re: GPX editor / route planner wish lists?
« Reply #45 on: September 21, 2012, 05:32:48 pm »
Varies per rider and there are a lot of factors. Node to node costs probably need to be pre computed for speed for routes of any significant distance.

Yes, agreed but you could assume an average rider and replace the node to node distances by node to node calories or kJ. The computed number of calories wouldn't be very accurate but the routing use the data as a relative number and it would favour flat routes.

And you could indeed make it extremely complicated by integrating wind, road surface actual rider weight, Cx (even Cy for side winds)....

It's a good idea and intuitive for cyclists for 'cost' to be given by 'energy required', and for climbing, it's relatively easy to assume an average weight to give you the potential energy gained.
But if you're using energy for cost, then you have to work out the energy required to go along on the flat. This is what's less easy. Wind resistance is proportional to speed squared (I think?) so you could assume, measure or calculate this for an average rider. But there are several problems with this,
(a) it would have to be relatively accurate compared to the potential energy gained for climbing otherwise it would either place too  much importance on climbing, or on distance, and
(b) for the 'flat' cost to be static (and thus for the router to return a route in an acceptable time), you would have to assume that the rider is always going at their average cruising speed. This assumption isn't always valid, for instance in town, having to re-accelerate again after stopping at traffic lights consumes more energy than when batting along a country lane.
Could you take account of this? Possibly, but it would involve each node traversal having a speed associated with it, which would be a function of the previous nodes, and then you get into what parameters of the previous nodes is it a function of, if not just the length ... and it becomes a bit of a nightmare.

Cut short, you don't want to make it too complicated otherwise it would be impossible to fathom why it had chosen a particular route.
What I was thinking of is a simple slider that relates importance of height gain on cost to that of distance. So for example, I might like hills, so I could just move it and see how the route changes.


Panoramix

  • 50 61 6E 6F 72 61 6D 69 78
  • Suus cuique crepitus bene olet
    • Some routes
Re: GPX editor / route planner wish lists?
« Reply #46 on: September 21, 2012, 10:50:05 pm »
Varies per rider and there are a lot of factors. Node to node costs probably need to be pre computed for speed for routes of any significant distance.

Yes, agreed but you could assume an average rider and replace the node to node distances by node to node calories or kJ. The computed number of calories wouldn't be very accurate but the routing use the data as a relative number and it would favour flat routes.

And you could indeed make it extremely complicated by integrating wind, road surface actual rider weight, Cx (even Cy for side winds)....

It's a good idea and intuitive for cyclists for 'cost' to be given by 'energy required', and for climbing, it's relatively easy to assume an average weight to give you the potential energy gained.
But if you're using energy for cost, then you have to work out the energy required to go along on the flat. This is what's less easy. Wind resistance is proportional to speed squared (I think?) so you could assume, measure or calculate this for an average rider. But there are several problems with this,
(a) it would have to be relatively accurate compared to the potential energy gained for climbing otherwise it would either place too  much importance on climbing, or on distance, and
(b) for the 'flat' cost to be static (and thus for the router to return a route in an acceptable time), you would have to assume that the rider is always going at their average cruising speed. This assumption isn't always valid, for instance in town, having to re-accelerate again after stopping at traffic lights consumes more energy than when batting along a country lane.
Could you take account of this? Possibly, but it would involve each node traversal having a speed associated with it, which would be a function of the previous nodes, and then you get into what parameters of the previous nodes is it a function of, if not just the length ... and it becomes a bit of a nightmare.

Cut short, you don't want to make it too complicated otherwise it would be impossible to fathom why it had chosen a particular route.
What I was thinking of is a simple slider that relates importance of height gain on cost to that of distance. So for example, I might like hills, so I could just move it and see how the route changes.

I wasn't suggesting to make it complicated, for the calories thing to be workable, you would have to keep it as simple as possible, the easier would probably be a hash table that for each gradient give a number of calories per km. I suppose you really want to, you could also say twistiness coefficient = travelled distance / (Node A to Node B distance as the crow flies) and then make up an empirical coefficient to guesstimate the extra effort.

I won't blame you for using your method though as coding a routing algorithm isn't trivial!

simonp

  • Omnomnomnipotent.
Re: GPX editor / route planner wish lists?
« Reply #47 on: September 22, 2012, 01:11:43 am »
Naismith can be applied to cycling. It's not as good as for walking but:

X minutes per km

Add Y minutes per metre climb

Subtract Z minutes per metre descent

Then allow rider to set flat speed and weight and choose X, Y, Z based on these. Each edge of the graph has length and climb and descent. If edges only know about height at nodes, consider adding nodes at significant summits or troughs.

I can estimate my moving time on Audax events reasonably well using Naismith.

Re: GPX editor / route planner wish lists?
« Reply #48 on: September 22, 2012, 08:38:24 am »
I know absolutely nothing about the technology of using the Google data, but these "solutions" to the gradient effect seem overly complex.  Could one not just adjust the distance for the gradient, so a 10% slope was deemed to be 10% further (or 10% slower if you are using times rather than distances)?  Would you be prepared to go 10% further to avoid a 10% slope; an extra 500m to avoid a 5km climb?  It might need a multiplier, but probably not as a rough and ready solution.  And I presume you would ignore the "benefit" of the downs.


frankly frankie

  • I kid you not
    • Virtual Alps
Re: GPX editor / route planner wish lists?
« Reply #49 on: September 22, 2012, 09:39:51 am »
So the 'cost' attached to roads would have to be directional.
Anyhoo, I suspect we're reinventing the wheel here ...
"This is a complex subject, with a need for more than one highlighter pen."