Yet Another Cycling Forum
General Category => The Knowledge => GPS => Topic started by: David Martin on 26 January, 2009, 11:49:04 pm
-
Should we be encouraging the CTC to use/enhance Open Street Map? It would provide an extremely valuable resource, particularly in the ability to grade routes and add details of barriers, steps etc.
With the input of CTC members it could make a big impact on the blank map areas as well as inspiring people to look at the roads and paths less travelled.
..d
PS. I mooted the idea of the council releasing their GIS data into the public domain during a discussion at the local access forum on heritage paths. The idea was not met with a scowl and a snort of distrust but with genuine openness.
-
Barriers and steps on cycle routes matter - that's the quality of data that makes the difference.
-
PS. I mooted the idea of the council releasing their GIS data into the public domain during a discussion at the local access forum on heritage paths. The idea was not met with a scowl and a snort of distrust but with genuine openness.
The trouble is that even if the council are willing, the data will very likely have been digitised against an OS background map and thus be infected by the OS copyright.
You would need to go through it with the OSM people as well as the council.
It's certainly worth getting it publicised by the CTC, both on their forum and the magazine from the point of view of roping in extra bodies for surveying.
I would say that it would be worth getting them to switch their website over from Google Maps, but I wouldn't be sure they are up to it.
-
PS. I mooted the idea of the council releasing their GIS data into the public domain during a discussion at the local access forum on heritage paths. The idea was not met with a scowl and a snort of distrust but with genuine openness.
The trouble is that even if the council are willing, the data will very likely have been digitised against an OS background map and thus be infected by the OS copyright.
You would need to go through it with the OSM people as well as the council.
It's certainly worth getting it publicised by the CTC, both on their forum and the magazine from the point of view of roping in extra bodies for surveying.
I would say that it would be worth getting them to switch their website over from Google Maps, but I wouldn't be sure they are up to it.
In Scotland there would be a lot of blank bits to the map. Google maps is a good solution at present. When OSM has more complete coverage, it would be worth a transition.
Even if people just upload GPS traces, that would be an advance on the present situation.
..d
-
The issue I would see with it, although I am in general favour, is that it is a wiki.
I can see "CTC" or "Cyclists like me" putting notes on there such as "Do not use this route, it contains steps, the best cycle route is to use this quiet road", which would then be removed/replaced by a sustrans/cycle campaigner with "Ideal cycle route".
It would end up like the wikipedia pages on cycle paths, where there is a label of "contentious subject" (or there was last time I looked), and constant changes would be made to open street map as to whether a cycle route was a cycle route; irrespective of whether it was a cycleable route or not a cycleable route.
-
The issue I would see with it, although I am in general favour, is that it is a wiki.
I can see "CTC" or "Cyclists like me" putting notes on there such as "Do not use this route, it contains steps, the best cycle route is to use this quiet road", which would then be removed/replaced by a sustrans/cycle campaigner with "Ideal cycle route".
It would end up like the wikipedia pages on cycle paths, where there is a label of "contentious subject" (or there was last time I looked), and constant changes would be made to open street map as to whether a cycle route was a cycle route; irrespective of whether it was a cycleable route or not a cycleable route.
But it soes not, unlike wikipedia, contain opinion. It contains facts. If there are steps, then there are steps. No debate over that. Yes the route may be NCN4567 but the steps are steps.
And when the steps are replaced by a ncie graded ramp then that can also be noted, rather than waiting years for the sustrans map to be reissued.
There may be debate over whose miscalibrated GPS is to be believed, but that is a minor issue.
Most debates on OSM are due to politically charged place naming - Cyprus being a case in point.
..d
-
I agree, more people generating GPS traces outside towns would be good for open street map.
You're right about steps being fairly clear, but...
I think in marking "rideable" you end up wanting a lot of well defined tags to be able to see whether a route is: decent tarmac, sort-of-hard surface track, muddy / rutted path, this footbridge isn't rideable but still quicker and nicer than alternative ways of getting to the other side, this section is interestingly rooty on a mountain bike, just a little jump, here is an underpass that the local yoof get wasted in of an evening but is OK in the day etc. Somewhere in there you probably end up with opinion, or at least needing "type of bike" and a second opinion.
Sounds like an interesting tagging project. Maybe if we could carry accelerometers as well as GPS a scale of roughness could be automatically extracted. But, that's probably just because I could claim developing that as my day job ;D
-
I agree, more people generating GPS traces outside towns would be good for open street map.
You're right about steps being fairly clear, but...
I think in marking "rideable" you end up wanting a lot of well defined tags to be able to see whether a route is: decent tarmac, sort-of-hard surface track, muddy / rutted path, this footbridge isn't rideable but still quicker and nicer than alternative ways of getting to the other side, this section is interestingly rooty on a mountain bike, just a little jump, here is an underpass that the local yoof get wasted in of an evening but is OK in the day etc. Somewhere in there you probably end up with opinion, or at least needing "type of bike" and a second opinion.
There are already five grades of unsurfaced track. I think Grade 5 is packhorses only. Grade 4 is just about passable by a good 4wd. 1 is smooth, almost like tarmac.
There is also the bicycle= tag for whether access is possible on a bike. I tend to under-annotate as I am not yet aware of the full richness of OSM tags (I think they need developing into a proper ontology - again I could then claim that almost as part of my work).
..d
-
I've never spotted this information when I've looked, so it seems I need to spend a bit more time with it.
-
I've never spotted this information when I've looked, so it seems I need to spend a bit more time with it.
The wiki has a lot of information. There isn't really a very good beginenrs guide to tagging your routes/features.
In principle it should be possible to plan a route that avoids main roads, avoids excessively bumpy off road tracks, will go via steps if there are less than four in a flight, and will pass a cafe every 20 miles.
However, the annotation isn't up to standard there for that.
Yet.
..d
-
And when the steps are replaced by a nice graded ramp then that can also be noted,
Just picking this specific example, there will always be "relevant opinion". The NCN route from Chepstow to Bulwark is a graded, tarmac ramp. With a gate, leaf-mulch-and-debris-a-plenty, and 0.5mph switchbacks. I would rather take the direct road route (especially in company).
(I would never have found this route without the Sustrans site - to most it is just a legend ...)
-
The wiki has a lot of information. There isn't really a very good beginners guide to tagging your routes/features.
I find the wiki page 'Map Features' quite helpful
http://wiki.openstreetmap.org/wiki/Map_Features (http://wiki.openstreetmap.org/wiki/Map_Features)
"This page contains a core recommended feature set and corresponding tags."
But I rather agree with Nutty that there just seems so much scope for inconsistency of approach.
Even where guidelines exist they are rather Anglo-centric. Deciding between 'secondary' road and 'tertiary' in rural France, can be a bit hit-or-miss even with good personal knowledge of the road - and yet it makes quite a visual difference on the map render.
And I've come across some French cathedral towns tagged as 'village', and a place I often stay in, that has at least 6 large hotels arrayed along the High Street and twice as many restaurants, is tagged as a 'hamlet'.
-
I find the wiki page 'Map Features' quite helpful
http://wiki.openstreetmap.org/wiki/Map_Features (http://wiki.openstreetmap.org/wiki/Map_Features)
"This page contains a core recommended feature set and corresponding tags."
But I rather agree with Nutty that there just seems so much scope for inconsistency of approach.
There is much inconsistency at present, even amongst one editors approach. But this will evolve, style guides will be tightened and annotation will be regularised over time. That is the joy of a wiki.
The best way to improve it is to, well, improve it and join in the discussion. An annotation workflow plugin to eg JOSM would be really useful to guarantee minimum annotation standards and consistency.
..d
-
And when the steps are replaced by a nice graded ramp then that can also be noted,
Just picking this specific example, there will always be "relevant opinion". The NCN route from Chepstow to Bulwark is a graded, tarmac ramp. With a gate, leaf-mulch-and-debris-a-plenty, and 0.5mph switchbacks. I would rather take the direct road route (especially in company).
(I would never have found this route without the Sustrans site - to most it is just a legend ...)
My point exactly. I can imagine a route where two groups of contributors constantly fight over whether it is Grade 5 or Grade 2.
Take for example the Piste runs at the last ski resort I was at. We did a black "oooo scary", then later in the day did a steeper and harder blue that everybody said should be harder than black. Some folk were complaining about a green - only accessible via red runs - yet another green was so flat that skiers were walking.
If we had mapping via a wiki with arguing contributors, would the end result be useful or usable? I'd hate for somebody to try to cycle a "Grade 2 unsurfaced" then find it was actually a grade 5 that they can't ride. Or for them to avoid a grade 5 and instead use a major road, when in fact the path is a Grade 1.
-
If we had mapping via a wiki with arguing contributors, would the end result be useful or usable? I'd hate for somebody to try to cycle a "Grade 2 unsurfaced" then find it was actually a grade 5 that they can't ride. Or for them to avoid a grade 5 and instead use a major road, when in fact the path is a Grade 1.
Well, there is an objective description of the different grades. It would be hard to slip two grades, though one might find that a 2 is a 3 in some parts or a 4 is a 3. But you are unlikely to determine that a 3 is a 5.
..d
-
Well, there is an objective description of the different grades.
... if the contributor bothers to read the guidelines.
Or the Powers That Be may move the goalposts over time - adding or removing a grade.
Don't get me wrong - I think its a great and fascinating project - but its as well not to be blind to the possible pitfalls, or to the fact that many of the contributors are (like me) still on the steep bit of the learning curve, or to the fact that its a lot of different things to a lot of different people.
-
Google Maps and even OS don't get it right regarding the exact road details so what chance the general public. Looking on GM where I am all the forestry tracks are shown as normal roads, so if you planed a route with your car up there you would be in for a surprise.