On Tue, May 1, 2012 at 10:39 AM, Ian Sergeant <inas66+...@gmail.com> wrote: > However, certainly in New South Wales this is likely not to be the > case. Any cyclist around Sydney will be all too familiar with the > cycle lane that disappears, leaving you on a 80km/h full-on three lane > highway, with some semblance of a cycling facility possibly resuming > some distance up the road.
Sure - but we cover that by being precise about cycleway=lane etc. In my area, I often map to a precision of a metre or so. > It comes back to what a cycle route is. Cyclists who just want the > shortest trip between two points don't need to look at cycle routes, > they can just go. A cyclist following a cycle route is looking for > some kind of cycling amenity, whether than be quiet, flat, lanes, etc. I think you're focusing a bit too narrowly on one kind of route. There are other kinds of routes: historical interest, fast commuting routes, easy to navigate etc. As a cyclist, I'm interested in knowing about a route that I can follow without a map. Whether or not it has bike lanes, quiet streets etc is also interesting - but is not the only consideration. (Similarly, a published route that has bike lanes, quiet streets etc but no signage could also be worth including.) >Where there is a > gap in the route, we should make that apparent, and not disguise it. Yep - through the cycleway=* tags, not gaps in the route relation. If it helps to explain my view, I see the network of intersecting routes as a valuable navigational tool for cyclists - knowing "I jump on this trail, take it up to that road, then switch to this other trail, and 20 minutes later I'm there". Having gaps in the (conceptual) network makes that work less well. > In other words we're human mappers. Official cycle routes are often > wrong, as other mapping services are often wrong. And yes, we should > be able to correct cycle routes when they don't correspond to an > amenable through route on the ground. It's a fair jump from this > position to being able to enter personally preferred routes, or that > we should automatically link two route sections just because they are > close. Ok, on those two things: 1) Personal routes: my actual (toned-down) suggestion is that such "personal" routes might be of value in the absence of more official ones. (Which also solves your slippery slope argument...) 2) Implicit connections between routes. Let's take this concrete example: http://osm.org/go/uG4JzYKDD-?layers=C The Rosstown Rail Trail runs along that bike path along the rail line then down Freda St etc. I don't think there is any signage for the Anniversary Trail (aka Outer Circle Trail) until Boyd Park. So technically there is a gap of a couple of hundred metres between them. But clearly, they're intended to link - there is even this bike/pedestrian path connecting the RRT to Poath Rd. So I'm comfortable interpolating a route between them. What do you think? Steve _______________________________________________ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
