I think that's a "we'll see". It shouldn't immediately affect it as cycle.travel takes a monthly download from Archies, so I can just continue with the final dump. As time goes on, though, that data will become more out-of-date.brianleach wrote: ↑31 Aug 2022, 2:12pm Will the Archies web page closure affect your site Richard???
When I first added the Archies data, it was much better than that available in OpenStreetMap. My sense is that in the intervening years OSM has gone a long way to catching up in campsite locations, if not in metadata (phone numbers/opening times). Certainly all the campsites near me are on OSM now and I'm sure that'll be the case in the German-speaking countries too. Whether it's true further afield, who knows, but I'll spend some time looking into it.
Exactly the latter! They cover a multitude of sins. I've had suggestions from German cyclists that c.t is too lenient on cobbles which I guess goes to prove you can't please everyone.RickH wrote: ↑31 Aug 2022, 6:26pm A probably mainly Northern-related question. I noticed a few, mainly short, stretches I often cycle, & get included in routes I plan, being marked as unpaved & it seems to be because they have the surface labelled as "cobbles". Strictly speaking they are setts - square or rectangular flat(ish) topped stones - but further delving shows that even Le Mans Crescent in the centre of Bolton, which has the surface marked as "setts", is marked as unpaved. Is this a deliberate choice in Cycle.travel or just a bit of "collateral damage"? I know the term "cobbles" can convey a wide range of meanings.
Internally, c.t's weightings are much more subtle than just paved vs unpaved - so it has different scores for "sett", "cobblestone", "unhewn_cobblestone" and "cobblestone:flattened", all of which can be found in OSM. One of the problems is indeed that OSM mappers often record setts as "cobblestone" which gets the lower ranking. I hope that in time the tagging will get more consistent.