Do you know how accurate the geonames data is for Kosovo? It's not always terribly good, so it's worth checking.

I didn't think it made any difference to opencyclemap whether it's a tag or a relationship. I tend to go for tags when it's just a local route and a relationship when it's a sustrans route.

GPS is generally pretty terrible at altitudes, it just gives a very rough idea. Also, it'll be altitude above mean sea level not actual sea level where you are.

OK, landuse=industrial has been added, let's give it some time to render and then see what difference that makes.

Cool,

I really wish we could find a suitable source of campsites for the UK. Combined with the opencyclemap that would be fantastic!

Oh and I meant csv2osm, not svn2osm.

I wish you could differentiate ncn routes from rcn routes just by the number. I've been mapping ncn_66 recently, and whilst riding it keep bursting into song

(It goes up through Bierly, down to West Bowling, Bradford City looks oh so pretty...)

It is possible to import from a csv file, which in turn can be exported from excel. However, for this to be ok, the original data would need to be correctly licensed and your calculations of latitude/longitude would have had to be done using public knowledge (ie. not geolocated using Google or similar)

If those requirements have been met (and I know I get annoyed at finding seemingly open data then failing to get it licensed correctly) then there is a script in svn called svn2osm. It's very old and needs updating to API 0.6 though.

The trouble with a spare inner tube is I'd need two - one for each wheel since they are different sizes. And yes, I know, standardising my wheel sizes would also be sensible.

In future however, I do intend on carrying puncture repair kit *and* spare inner tube since a puncture repair kit doesn't take up much space.

Chris - If it really was unsuitable for motors then yes, a track is the solution. However, it was better quality than at least one other road I rode along.

Ed - motor_vehicle=unsuitable sounds like it's the solution, I was also thinking about the designation key since it's not actually unsuitable it's just designated as such.

I've seen a few places where duplicate roads have appeared but didn't think anything of it at the time. Perhaps one of the editors has a rare bug which creates duplicate ways.

I like the look of that logger, you'll have to do a proper review after using it for longer. What mode are you using it in? What's the accuracy, speed and sensitivity like?

It's certainly nice to see somebody working on an alternative to OpenLayers. I don't have time to do a full test, but straight away I can tell you it fails in IE 7 (object expected on line 15)

Where does xapi fit into this? It is just a part of the api or does it fit in somewhere else?

Awesome, if the weather stays dry I might be tempted to go and ride some of those tomorrow.

Cool! Shame it's not working with Open JDK without Compiz though.

And if re-routing cycle routes was as easy as editing the OSM data then we could make them much nicer to ride.

Added all tagged routes in the sixties to the relevant relations (except 67 which doesn't seem to have a relation). I wish people would do this themselves when editing routes.

Route 65 should now not be too hard to finish mapping between Middlesborough and York, presuming it's signposted ok. Which just leaves a stretch from Selby out towards Hull where Chilly has been mapping it.

3ld - the road changes names between the two mini roundabouts, so the linking road (which doesn't exist) could have either name and still be incorrect. I think Thomas is correct, it doesn't need a name.

Looks good to me, Amesbury was a bit messy so I've done some tidying up of existing features so it's easier for you to start mapping properly.

Absolutely perfect thank you!

As for the bug, ant clean and then rebuilding worked. I now know to try that in the future before filing bug reports.