On Wed, Jun 11, 2008 at 6:44 AM, Karl Newman <[EMAIL PROTECTED]> wrote:
> On Wed, Jun 11, 2008 at 12:35 AM, Simon Wood <[EMAIL PROTECTED]> wrote: > >> Hi, >> I have finally gotten around to uploading some of the Calgary Trails >> Project data... under the username 'catmp'. I made a slight mod the >> converter to remove attribution/RoadID on nodes within a way as this was >> making/will make cleaning up of traces difficult. >> >> I think there is a bug with the POLYGON stuff. Look in the Calgary area >> (BOW RIVER polygons), they are mostly right but 'jump' to nodes not in the >> original polygon. So it looks like something is up with the 'previously >> used' node stuff. >> >> For example (frigged the name to make it findable in output) >> --- >> [POLYGON] >> Type=0x48 >> Label=BLOW RIVER >> >> Data0=(50.93128,-113.99972),(50.93116,-113.99959),(50.93172,-113.99959),(50.93214,-113.99998),(50.93279,-114.00093),(50.93288,-114.00122),(50.93270,-114.00196),(50.93262,-114.00212),(50.93253,-114.00199),(50.93227,-114.00067),(50.93163,-113.99990),(50.93150,-113.99993) >> [END]RoadID=1459 >> --- >> ends up with nodes >> --- >> <node id='-1' visible='true' lat='50.93163' lon='-113.9999' /> - yes >> <node id='-2' visible='true' lat='50.93227' lon='-114.00067' /> - yes >> <node id='-3' visible='true' lat='50.91909' lon='-113.99369' /> - no >> comes from RoadID 1458 >> <node id='-4' visible='true' lat='50.93262' lon='-114.00212' /> - yes >> <node id='-5' visible='true' lat='50.93214' lon='-113.99998' /> - yes >> <node id='-6' visible='true' lat='50.93279' lon='-114.00093' /> - yes >> <node id='-7' visible='true' lat='50.93116' lon='-113.99959' /> - yes >> <node id='-8' visible='true' lat='50.93288' lon='-114.00122' /> - yes >> <node id='-9' visible='true' lat='50.9327' lon='-114.00196' /> - yes >> <node id='-10' visible='true' lat='50.91991' lon='-113.9935' /> - no >> comes from RoadID=1459 >> <node id='-11' visible='true' lat='50.92262' lon='-113.99522' /> - no >> comes from RoadID=1459 >> <node id='-12' visible='true' lat='50.9315' lon='-113.99993' /> - yes >> --- >> >> >> >> I am not 100% convinced of the tags, apparently the route up Turtle >> Mountain is 'bicycle=yes'... this is definately NOT the case!!!! Don't >> know whether this is source data problem, or with the parsing of the >> garmin attributes. >> >> Places uploaded: >> >> http://www.openstreetmap.org/?lat=49.5908&lon=-114.4056&zoom=14&layers=B00FT >> >> http://www.openstreetmap.org/?lat=49.5846&lon=-114.5279&zoom=13&layers=B00FT >> >> http://www.openstreetmap.org/?lat=49.3548&lon=-114.2812&zoom=13&layers=B00FT >> >> http://www.openstreetmap.org/?lat=49.0474&lon=-113.9209&zoom=13&layers=B00FT >> >> (these are revertable/removable quite easily as there are isolated on >> their own) >> >> Way too late to do much more on this tonight (errr.. this morning). >> Cheers, >> Simon. > > > I don't think I did much of anything with the polygon part--I mostly > focused on the polylines. I'll look into it but I'm not sure that I'll be > able to get to it right away. As for the bicycle=yes, I'll double-check the > parsing but I suspect it's a problem with the source data. Part of the issue > is that if it's not marked as restricted/forbidden in the .mp file, then I > had to assume that meant "yes". > > Karl > Okay, it was a one-line fix. I wasn't clearing out the routing nodes so the polygons were trying to reuse nodes from previously parsed roads. It's checked in to SVN. As far as Turtle Mountain, it's in the source data, so apparently someone thought bicycles were okay (at least not restricted). Don't forget bicycle can include mountain bike, too. I think there may be some proposals about distinguishing between them, but in any case there's not enough information in the source data to make that distinction. Karl
_______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
