Current weightings are for bicycle travel. That is to say, it's ignoring freeways and preferring bike paths.
If the SPT is spreading out in a diamond shape, it's because the SPT has a diamond shape. Also, it's minimizing distance in geographical units (hah!), which will skew things at Seattle-ish latitudes. -B On Mon, Mar 12, 2012 at 12:21 PM, Steve Coast <[email protected]> wrote: > Very pretty > > * 520 bridge doesn't work for some reason, click in Bellevue and everything > traverses the i90 bridge > * ie 9 & 10 don't work > * would be nice to have more graduations of color > * confusing to me why the colors spread in a diamond shape implying a square > grid rather than a circle > > -----Original Message----- > From: [email protected] > [mailto:[email protected]] On Behalf Of Brandon > Martin-Anderson > Sent: Monday, March 12, 2012 9:12 AM > To: osm-dev; [email protected] > Subject: [Geowanking] shortestpathtree.org - a tool for quickly checking OSM > data integrity > > Behold! I made a thing. > > http://shortestpathtree.org > > It creates shortest path trees, which are pretty, and have a variety of uses. > My favorite use is quickly and phenomenologically checking OSM referential > integrity across entire cities. Also, potentially, it can tell you how to get > places. Tell me how you like it. > > Colophon, for the interested: > Server and client-side code is at https://github.com/bmander/vtp. I took > Migurski's city extracts in PBF format and popped them into a Mongodb > instance using a homebrew script in node.js. Then I applied a series of > map-reduce runs to slice the ways at shared intersections, and to collect > them into tiles. This is slow, but there's some home of parallelization. A > simple node.js script serves the vector tiles to the client, where all > routing is done; printed to a homebrew canvas-based client. The disadvantage > is that routing is slow for you. > The advantage is the server doesn't have to do anything except hand out > tiles, which, ideally, should be pretty small. > > -B > > _______________________________________________ > Geowanking mailing list > [email protected] > http://geowanking.org/mailman/listinfo/geowanking_geowanking.org > > _______________________________________________ Geowanking mailing list [email protected] http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
