some years ago, i made route application using local map data. at that time, dijkstra's algorithm was a perfect solution. now I'm downloading 'asia.osm.bz2'. it is 2.3 Gbytes.. the size is reasonable for recent mobile external memory.
i'm gonna make sample route program and test on my n900. :) there are some open sources for that, but i do make it from scratch. okay.. as everybody said. local feature is needed. i'll focus my brain to that feature for the time being.. Thanks. > Date: Sat, 30 Apr 2011 11:33:51 +0900 > From: Carsten Haitzler (The Rasterman) <ras...@rasterman.com> > Subject: Re: [E-devel] splitting elm_map into a new library > To: David Seikel <onef...@gmail.com> > Cc: enlightenment-devel@lists.sourceforge.net > Message-ID: <20110430113351.427703e5.ras...@rasterman.com> > Content-Type: text/plain; charset=US-ASCII > > On Sat, 30 Apr 2011 08:56:11 +1000 David Seikel <onef...@gmail.com> said: > > > On Sat, 30 Apr 2011 02:12:49 +0900 sangho park <gouach...@gmail.com> > > wrote: > > > > > as basic concept, route information can be acquired by service, not > > > local data. > > > we can extend route for shortest path using local backing store, but > > > it's just option. > > > yes..yes.. it'll be very good feature that uses local map and route > > > data, but we should overcome so many license issues. > > > > Personally I think it's absolutely critical to be able to do local > > routing. That's my main use case, navigating out in the wild when I > > have no data connection. "Semi manual" routing is OK, that's even > > a desired feature. Pick out a few waypoints, draw some route > > segments, tell it to follow this road or river from waypoint C to > > waypoint D, but keep the A to B to C and D to E segments... > > > > Perhaps, if we have to, precalculate route segments when there is a data > > connection, store them; then manipulate them later, adding or removing > > them to a particular route. > > > > /me does not want to be stuck out in the middle of nowhere with no > > routing ability. > > to me this all seems like it would be best done as client-server locally > too. > unlike remote map server, it has 1 shortcut - you don't DOWNLOAD map tiles. > you > are directly pointed to the file. this way routing can be offloaded to the > local mapserver (and calculating a route can take a long time, and need a > lot > of route data, so this makes sense). all the abstraction needs is the short > cut > ability for tiles. > > someone still then has to write that local server and routing engine :) do > you > want to? :) (a local tile server would be trivial, but routing... that's > another matter as technically it may not be that hard as its just solving a > graph, the problem is having the right graph data with the right weighting, > and > then a whole tonne of "fuzzy logic" thrown in to cover up "odd solutions" > that > may come out of the graph that a human would never make). > > -- > ------------- Codito, ergo sum - "I code, therefore I am" -------------- > The Rasterman (Carsten Haitzler) ras...@rasterman.com > > ------------------------------------------------------------------------------ WhatsUp Gold - Download Free Network Management Software The most intuitive, comprehensive, and cost-effective network management toolset available today. Delivers lowest initial acquisition cost and overall TCO of any competing solution. http://p.sf.net/sfu/whatsupgold-sd _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
