View Single Post
Posts: 317 | Thanked: 787 times | Joined on Oct 2009 @ Krakow, Poland
#65
Originally Posted by arne.anka View Post
what about using the maps navit uses?
there are preprocessed maps available of all of osm and the maps need by far less space than the tiles maep or tiles baesd approaces use (germany is at about 1.6g in the current osm navit map while the tiles will be a multiple of that).
besides, navit is open source, you could easily re-use the code to read those maps.
I know vector based maps have many advantages. But there is already plenty of work to do for tile approach (see post #50). For now let me follow that roadmap and then we will discuss whats next.

But let me clear one thing. Vector based navigation has been researched and developed by companies for about a decade now and it would be very difficult to catch up in short time (read: one year). I am not saying it is impossible but still very hard. I preffer to look at simpler tile based approach as some kind of warranty that it is doable in reasonable time. For now I can olny say that I would like to approach vector based navigation when CloudGPS is "complete".