Apple, Inc., Cupertino, CA
Maps C++ Software Engineer - Public Transit
As a member of the Transit Routing team, you will work on one of the most anticipated features of Apple Maps. You will design and implement functionality that will be used by millions of users worldwide. Being part of a small team, you will have an influence on the future product. You will be responsible for implementing new features and resolving complex issues. You will work on high-performance server code using C++ in a Linux environment. Work with engineers and QA to deliver high-quality routes to our customers. Challenge the status quo and improve the design of existing software. Participate in design and code reviews.
simpler than a restaurant: a coffee shop.
coffee machines have been around for quite some time, they are easy to operate and they produce decent coffee.
yet "the total number of coffee shops in the US increased by 70% between 2000 and 2005" [1]
AI seems necessary but not sufficient for automation...
I wonder what would happen if everybody who wants Google to continue Reader would start clicking on the Ads in Reader on a daily basis.
Oh wait, which ads...
Data structures called "distance oracles" handle such problems. However, most research results are for static graphs only.
If the graphs are grids (as in the cstheory.stackexchange.com question) they are planar and some dynamic data structures exist (unclear whether the constants are small enough for the application in question):
Exact shortest paths:
Jittat Fakcharoenphol, Satish Rao: Planar graphs, negative weight edges, shortest paths, and near linear time. J. Comput. Syst. Sci. 72(5): 868-889 (2006)
Approximate shortest paths:
Philip N. Klein, Sairam Subramanian: A Fully Dynamic Approximation Scheme for Shortest Paths in Planar Graphs. Algorithmica 22(3): 235-249 (1998)
I wonder why some of these sites would want to participate in an information exchange that seems rather asymmetric. I couldn't find detailed information on the service but I assume they have some kind of mechanism, similar to those in P2P systems, to "encourage" sites to contribute large amounts of information(?) or maybe sites can buy information without having to sell?