On Thu, Feb 26, 2009 at 5:51 PM, Dossy Shiobara <do...@panoptic.com> wrote:
> > On 2/26/09 7:19 PM, Nick Arnett wrote: > >> Yes, you probably don't, unless you just want to do it occasionally. >> I'm working on some graph manipulations like that... the challenge is >> that Twitter is so open that the graphs are enormous for many people. >> This isn't a very hard problem on a small scale, but generalizing it >> looks like it will take a lot of resources. >> > > Huh? Now that Twitter has added the two social graph APIs that return the > entire list of IDs, this kind of social graph inspection becomes trivially > easy: two API calls, and one set operation (intersect). Even at 1M > followers, if the two lists are sorted, this is an O(N) operation. I guess I read it as a second-order problem, but now I see how it was probably just the first-order intersection. Still, the second-order problems are interesting and difficult, I believe. Nick