Lars Vilhuber writes
for development including undergraduate RAs etc. (regardless of how good they are), I prefer to do the development elsewhere. I have a bunch of virtual servers. If you let me know what collec.repec.org is running (ubuntu, debian, centos, opensuse, etc.) I can replicate that on the development server.
That's good. Load on katri (the server we use) is very high. It maintains now close to 1 billion paths. I run debian testing. If you set up the server, I suggest you create an "icanis" account, and I can rsync the data to you.
The visualization itself may have some requirements that we'll need to figure out. Right now, exploration is in pure javascript, but that may not scale well (that's part of the project to figure that out).
The CollEC site uses XSLT, so thet data that you generate needs to be in XML to fit. Sure, it could point to graphics say is a different format, or to files with BTW I was last year in Ithaka. Too bad I did not think about contacting you. -- Cheers, Thomas Krichel http://openlib.org/home/krichel skype:thomaskrichel