1 Loads of coding dealing with trying to get my reference software to interact with pubmed.
3 Looks like I'm going to have to duplicate some php code in perl, as php just isn't powerfull enough, nor multi-thread capable enough to deal with pubmed's slow generation of sgml. (I'm going to multiplex to pubmed almost 100 connections, and doing it syncronously isn't going to work.)
5 Got to take care of making the jacked figures for this paper. Will happen very shortly.