Did a lot of research into Java RMI. The current plan (DCOMP-type design) has several flaws in it that inhibit applications from doing some slick things without a lot of pain on the part of the programmer. Also--with the current trend in companies suing each other about Java applets in web-browsers, it might be a good thing for me to change the system design now before I get too much further.
Since DCOMP relies upon the client browsers to host the applets, I will almost certainly run into problems with compatability between browsers and platforms. It might make more sense to either target a browser/platform combination or to switch from applets to applications.
I'm moderately irritated that I've done a lot of work cleaning DCOMP and moving it to the next stage and it looks like I might end up building a new system entirely from the ground up using different stuff. However not all is naught--the work that I have done so far on DCOMP has given me a better idea of what the system I want to build needs to accomplish and what components and functionality is necessary.
I need to do a lot of research into various possibilities before I go off to programming again. As they will most certainly affect the system design again.