I don't know if you have seen the recent question about RAND for DoC in one of the discussion threads, but it actually made me seriously think about doing this for the first time. So I skimmed the RAND code to find out how that would work. Stuff I found out: - it uses a modified version of the Terra.py script from BtS that is made to look more like an RFC map - code for dynamically choosing starting locations (among other factors, based on the proximity to similar civs) - randomizable starting dates with appropriate techs (!) - dynamically chosen barbarian spawns and independent cities - a rudimentary CNM for context sensitive city names - different UHV conditions that do not rely on specific map regions So basically everything that relies on map knowledge. (Edit: Rhye's readme on RAND attached, in case people want to know more about it.) A naive merge is unfortunately impossible because DoC has already diverged so far from RFC, and I don't really want to include more of Rhye's code here, so there'll be mostly manual work involved. There are also parts that can be copied as a whole (like the map script) with the downside that I know very little about map scripts. But there are also advantages because some aspects here intersect with stuff I want to do anyway. For example, RAND only spawns a selection of civs based on map size, and I want to implement some order of priority which civs can or should exist anyway. And dynamically determining techs for (re)spawns has already been discussed for DoC, so I can maybe work from the RAND code here and later reuse it. But first I'd like to know how many people care about this? Spoiler Important note for people who do not care about this : There's no need to tell me about it. I know what I should focus on first, and I'm going to do that. This is just an open-ended discussion, and I'm not really interested in that sort of comment.