Tweet
Noting the computational efficiency of the Freeciv AI at expanding cities and infrastructure, I decided to recompile a CVS snapshot with the AI "desire to build" a settler set to 10 out of 100 by modifying server/settlers.c. Then I started up civserver with AIFILL set to 4.
In trial runs with logging enabled I noticed that the AI built far fewer cities, unsurprisingly. The surprise was a repeatable pattern that began to emerge. One AI civ develops *in tandem with the course of human history*. The game years in which keys techs are discovered and Wonders are built by that civ correspond to their historical analogs. The other two civs were second- and third-world equivalents, lagging somewhat or very far behind the winning AI. As for my civ, I did nothing except to click "Turn Done".
Anyhow, Freeciv should be much more accessible to me now, and I plan to play through with the above hack. If I may gripe a bit, the server console help feature is incomplete. There is no obvious explanatory text for setting the AI difficulty, for example. The server settings (such as AIFILL) are well documented, though.
In trial runs with logging enabled I noticed that the AI built far fewer cities, unsurprisingly. The surprise was a repeatable pattern that began to emerge. One AI civ develops *in tandem with the course of human history*. The game years in which keys techs are discovered and Wonders are built by that civ correspond to their historical analogs. The other two civs were second- and third-world equivalents, lagging somewhat or very far behind the winning AI. As for my civ, I did nothing except to click "Turn Done".
Anyhow, Freeciv should be much more accessible to me now, and I plan to play through with the above hack. If I may gripe a bit, the server console help feature is incomplete. There is no obvious explanatory text for setting the AI difficulty, for example. The server settings (such as AIFILL) are well documented, though.