The first warning message is in: a “trespass” warning from the AI when you have ships orbiting his colonies. The severity of the warning is calculated based upon the firepower of your fleets vs. the planetary shields, and is doubled if you are at war or if he is xenophobic. There’s a minimum severity so that unarmed ships can potentially trigger this warning.
Since this is the first possible warning of many to come, it involved setting up some code infrastructure to look for possible warning situations initiated in the current turn, determine their severity, and then issue a warning if their accumulated severity exceeds a value. If so, you are issued a warning about the most severe violation.
Since trespass violations are checked every turn, you will be warned continuously if you do not retreat your ships. Eventually the accumulating diplomatic penalties could result in a declaration of war.
Since the diplomacy system in MOO1 is overly simplistic and easily exploited, I am reworking the internals extensively although the outward effects to the player should appear the same.
To avoid AI spam to the player, an AI opponent will initiate no more than one message to the player per turn. The most severe messages are checked first (declare war?) and then progressively less-severe messages are checked (break treaty? issue warning?) until finally the AI is cleared to offer a request such as a treaty.
This means that constantly irritating an AI to cause warnings will result in him never proposing anything constructive. Why should he? You’re being a jerk!
Here’s a screenshot of an example trespass warning (note that the Meklar graphics are currently used as placeholders for all of the races).