Instead of setting that to 0, it might help to set this value to be as high as it can go (1.0 if possible). If that doesn't help then there is something going on that I didn't notice in the code.
The file that uses this in RoM:AND is Assets\Python\Revolution\Gameready\BarbarianCiv.py.
I didn't examine this in much detail, but...
It should be the percentage of teams that must already know a tech for the barbs to get the tech, unless the regular barbarians already have it in which case they automatically get it. But that only happens in a "non new-world" situation ("bNewWorldScenario" set to false - I think this is set to true if they are spawning on a continent that has no other non-barb city on it or set to false if there is a city belonging to a non-barb civilization), otherwise it is not used at all. There is also a separate section of code that can give them a few techs, in addition to all the ones the barbs have, in a way that is not related to this value (something in the 2 to 5 range, depending on other settings that I didn't check to see how they were set), and yet another setting somewhere ("BuilderBonusTechs", or something like that, based on the name of the function used to get the value) that can add some more in some cases.
If it is a situation where it is being used, having it set to 0 seems like a bad idea since it looks like the new barbarian derived civ should then get all techs, including those that nobody else has yet (like, say, Fusion and Future Tech) since they get a tech if the fraction of teams that know it is >= this value (nobody else already knowing it makes the number of teams that know it 0, but 0 out of N civilizations is a fraction of 0 which is >= 0 so they'd get it anyway).
) ill go ahead and just ask. Does changing a xml (like, lets say GlobalDefines.xml, i increase how many wonders i can build per city) influence a saved game ?
