Incorrect magnitude of Witwatersrand Gold Rush national immigration modifier

[Sillken]

In the event_modifiers.txt file, the witwatersrand_gold_rush modifier (which is to be nation wide) uses immigrant_attract instead of global_immigrant_attract. On top of that, it's supposed to give you 800% immigrant attraction. I imagine whoever coded this either thought this was the modifier johannesburg itself got or kept testing random amounts without noticing they were using a different effect, so you might want to tone that down when fixing it.

[moretrim]

A solution is not so obvious:

https://github.com/moretrim/ccHFM/blob/a5657881e9d803e863fcdd0a514076ae90be8776/ccHFM/events/Goods.txt#L224-L235

• the gold_rush province modifier is in charge of domestic migration, in the form of immigrant_attract and life_rating
• the witwatersrand_gold_rush country modifier is also used to identify the country as the potential target for a Second Boer War (more or less, depending on other happenings)—it is removed if the GP takes the CB

So it is suggestive to think of the latter as the modifier in charge of international migration very deliberately, not by accident. On the other hand as mentioned +800% does blow the scale entirely. So what to do? At a quick look Boer republics can get the following:

• trek_boers: +310% over 730 days
• boer_republic: +100% over 7300 days (there is overlap with the above, best to ignore to keep things simple)
• gold_rush_country: +50% over 5475 days
• mineral_revolution: +200% over 1825 days
• access_to_the_sea: +50% with indeterminate duration, though apparently designed to mitigate the full -150% from landlocked_nation

Numerically as things are the gold rush would have 6 times the impact of boer_republic, the next most powerful modifier, if translated to international migration. It’s tempting to pick one of the other values as a reference and go with that, e.g. gold_rush_country is the country modifier that pairs with the gold_rush province modifier and has the same duration, it’s a great match.

However HFM was designed to add these delicate pop & migration tweaks to better reflect historical patterns, so ideally we would have historical numbers to use as a reference. Consequently I’m going to leave this issue open without a milestone.