Experimental and Migration

Various topics about the game, the website, or anything else Freeciv related that doesn't fit elsewhere.
User avatar
Arbogast
Elite
Posts: 412
Joined: Tue Jul 16, 2013 9:09 am
Location: France

Experimental and Migration

Post by Arbogast »

I'm getting more and more into the Experimental ruleset which I'm liking it more and more.
But, the migration thing is bothering me more and more. I find it too strong : I'm losing about half of my cities by turn ~70 or so.
I found in the effects.ruleset that the value is 10 or 20.
My question is : Lowering that value, does that lowers the risk of migration ?
User avatar
Alien Valkyrie
Elite
Posts: 513
Joined: Sun Feb 10, 2013 10:21 pm
Location: Stuttgart, Germany

Re: Experimental and Migration

Post by Alien Valkyrie »

If all your citizens move away, maybe that's not migration being too powerful, but you not taking care of your people. Make sure to keep your citizens happy, build a wonder in each of your cities, don't found cities too close to other, bigger cities or maybe send whole groups of (up to 8) settlers to start new cities at greater sizes. Make sure the terrain around your cities is fertile so that the city has some food surplus (that factors into migration as well), and maybe try switching to more libertarian governments if nothing else works.

Oh, wait, NVM, you probably still have me blocked -.-
~ AVL
User avatar
dunnoob
Elite
Posts: 327
Joined: Mon Dec 23, 2013 3:13 am
Location: Hamburg
Contact:

Re: Experimental and Migration

Post by dunnoob »

Your oldest cities will grow to size 50 or more, that requires big distances. You can pump up small cities to size 8 with "help build" settlers, and you can buy one building per turn making it more desirable for the inhabitants. Vague impression, without checking the code or trying to find Linux man-pages: Trade routes are good against migration.

Experimental has radius+3 (distance), 75% national chance, and a 25% chance for migration to from/other nations. I changed that to 66% national, 22% international, and use the default mgr_turninterval, IIRC that's 5, experimental explicitly sets 3. It's not locked in the ruleset. you can change it as you like before you click start.

You can't completely disable migration in experimental. I actually do the opposite, I enable it in the saved settings also for classic. I don't use plague so far, and no hospital/monsanto. The latter is one of the things I hate in experimental, but it's full of brilliant ideas, okay, maybe migration is only good, not brilliant. :D But the variable city radius, the root reqs for techs, and no infinite moves on rails, hopefully classic inherits more of these good experimental features in future versions.
User avatar
Arbogast
Elite
Posts: 412
Joined: Tue Jul 16, 2013 9:09 am
Location: France

Re: Experimental and Migration

Post by Arbogast »

As for your tips they are good as that's exactly what I do...
But my question is about lowering the migration value...
Where's is that ?
Last edited by Arbogast on Thu Jul 28, 2016 1:20 pm, edited 1 time in total.
cazfi
Elite
Posts: 3077
Joined: Tue Jan 29, 2013 6:54 pm

Re: Experimental and Migration

Post by cazfi »

There are server settings about migration, and there's ruleset values (effects) about it. Usually you should be fine by adjusting the server settings (those include completely disabling migration, I think)
User avatar
Arbogast
Elite
Posts: 412
Joined: Tue Jul 16, 2013 9:09 am
Location: France

Re: Experimental and Migration

Post by Arbogast »

I'm getting a little confused OR confusing...
First, migration is blocked in the ruleset so modifying Migration to DISABLED in the XXX.serv doesn't work.
Secondly, the values I'm talking about are in the effects.ruleset, that is the value is set to 10 in all governments except Democracy which is 20.

So, again my question : Would lowering those values will it discourage migration ?

dunnoob : where did you those numbers ? 75% national chance and other such ? As for trade routes they are disabled in national roads/cities as far as I read.
So back to the first line in this post.

Oh, I forgot to mention that the 10 value is a minus (-10) and the 20 is a plus one. THAT's what is confusing me.
cazfi
Elite
Posts: 3077
Joined: Tue Jan 29, 2013 6:54 pm

Re: Experimental and Migration

Post by cazfi »

See server settings mgr_distance, mgr_foodneeded, mgr_nationchance, mgr_turninterval, and mgr_worldchance. Values in effects ruleset are about how much people like certain city. Bigger difference there is between the cities, more likely people are to migrate.

You can think ruleset values and server settings roughly as multipliers for each other (except that those ruleset effects are only part of the overall value of a city). You can use either to control things, depending who you are (ruleset author or user)
User avatar
dunnoob
Elite
Posts: 327
Joined: Mon Dec 23, 2013 3:13 am
Location: Hamburg
Contact:

Re: Experimental and Migration

Post by dunnoob »

Arbogast wrote:75% national chance and other such ?
I looked into data/experimental/game.ruleset, all game.ruleset files have a section [settings] typically at the end with server options. That's where rulesets can optionally tweak and lock settings, we're talking about the mgr_xxxx stuff at the moment.

I think you're right, there's no compelling reason to lock set migration TRUE. The ruleset does not depend on this setting, and as the name experimental says it should be as open for experiments as possible. Ignoring radical approaches with a text editor != notepad.exe ;)
I forgot to mention that the 10 value is a minus (-10) and the 20 is a plus one.
The data/experimental/effects.ruleset says type = "Migration_Pct", so if something is normally 75% it could be 65% or 95% depending on the government. And pray that the code handles less than 0 as 0 (never) and more than 100 as 100 (always). Don't worry, that's a joke. "Unlock" submitted as patch 7518.
cazfi
Elite
Posts: 3077
Joined: Tue Jan 29, 2013 6:54 pm

Re: Experimental and Migration

Post by cazfi »

dunnoob wrote:more than 100 as 100 (always)
I've not checked that particular percentage, but it's likely to be 100% = normal level, 200% = twice the normal level...
User avatar
dunnoob
Elite
Posts: 327
Joined: Mon Dec 23, 2013 3:13 am
Location: Hamburg
Contact:

Re: Experimental and Migration

Post by dunnoob »

cazfi wrote: it's likely to be 100% = normal level, 200% = twice the normal level...
Okay, those effect types should be documented at some point in time, not necessarily before 2020 and version 3.0, they tend to confuse me:

Some numerical types are additive, y means new x := old x + y.
Some are multiplicative, new x := old x * y.
And if it's not only a hallucination on my side there are also absolute values new x := y.
Post Reply