[Freeciv] Re: How to solve ICS - increasing settlercost

[Reinier Post <rp@xxxxxxxxxx>]

On Wed, Feb 26, 2003 at 02:32:51AM -0400, Mark Metson wrote:

> I have always thought it strange that lightbulbs accumulated researching 
> one tech can suddenly be deemed to have been working toward a different 
> tect entirely. Thus I already had reason to think it might be good to 
> throw away old lightbulbs any time that you change your research goal.

This is controllled by a server variable.
Server variables are the Freeciv way to deal with obscure cutoff limits
in the code: they at least make them visible and modifiable.

Many of these are brought about by Civ II compatibility.
 
> If one is prepared to do that, then it ought not be too hard to make 
> research of certain techs require certain buildings. For example, any tech 
> that has University among its ancestors in the tech-tree could require 
> lightbulbs from cities that have a University. If you choose to research 
> such a tech, suddenly your lightbulb production is effectively limited to 
> only those cities that have Universities.

This looks feasible.  But Universities are special in that they boost
research.  Would Barracks be required to research new military units?
If you make it that general, do you require *all* buildings enabled
by *any* technology on the path?  That seems wrong.  And how would you
make this mechanism clear to the user?  By adding information in
the science dialog?  Is that good enough?

> This would open to question "what happened to all the other potential 
> research". So a more general solution would be to allow research on a 
> per-city basis instead of on a per-civilisation basis.

I think that would be too much micromanagement.

> [...]  So maybe give 
> each city that lacks the crucial building one "luxuries" point for every 
> two (or three, or whatever) "wasted lightbulbs".

Argh.  Yet another computation to make the exact effect of settings even
more obscure.  With new hardcoded cutoff values, too.  Brrr.

> The idea of allowing unit production to be able to depend upon buildings 
> is also good.

Yes.  But this is already implemented (the Factory effect).
Essentially you can just create more buildings with a factory effect.

> Would starting without a Palace have much effect upon smallpox? Maybe 
> making them at least want to build a capital city?

No palace -> no research at all.  Hence your only hope is to conquer
someone else's palace.

> Another idea would be to make research of tech as an attribute of a 
> civilisation - a transmittable thing communicable throughout a population 
> - require a "school" building, so that no city that lacks a "school" can 
> produce any lightbulbs at all.

This is simple - I quite like this,  The rule can be that as soon as
you can build a research enhancing building, like a Library, you *must*
build them in order to be able to do any further research.  However that
would create the cutoff problem again: you have to be very aware of this
effect before you research a Library-enabling tech and you have to
be careful not to discover them by accident!  A more continuous rule
is better.  For example, the rule can be that tech development speed
decreases rather sharply and that the effect of Library-like buildings
is much increased.  This rule has the additional advantage that Freeciv
already supports it.

>  Schools would require some tech such as 
> "culture" or "language" or both that Civ and FreeCiv seem to already 
> assume all civilisations, and in particular all units capable of founding 
> cities, possess.

Well, you could use the Library instead.  Without Settlers at the start
you need a way to still have cities or some other source of development.

-- 
Reinier