Re: [Freeciv-Dev] PATCH: AI governments + PATCH hell level

[Peter Schaefer <schaefer@xxxxxx>]

Several people wrote:
> >
> > When you write *tmpcity=*pcity, what does It do with
> >  int ppl_happy[5], ppl_content[5], ppl_unhappy[5];
> >
> > Does it really copy the array or only int * ?
> 
> No, the whole thing is copied. I think this is fairly
> standard behaviour...
> 
Yes, this is an interesting ambiguity in C:
although pplhappy is an array (as in int ppl_happy[5]),
it is also a pointer to an int. But you can be sure that 
for multidimensional arrays lots_ppl_happy[5][5] the value
lots_ppl_happy[1][0] is stored after lots_ppl_happy[0][4].
So being an array comes first.

[Back on topic]

> So the question is (***opinions_are_sought!***):
> 
> - Do we want the AI to be able to change governments at all levels?
> - Should we create a new level (as Nicolas suggested) ?
> - Or should we allow the AI to switch between different governments
>   at the hard level (or normal too)?
> 
> Personally, i'd vote for the 3rd option (ie allow the AI to change
> governments in normal/hard - or at least hard level). I think that
> even though the AI becomes more powerful this way, it will still
> be lagging behing good players in development and therefore all
> this will be of little (if any use).
> 
> Once the above is agreed, i can prepare the final patch.
> 
> --rizos

If I may offer a definition,

Normal mode should have the AI play at best as it can without cheating;
if you wish you may consider excessive micromanagement, which the human player
can't do well, as cheating ( like buying settlers the right time, knowing 
in advance that a city will grow and revolt ).

Hard mode should give the AI advantages, especially in those fields
were it can't compete with a human player, like gathering intelligence
(i.e. were has this battleship gone that has moved into the Black).

The advantage the AI has in allowing in 100% rates IMHO should be moved to
a level above hard; (maybe hell or unfair). ( You could defend the 100% rates 
by saying that later in the game the AI often alternates between 100% and 0%
rates. But this is not true in the beginning. ).

So if, as I understand, Rizo's AI work better than the old one,
AI should use it starting from normal level.
If people wish to test the old AI, there could be a subnormal(or test)
level.

As testing different AIs can be fun, maybe there should be a way to do this;
Maybe a custom AI that reads some flags from the ruler name
( like Alex_0x32a1 custom ).

My favorite Ai handicaps would be(see stdinhand.c):

/*worst     */ -1, 
                0, 
                0, 
/*easy      */  H_RATES|H_TARGETS|H_HUTS|H_OLD_EVAL_BUILDINGS,             
/*subnormal */  H_RATES|H_TARGETS|H_HUTS|H_OLD_EVAL_BUILDINGS,
/*normal    */  H_RATES|H_TARGETS|H_HUTS, 
                0,
/* hard */      H_RATES,
                0,
                0,
/* deity*/      0,


My main point is  Rizo's AI should go into the normal level.

-- 
Peter Schaefer - schaefer@xxxxxx, schaefer@xxxxxx
"Buy AMD processors or you will have to pay $4000 for the cheapest CPU in 2020"