PROJECT: Playtest

[Martin Gühmann]

Re: Mid-Game Report

Originally posted by Tamerlin
Another nuisance is the number of caravans needed to open a trade road, the number needed is so high (around 25 to 34) that building caravans is pointless and the AI Civs have not open any trade road neither.

And why don't you build roads? However you can play around with the CARAVAN_COEF in const.txt, you should try what happens if you half it.

-Martin

[drulius]

Found a new bug ( I think )
You can place tile imps on unexplored squares near your cities .

I found you can place Fortresses on any square outside your borders (and not inside someone else's borders) that one of your units can see. I don't recall ever being able to place Fortresses outside my borders before.... But then again, I've never really had much use for Fortresses.

[Tamerlin]

Re: Re: Mid-Game Report

Originally posted by Martin Gühmann
And why don't you build roads? However you can play around with the CARAVAN_COEF in const.txt, you should try what happens if you half it.

In my game Suez and Memphis are 7 tiles away from each others an are indirectly linked by a 11 tiles long road passing through Thèbes. I am asked 30 caravans to create a trade road that would grant me 61 Gold per turn. Certainly worth the pain in the long run but there are more important things to build during the time needed to build such a high number of caravans.

I can mod the CARAVAN_COEFF (thanks for the tip ) but this is nonetheless something we should think about at this is not good IMO.

[The Big Mc]

To be honest I never really got this you got to build caravan thing. I think you should pay to create a “trade road” now kill you production for five or six turns for little gold.

[Martin Gühmann]

Originally posted by drulius
I found you can place Fortresses on any square outside your borders (and not inside someone else's borders) that one of your units can see. I don't recall ever being able to place Fortresses outside my borders before.... But then again, I've never really had much use for Fortresses.

That's the sense behind Fortresses to extend your empire without the need to build cities.

Originally posted by Tamerlin
I can mod the CARAVAN_COEFF (thanks for the tip ) but this is nonetheless something we should think about at this is not good IMO.

Well you could figure out what the best value for the CARAVAN_COEFF, or I can hardencode the change. However we should lower the costs to build a Caravan.

Originally posted by Tamerlin
It looks like a hopeless situation then. If you do a /reloadslic, you will reset all variables - so the game state is destroyed, and it just isn't a continuation of the saved game any more. But if you don't do a /reloadslic, stored references (indices in arrays, probably) become incorrect, and anything may happen.

Most slic code I know can handle this, FrenzyAI starts from scratch to rebuild its state, maybe Diplomod with its Embassy code might be a problem, but the code itsself is a problem as well, it makes the game scrash when a civ is added to the game, and even if wouldn't make the game crash its embassy matrix would be totally screwed up. My GoodMod considers a /reloadslic and also Peter's CTC considers a /reloadslic. And if I look on the ConsiderProposal function in Diplomod then I must see that this has to be rewritten with Peter's changes.

For scenarios, if it is done at the beginning of the game then it should be ok, actual the slic should be reloaded with the start of a scenario anyway. It shouldn't be taken from the scenario file that may have some buggy slic that doesn't match with the slic you find in the scenario folder. However I am just talking about reloading slic onces or twice for a whole game and not everytime when the game is reloaded.

-Martin

[Tamerlin]

Originally posted by Martin Gühmann

Well you could figure out what the best value for the CARAVAN_COEFF, or I can hardencode the change. However we should lower the costs to build a Caravan.

As far as I remember the values in the SAP2 seemed right, I will check and compare them with the vanilla game (if they differ at all). Someone suggested the changes could come from the way the trade paths are now computed. If this is the case it will not be easy to adjust the values so that they produce the same result.

Most slic code I know can handle this, FrenzyAI starts from scratch to rebuild its state, maybe Diplomod with its Embassy code...

BTW, it would be great if the Embassies were restored once you are no more at war with another civ... as in the SAP2.

... and the latest quote in your post is not from me, I can't write as well in English.

Another thing, the entries of the Great Library should be reworked, for example Galileo's Telescope's entry does not state clearly that it gives you 5 "free" Scientists in the city it is built in. And this is the case for almost all the advances and Wonders.

[J Bytheway]

Originally posted by Tamerlin
As far as I remember the values in the SAP2 seemed right, I will check and compare them with the vanilla game (if they differ at all). Someone suggested the changes could come from the way the trade paths are now computed. If this is the case it will not be easy to adjust the values so that they produce the same result.

Well, assuming the poposed explanation was correct then the new system will give the same values as the old system if there are maglevs present. So, if you want to adjust it to give the same behaviour as the old system if there are roads present, rather than maglevs, you need to multiply CARAVAN_COEFF by 5/33.

[Tamerlin]

Originally posted by J Bytheway
Well, assuming the poposed explanation was correct then the new system will give the same values as the old system if there are maglevs present. So, if you want to adjust it to give the same behaviour as the old system if there are roads present, rather than maglevs, you need to multiply CARAVAN_COEFF by 5/33.

Thanks for the COEFF J Bytheway, it will spare me many tests.

I think I will wait for the next playtest build before playing with the CARAVAN_COEFF, if the new system sounds good I won't modify the coefficient. I suppose that one of the goals is to keep the human player from earning too much gold through trade.

In my current game the only trade road I have seen was between two cities that were linked by a 4 tiles long road.

[The Big Mc]

I have to crush the Greeks(or at least make it so they only have 1 city) then I may start trading.

[Fromafar]

Originally posted by Martin Gühmann
Most slic code I know can handle this
[...]
For scenarios, if it is done at the beginning of the game then it should be ok, actual the slic should be reloaded with the start of a scenario anyway. It shouldn't be taken from the scenario file that may have some buggy slic that doesn't match with the slic you find in the scenario folder. However I am just talking about reloading slic onces or twice for a whole game and not everytime when the game is reloaded.

I agree with the beginning of the game. In that case, you will lose nothing.
From any other point, it will depend on the specific Slic code. If you have no real state variables (e.g. only random, recomputable, or periodic events), doing /reloadslic is relatively harmless. But when you have some variable storing cumulative battle wounds, or a flag storing some essential achievement, resetting it will cause the game to change completely.

Originally posted by Tamerlin
I think I will wait for the next playtest build before playing with the CARAVAN_COEFF
[...]
In my current game the only trade road I have seen was between two cities that were linked by a 4 tiles long road.

Don't bother building roads. In the current playtest build, you should connect cities with farms and mines to make the caravan cost go down .
However, don't count on this strategy to still work in the next release.

[Tamerlin]

Originally posted by Fromafar
Don't bother building roads. In the current playtest build, you should connect cities with farms and mines to make the caravan cost go down .

You mean that to the prohibitive cost of the 25 or so caravans needed I would have to add the cost of a myriad of farms linking the two cities (farms that would be better used elsewhere, by the way)...

However, don't count on this strategy to still work in the next release.

I feel reassured.

[Flinx]

Originally posted by Tamerlin
Another nuisance is the number of caravans needed to open a trade road, the number needed is so high (around 25 to 34) that building caravans is pointless and the AI Civs have not open any trade road neither.

In my games it appears the AI has kept up with me in terms of the number of trade routes, so the AI has not been put at a disadvantage by the change as far as I can tell. I did not like this ‘new’ trade system first either but I have changed my mind. There is too much gold in the game; now that trade routes are more difficult to get in the early game, this will reduce the amount of extra gold. As you build roads, then rail and finally maglevs the number of caravans per trade route decreases, increasing the amount of gold from trade in the late game. I want to play a few more games before I consider minor changes to CARAVAN_COEF or the build cost of caravans.

Originally posted by The Big Mc
To be honest I never really got this you got to build caravan thing. I think you should pay to create a “trade road” now kill you production for five or six turns for little gold.

I disagree....

[Ekmek]

Is it possible to add a gold cost to building caravans and tile improvements?

[J Bytheway]

Anything is possible*. That would probably be relatively easy to do for tile imps, but tougher for caravans (as a rough initial guess).

*exceptions apply

[The Big Mc]

Just had to abandon my play test. I started with 3 AI I killed two before they got too many cities but the Greeks keep building more cities ( A lot of new cities) I ended up with 61 cities and the AI had at least 11 left So baring a perpetual period of anarchy I can’t conquer the world .

the AI on easy ( easy making it better to spot bugs as you not concentrating on the AI ).