Final Frontier Plus

Final Frontier 3D?

First time ever Civ poster here so apologize if I have the wrong forum for this. I have been playing this game since my little box black and white Apple 2e computer, and most canned programs since. Wasn't aware of things like FF+ so will be excited to explore. Anyways, I have only played one game of Final Frontier (BTS version) and truly love the concept. Hopefully your tweaks will help. I thought perhaps this was an abandoned project. So glad it's alive.

My main thought is could programmers not layer similar grids to create a truly 3D Civ? With the basics of Final Frontier in 3D space, with the great graphics. That's a Civ I would buy!!! Thoughts?
 
Final Frontier 3D?

First time ever Civ poster here so apologize if I have the wrong forum for this. I have been playing this game since my little box black and white Apple 2e computer, and most canned programs since. Wasn't aware of things like FF+ so will be excited to explore. Anyways, I have only played one game of Final Frontier (BTS version) and truly love the concept. Hopefully your tweaks will help. I thought perhaps this was an abandoned project. So glad it's alive.

My main thought is could programmers not layer similar grids to create a truly 3D Civ? With the basics of Final Frontier in 3D space, with the great graphics. That's a Civ I would buy!!! Thoughts?

....Wau.....maybe you should try Sword of Stars or something similar.....SoSE is good too. I myself never much liked any of the "true" 3D RTS or TBS games.
 
an awesome mod... almost a standalone game.
really tnx from a veteran player for making civ IV an evergreen on my pc.
 
Hi,

Been a huge fan of this FF+ and play it just about every month since inception.

The new version adds more squadrons to the AI order of battle. This is good.


The nagging problem for me with FF+ is that the AI simply doesn't build enough units. It's almost like the polar opposite of other Civ4 mods...sheesh the stacks of units that come flooding in (yes, mostly the same unit type but still a serious threat) in Master of Mana, for example.

Unless I set the AI with huge advantages in difficulty I can avoid wars long enough to build up and pick my victim.

See for yourself. Start a game until about the mid-game...then go into worldbuilder and see just how empty the AI order of battle is.

Cities taken over by pirates. Cities with a single destroyer as guard. Entirely undefended cities. Colonizers wandering around by themselves. No "standing navy" except for some squadrons here and there.


Near-term request:

default garrison set to three defense ships for all AI cities.

aggressive AI game setting: make it MORE aggressive. make each AI civ an angry little warlord waiting to crush and destroy. I have to assume that the AI doesn't bother to build a worthwhile navy until somebody doesn't like them. So let's have everyone at war with *someone* real quick-like.

option: don't count starbases as civ military strength. This is one thing that "artificially" inflates human player military strength, IMHO...which helps humans like me to avoid war, grab the desired resources, and build navy to victory while the AI plods aimlessly.

Long term:

A game setting for "AI fleet anchorage": have AI *always* have stack of doom somewhere. Capital ships preferred.

____

rambling:

Is the AI *unit quantity* building behavior exposed much in python for modders? I'm a newbie to programming but am closer and closer to be motivated to finding a "warfare" solution to this, one of my most loved Civ worlds. I've only seen some mention of "unit build preferences can be modified to favor particular unit types in general" sort of thing....
 
I found, and fixed, a couple of things in the AI that prevented it from making units but I still don't really know why it is building so few. My main suspicion is that it is mostly because in FFP, unlike FF, the AIs can actually build things on all of the planets when in FF the only place it ever changed what the build planet is happened in the Python AI overrides (in the function mentioned in the next paragraph). Because of that, regular FF runs out of buildings to build and then builds mostly units (or occasionally research or whatever) but it takes FFP much lonver to run out of buildings whcih can be built in a star system. The most puzzling part is the basic defense. That was bumped up slightly (as I recall it is having it round fractions up instead of down, basically, so in either the the 2nd and 4th eras or the 1st and 3rd eras, I don't remember which, it wants 1 more than it used to) but it still does not seem to want to build the specified minimum even though the code in the DLL should be making it do so unless one of a very few high priority things overrides it in there. And yet it doesn't seem to actually do so all of the time. I did not walk through that code using the debugger during a game enough to see what the problem was.

It is possible to make it the AIs build units via Python. In the CvAI.py file the doCityAIProduction function is overriding the AI in some specific ways when it is starting the build decision making for a city. Almost all of what is done in that function involves making it build buildings to solve specific problems (like building a sports arena if it is at or very near the happiness limit, for example). Originally the only units it built were construction ships. Now it can, in recent versions, queue up a few military unit types: carriers and squadrons. It won't have a city build any other unit types.

There is an even easier thing to try adjusting. In the XML it might help to increase one of the settings in the CIV4LeaderHeadInfos.xml for all the leaders. Each leader has a value called iBuildUnitProb. Currently the barbarians are at 50 and the real civs are in the 20 to 40 range. Bumping these up some more might help a little (they were all increased by 5 a long time ago). Another 10 or 20, or maybe even 30, added to each of them might help. I do not expect this to solve the problem, but it would probably help some. It may also help to adjust some of the parameters relating to planning wars, to make it more likely. These are the iMaxWarRand, iLimitedWarRand, and iDogpileWarRand - for these a lower value makes them more likely to want to go to war so subtracting 5 or 10 from each (or maybe a 10% reduction to the current value for each) could be interesting. I seem to recall there being a dividing line at 100 where going below that value triggers some other stuff too. Unfortunately adjusting this could have them endlessly planning for war without ever actually declaring war, or even worse just going to war a lot without being more prepared, instead of actually making them more aggressive in a way that makes them more dangerous.
 
Much obliged with the response! As Darth Vader said in Empire Strikes Back: "Impressive!"

I'm going to have to parse each of the recommendations and make a list and give each a try.

That statement: "the doCityAIProduction function is overriding the AI in some specific ways when it is starting the build decision making for a city. Almost all of what is done in that function involves making it build buildings to solve specific problems (like building a sports arena if it is at or very near the happiness limit, for example)

...is very interesting. Speaking as a newbie-programmer: The medium/large sized cities *without access to health/happiness resources* are subject to a large amount of unhappiness/unhealthiness. Perhaps the speed at which AI cities grow would mean that the city would reach the thresholds and perhaps feel the need to queue buildings to address the problem. Cities continue to grow...and continue to try to build the solution.

Another factor: AI is subject to map luck, and only fair at, finding and exploiting resources as well.

The truth is: a city can be *plenty* effective without a single building to offset these problems.

...of course my inclination is to disable research and culture victory conditions so I'm shamelessly biased towards making this a more conquest game (if I choose).


Further rambling: if we were to halve the amount of unhappiness and unhealthiness for testing...wonder if the AI would then settle down to unit building....
 
A problem with the AI, in the DLL parts, which I did not mention is that it does not know how much population is actually useful. The planets are done in Python and it does not know how much population can actually be working. Because of this it likes to increase the population above the level that can actually produce additional yield.

The Python AI override should probably be turning the "avoid growth" setting on & off for cities but it doesn't. The problem is that this is actually complicated - it is done in the main interface via a specific "emphasis" widget type and there is no function to directory turn it on or off. So the easy way to do this is probably to have the DLL expose some sort of "setAvoidGrowth" function to Python. So I therefore never messed with it.

Because of this, I'm sure the AI is trying to make people happy and healthy who can not possibly produce any yields, due to the planets being full already. The Python AI override takes this into account, only caring about happiness up to the population limit of the star system and likewise for health, but the DLL does not have any idea what is going on with this.

So, in short, the AI really could use some more work.
 
Thanks. For fun, I've also tried to play with the Military-centric aiWeights in CvAI.py to see if I could make "military" in general have greater weight than buildings in the AI city queue(s), or make the AI more likely to build a barracks etc. but without any results that I could see. Will work through your specific recommendations and observe at the 50 turn/100 turn checkpoints in each case. Might take a week or so.

Pardon, my first look at Python. Seems pretty clear in CvAI.py where:

# Number of military units per city
aiWeights[iMilitaryWeightType] += 100 * pPlayer.getNumCities() / pPlayer.getNumMilitaryUnits()

...would, as the number of AI player units is low...cause the resulting value added to military aiWeight to be *higher*..etc. But I think this is just what you were saying..."something else likely in .dll or in other AI behavior is overriding tweaks like the above.

...if it doesn't already...there should be a code snippet to display the starting/ending values of each type of aiWeight as the code progresses in each section...or maybe I'm waaaay getting ahead of myself.

Starting to peek at the components of the Civ4 AI etc.

Like in:
[TUTORIAL] An Idiots Guide to Editing the DLL
http://forums.civfanatics.com/showthread.php?t=314201

...I'm just starting to learn c++ so might as well start the trek towards creating a view on what the Civ4 AI code is doing in c++, and in python. Would be wicked to see the code being stepped through in a debug dll and see where the AI is deciding to do what.
 
Can somebody from here, especially GE, join our PBEM game please? We need a replacement in a very cool game!:help::help::help:
 
Is this still being worked on or has this died?
 
Is this still being worked on or has this died?

It's been a while, but you (and this thread) deserve an official response, so see this thread (and my post in it) from the forum.

It hasn't "died", exactly, but it is fair to say it's in suspended animation. I'm busy with university and many other things these days; every so often I do something, but rarely have the time or feel like putting in a lot of work.

So the answer is that there are unlikely to be more FF+ releases in the near future, unless additional people are interested in helping out. Perhaps one day I'll feel like working on it and do a bit of work and release some small updates, otherwise.
 
Why does this mod fail to introduce a spy unit? All the infrastructure is already in the game!
 
If a spy is introduced… please, please, please, make it something that can be turned off in the options without affecting anything else! The only reason that I don't turn off espionage in BTS is that it messes up culture to do so! I detest the espionage in BTS.
 
In order to make use of the espionage related missions it would be necessary to go through and check which of them still make sense. For example, poisoning a well might make a city unhealthy, but it would most likely not be effective against population spread across multiple planets in a star system... Some may not make sense and some may not have a reasonable effect.

Nobody has taken the time to do so (particularly TC01 and myself). I seem to recall that there was a little talk about maybe assigning some missions to the stealth ships, or making a spy ship. Nothing ever came of it.
 
If a spy is introduced… please, please, please, make it something that can be turned off in the options without affecting anything else! The only reason that I don't turn off espionage in BTS is that it messes up culture to do so! I detest the espionage in BTS.

But why? It is one of the most exciting and realistic features (except for small costs in changing civic/religion missions).

Spy ships would be nice! Spy could chose a planet before conducting all the vanilla missions as usual.
 
I've been playing around with your mod for a few days now. Big thanks to you, God-Emperor and TC01. I won't say that I think Final Frontier is very good even with your mod - it has too many appalling shortcomings for that - but you've definitely improved and expanded greatly upon Shafer's original base. And FF remains a neat, if somewhat dysfunctional, mod.

I find that the main problem with the mod is the AI is unable to grow its colonies. This means the player will invariably end up in a dominant position with a population many times the size of what the AIs can muster. I believe the main cause of this has to do with what planets are selected for habitation within a system - I've seen the governor of my own world pick the blue 0-2-6 planets over the 2-0-4 ice planets and so on. In general, it appears the governor prioritizes hammers too much. That might work for a short while for warmongering but it won't help for building a long-term empire. Sadly, while the player can force working food-planets I doubt the AI has the same option.

Also, I must single out and nitpick on the Slave State civic as this one is particularly glaring. You've made changes to the civics but seem to have left the SS civic unchanged, which is surprising. Food doesn't work the same as it does in vanilla IV so it doesn't make any sense that SS provides only what appears to be something like 30 hammers/pop. Also, the uselessness of the civic is amplified by how it comes so late in the tech tree as opposed to early in vanilla (the idea in vanilla being that it's a way to convert unhappy citizens into something useful before you have happiness enough to support them) - and further amplified by how the cost of population point scales in FF, with later points being extremely expensive. Thus, in a size ~15 colony SS civic allows you to convert something like ~150 food into 30 hammers. So, this civic is terrible and the AI unfortunately likes using it and demolishing its already low pop growth even further, which means the late mid-game essentially has the AIs self-destruct.

Those were a lot of nays but once again: Thanks for creating this. Should you at a later point wish to take up development of this again, though, I would recommend looking at some of these eye-popping issues with the mod because sadly, it does detract from the value of the rest of your mod (which you clearly invested a lot into given the wealth of good ideas and their professional implementation in most areas of the mod).
 
I find that the main problem with the mod is the AI is unable to grow its colonies. This means the player will invariably end up in a dominant position with a population many times the size of what the AIs can muster. I believe the main cause of this has to do with what planets are selected for habitation within a system - I've seen the governor of my own world pick the blue 0-2-6 planets over the 2-0-4 ice planets and so on. In general, it appears the governor prioritizes hammers too much. That might work for a short while for warmongering but it won't help for building a long-term empire. Sadly, while the player can force working food-planets I doubt the AI has the same option.
I am forced to agree with Strategist83's assessment of the FF game. I've come across the same problem in the B5 mod and until recently have not been able to do very much about it. I've added 1 food to the benefits of having a trade route between 2 cities/solar systems. This has helped but I've also made having access to one of the food resources provide a benefit in food growth to all cities/star systems connected to it. The only problem with doing this is the AI is still not growing enough as, as stated above, the AI prefers production and/or credits over food planets.
Is there anything that can be done to get the AI to use the food planets as a higher priority?
 
Top Bottom