The Vault Dev. Canteen (General Discussion)

[Lib.Spi't]

have you tried out my quick speed file yet lark?

You seem to be getting through your games quicker than me (I am still working my way through my first).

I am really eager to hear your thoughts on it as a fast player, cause like I said I have no real context to tell if it is any good or not, most things seem super speedy to me, but then I am used to playing for 3x or more longer than a quick game.

 

[Lark]

have you tried out my quick speed file yet lark?

Distracted, I've been. But I just D/Led it and I'll try it tomorrow.

I'll play with your file and I'll think about it ...and then I'll think about the AI:

Spoiler :

I played a <Ctrl+Shift+X> AI-only game with my branch. Unfortunately, turning "Raging Barbarians" on makes the AIs play even worse than they do already. I went 500 turns and nobody looked close to winning.

Man, the AI is stupid. It plays even stupider than I do.

It makes roads all over. And Gardens all over, replaced by Cottages all over. A waste of money (OK, commerce, really, since it's fungible). Also, I'm fairly sure Gardens are slightly better than Cottages - and Farms than Towns - even before the consideration that building units uses up population and more food replaces it faster. If the highlit Worker actions are an indication, it also builds expensive advanced improvements instead of building the cheaper auto-upgradable base improvement.

It keeps turning research down, so it must be using a lot of money for something else, too. Upgrades? Rushing? -Both of those are usually inefficient.

It fails to build buildings, especially in later cities. I watched one city working on four different units over a hundred turns and not finishing any (unless it rushed them). It didn't build a single building even though it was way behind. One should generally only make units in cities that are close to caught up on buildings.
In a different game, I watched the Super Mutants build a Stupid Mutant Fighter in their capital for perhaps 80 turns without ever inserting newly-researched buildings in front of it. By the time they would have finished, the competition would have been able to build units twice as good with twice the production.

Trees were chopped everywhere. A human, with fewer more-specialized workers, can use the Treehugger promos for an edge.

And it only expands at the edge of its own territories (usually OK, but not always best). I was considering cheaper Settlers, but at present that would only hand a bigger advantage to the human.

Game 2: no raging, no villages, no techtrade, no events:
From turn 300 to 700, research was 0%! Why? -AAAARRRGHHHH!, that's why; 14 unhappy faces, ~2500 coins less, god knows what the reduction in research was -- because the AI set every single civic to the worst possible.
Guess how fast a newly built town grows when a) it can only use the center tile and b) it spends all its time making high-end units.

Most of the resources have Watchtowers on them (which I believe at least connects them), some have Cottages, only a few have the appropriate producing improvement. So, a large reduction in production.
Most ruins have Cottages instead of Scavenger improvements. And I don't see any Mines. So, no chance for more resources to pump up corporate bonuses.
I changed the Ghoul Farm prereqs to Rad Waste + Wasteland from r.w. + Fertile Wasteland (a combination that doesn't exist on the map) and it actually built some. Only: a) it's started replacing them with Cottages, which are worse, b) it's put some on non-Rad Waste tiles (I suspect it's an OR prereq rather than the AND intended), and c) the health bonus doesn't work yet (and I don't know how to figure out why). I will just delete the Wasteland part (result: only r.w. = only r.w + w.) and look at the dll folder to see if I can figure out what should be reading health.

It keeps building resource buildings. There is little advantage to having 8 sources of Junk and the opportunity cost is considerable.

Also, one city that had pushed its territory to two tiles from the Mutant capital and one tile from its second city ...flipped to the Mutants. I don't understand how that's even possible. Can Spies do that?

 

[Lib.Spi't]

Those are some very interesting observations that you have made about the AI, this is really helpful stuff, as I am too tired to notice alot of these patterns. At htis popint a lot of the stuff in the game is still in the let's try it stage, many things (like ghoul farms) I threw in quick as an idea, and probably copy pasted over stuff that I did not want. (Like the fertile wasteland bit for G. Farms.)

Lets break it down some.

Roads:
This is one of those AI features that bugs me, more from how I want the gameplay to go, more than whether or not it is a good move.
I tried to make roads in such a way that the AI would no longer spam them, increased costs and I even added -1 food at one point, it may be that the upgraded versions don't have that negative so it goes back to it's old spammy behaviour.

I really want roads to be really expensive, and really powerful but fragile. The idea that connecting the wasteland together is a massive logistic undertaking, keeping them free of bandits and such would be really hard, I want them to be used in an absolute minimum way (because they are so expensive) but the benefits of having them to be massive (the connection of bonus resources to every city is very valuable, as it leads to those building benefits from resources).

In many ways I want safe road networks to be like the pinnacle of a wasteland society, connecting your empire together is hard work but massively rewarding. I need to get the pain and pleasure mix for routes just right, and is also linked to those other factors like the resources boosts, etc. Also getting the AI to value them intelligently and use them sparingly is going to be the challenge too.

It may be that I need to increase the negatives 'on tile' for roads, so that they only go where needed, rather than everywhere..

Buildings/Units:
I have noticed several odd 'valuations' by the ai when it comes to unit building, for example I saw it build stuff like a fighter when it had access to plasma armed power armour, and the turn time was negligible, so why not build power armour?!?

The Super Mutants are an exception right now, as they are an unfinished concept when it comes to their mutant units. They need to be 'super' but this means they need to be limited in some way. I have yet to come up with the best limitation, so for now I am not that bothered by their failure to work well.

The more pressing concern for now is that non of the AI factions are much good compared to the player.

The fact that units seem be of a higher value than buildings is a problem, as you say, it may be a simple matter of adjusting the ivalue and ipower and aiweight type tags (I think that is what they are called...) so that buildings are viewed as greater value than units, so that they get built first and the city then becomes more powerful allowing it to build units in good order.

We would need to do some side by side reviewing of those tags for buildings and units and see if the units are disproprotionately higher than buildings, if so it may just be a simple adjustment of the two relevant values. (That would be nice right? )

Trees:
I think tree chopping was always a default AI move, with our new additions this is a pretty stupid move. For now we would probably have to 'turn off' tree chopping, or make it massively rewarding, so the AI at least gets something out of it.. Chances are it will use it early though and waste any benefit from it.

Perhaps the best solution is removing tree chopping for now, as trees have an ever increasing usefulness (with the tree hugger promos and logging improvements) thn chopping would only exist for rad weeds (jungle) which is just designed to be a nuissance anyway (or at least that is the future plan).

AI Expansion:
I am fairly convinced this is currently a nail to the head for the AI in terms of being competitive.
I don't think it has anything to do with settlers cost and such, as they aren't that expensive, and I know the AI builds them.
I think the current AI has something like a hard coded rule that says escort a settler with 2 units (or something similar to this) which is fine for a civ game as barbs are generally weaker or equal to the civs troops and they aren't that common, so the AI gets the job done and founds plenty of settlements.
I think also an AI settler stack has a thinking of 'take no risks' much like a worker, if it sees a bad guy, the stack is programmed to run away.

In FTTW it is a very different story, barbs are everywhere and they can be much stronger than civs.
So if you have a settler stack, with two defenders, it sees risks everywhere at it's borders and runs away, it repeats this cycle for let's say 20 turns, concludes it can't do anything and disbands the settler.
We need to make the ai take a very different approach to settling compared to civ.

Something like:
Take half th units from the city where the settler is built (most AI citys have 5-20 units stacked up) and add them to the settler stack.
Ignore dangerous units up to total number size of stack -3 (settler plus two units) (or move X squares outside of border, X being stack-3. )
Move to settlement site and found.

The idea being is that the AI/Player can afford to lose many of it's units, if the end result is a new settlement, as borders will protect it from animal attacks and a cities development will soon allow for even more units to replace the few that were lost, and the new city equals a big net gain in time.
It would probably also be good for the AI to add a worker to it's settler stack, so it canstart developing the area around the new city as soon as it has the funds.

I think the only reason it expands right now is because someimes a good site opens up inside it's borders or next to the border, we need to get that sphere expanded to about two or 3 squares, and in terms of the Vault civs, as far as it takes to get the next Vault.

Another approach may be creating a culture improvement that has borders around it (like a new city does) and then make the ai workers build 'culture paths' to potential new city sites, so that the settler is protected from animal threats.
It could be something cheap and quick to make, like a 'Patrol Lodge' or something..

Both solutions would need some kind of AI work though I think.

At the moment I had to switch off workers ability to leave borders, as the Superfort AI was seeing them suiciding trying to build forts outside borders. With some tweaks to that code, we may find a result we need.

Improvement Balance:
You are right, watchtowers should connect resources (just like forts). It should be working, I think I used it before...

Your comment about the choice between cheap improvement+Upgrade or Expensive improvement is an interesting one...
I wonder if we should consider removing the build option for later improvements and making them only improvable, then increase the time it takes to improve, so that civics with an improvement upgrade speed increase become more valuable at times when infrastructure needs improvement..
Not sure if that would make it better or worse for the AI...
They won't upgrade if the city is trash anyway, but if cities are not using their tiles because they are trash then the improvements themselves would be meaningless regardless...

The Ghoul Farm was really a half finished thought that I never went back to as I got distracted by more general things. (Like the SM's, Ghouls are a faction where I skipped ahead of my plans and added faction specific content before finsihing general design work)
So chances are it has loads of stupid mistakes like leaving in fertile land as a prereq or validator. It should be Rad waste and Feature Fallout that are two OR validators.
The health think may be from our switch to health as yields, I may need to move the health boost to the yield section, as yield heatlh was added after the Gfarm. (you can try this yourself by adding in the 5th iyield tag (4th is happiness) if you feel more productive than me!).

Village Vs Farm:
Eventaully I sort of want Maxed villages (calling them something like Suburbs/Districts as an upgrade path above towns.) to be the best improvement for standard flat ground. (ruins should probably benefit more from scav camp path, and hills from Mines (not least for the potential bonus resource pops) That will just be continuing to crunch the numbers and settle on the right ones. All these things are still very much WIP.

These ultimate villages should take a long time to get to, and be locked out by high end techs, but like a road network they are representative of a pinnacle of a recovered society.

These would also carry bombardable defense bonuses, so we can end up with some cool urban warfare and urban warfare specialists. (hopefully)

Resource Buildings:
These will continue to become more valuable as we move forward, right now Junk is only really good for getting access to melee troops, but we will continue to find places and buildings where junk (particularly multiple junk) will be more valuable and benficial.

The current Building bonus boost settings (that's some fun alliterration) are ported over from when each building could only use one resource, I now need to add in all the extra and alternate resources for each one. Like, for example, anything that benefits from building materials, could also benefit from junk.

I also eventually want to increase the resource/production system, so that 'base' resources are required to build quantities of production/manufactured resources which would then be consumed in the production of units and maybe buildings and such depending on how far we can take the system. (if we can take it at all)

So the overall fact that the AI prioritises resource buildings will be a good thing in the long run, as these will provide many more bonuses beyond the value of the building itself.

Platy's Manufactured Resources

That is a bare bones python version of the sort of thing I mean, but it would be an sdk version (hopefully) and much more advanced and deep. If I can convince people to help me do it )

The Culture Flip Conundrum:
This may well have been achieved by a spy. I think spies can cause rebellions, which may in erm have made it susceptible to a chain reaction leading to a culture flip. I don't know for sure though.
The rebellion would cause a loss of culture tiles, which would push an enemy border up to it, and then... 'flip!'

 

[Lark]

I did a cost pass on unit strength. My algorithm halved costs on average, but my second city on turn 220 still took 31 turns to build the best available unit (T51b PA). I was thinking I'd just do a 2nd fast pass and flat halve the cost again, but at turn 250 it's already down to 9 turns. So I won't.
On the Rads section below, I think I am able to do 1 & 2, so say so if you'd like to see one or the other and I'll add it into branch v03 and post it for trial.

Some ideas/suggestions/remarks:
Rads:

Spoiler :

A.) Should radiated tiles (Rad Waste & Fallout) be mostly impassable?
B.) Should Ghoul Farms be the only improvement buildable on radiation?

1: A & ~B: Make them impassable like Ice and have a new Worker at Hydro Power (to go with the Scrub Fallout ability) that has bCanMoveImpassable. Ghoul Workers would start with it. This lets these Workers build anything on radiation, though they shouldn't if they can Scrub it. (Ghouls, too, but Ghouls are supposed to enjoy radiation.) bCanMoveImpassable should also be given to Rad Beasts. Unless there's some way to split it up, it would let both Animals and some Workers move onto Peaks - not necessarily a bad thing. Also: more choke points and barriers on the map.

2: Only B: Another way is to only block ordinary Workers from radiation until Hydro Power using unit <FeatureImpassables> plus <FeaturePassableTechs>. This would continue to let all other units enter radiated tiles.

You could add an event at some tech that gives each researching civ a free [RadSuit/Armored/?? Worker] if Ghouls are too OP with this.

Also, how about a "Mister Raddy"? The only unit that can Scrub Fallout and Cover Rad Waste (slowly and very slowly, though twice as fast when you get the tech that gives Fallout's rad-eating plants), Mister Raddy will be glowing in the
dark before his work is done, so his last act is to bury himself.

Roads:

Spoiler :

I really want roads to be really expensive, and really powerful but fragile.
keeping them free of bandits and such would be really hard
I want safe road networks to be like the pinnacle of a wasteland society

I think if you make roads very expensive with yield negatives, what will happen is the AI will spend ever more time with research at 0%, saving up the money to
degrade its own territory. You don't fix an AI by giving it more ability and more incentive to screw up worse.
IRL, raiders and monsters would be idiots to try to loot or eat roads. IRL, warring nations only target bridges. Roads are the least fragile part of a civilization. If you are emulating a low-population scenario such as Fallout, where every unit, every building, every improvement employs some of your precious population ...roads won't. Roads are free in vanilla because they pay for themselves. In Fallout as IRL, most of the USA is flat enough you don't even need roads.
I would prefer Paths to be free or nearly so. Path-spamming isn't as ugly as it might be (whoever did the art is good - they're visible but only just). I hope the remaining route types will only get used to connect cities or improve tiles; at least they won't have tile connection as a reason for building them unnecessarily. Plus, free xp for treehuggery.
One pinnacle of wasteland society would be safe everything, not roads particularly. I'm pretty sure you think that too.
Also, (late game) roadbots! One only makes RRs, one makes Hwys, one more eachto patrol on (only on) and protect them. Laser trains! Flamethrowing robike swarms! Or whatever Clanky wants.

Raids:

Spoiler :

0.) Make barbs scale up appropriately with civ strength. Also: less wandering, more attacking.
1.) Place a few beasts with animal AI in neutral territory. Nothing else.
2.) Have most beasts do everything other barbs do except attack cities. Rad Beasts in Fallout have no genetic fear of Man.
3.) Partial exception to the above: Deathclaws. Place them early but have them use the animal AI except when they find themselves inside a civ's territory.
4.) Have the pillage chance for ruins and vaults be nil.
5.) Have low safety stir up barbs. Any tile that has a safety of 1-19?? has a .1%?? * (20-safety) chance of spawning a barb.
6.) Have travel stir up barbs. Entering a neutral or low-safety tile has a 5% chance of spawning an adjacent barb.
7.) Spawn waves. At turn 100 (+/- 10) (on Noble), any neutral tile that is next to a civ tile has a chance of spawning a barb. Repeat every 50 turns. Modify spawn strength by civ score; richer civs are more likely to attract more greedy raiders.
8.) What's Fallout without loot? Successful combat gains bottlecaps equal to strength defeated, lost battles lose similarly.

a.) With the above, you're controlling strength x number directly; no more jumping from Survivors to Securitrons in a few turns.
b.) AIs will feel safer (and be safer, between waves), so more expansion.
c.) More work to figure out how to do it (I have no idea).

Spreads:

Spoiler :

So long as buildings get flat bonuses, the best expansion strategy is spam -- as many cities as possible as close together as possible. It all seems very busy and not very Fallout-ish. Maybe increase the minimum city distance?
replace:

	<Define>
		<DefineName>MIN_CIV_STARTING_DISTANCE</DefineName>
		<iDefineIntVal>7</iDefineIntVal>	
	</Define>

with:

	<Define>
		<DefineName>MIN_CIV_STARTING_DISTANCE</DefineName>
		<iDefineIntVal>11</iDefineIntVal>	
	</Define>
	<Define>
		<DefineName>MIN_CITY_RANGE</DefineName>
		<iDefineIntVal>3</iDefineIntVal>
	</Define>

in GlobalDefinesAlt.xml (vanilla is 10 and 2).
One problem: with the current AI and barb strength, this makes AI civs expand even less. So this fix requires other fixes first.

 

[Lib.Spi't]

Have you tried my quick speed setting yet?

Lets look at these points:

Rads:
Pretty much all the things you said are in my plans. So lets break it down.

Spoiler :

Rad Tiles:
In fury road radiation was impassable and you could get a promotion (through research) that allowed units to move in it. This is something I want to reintroduce too. The alternative was fallout does damage instead, that then perhaps has a promotion that stops this (free for ghouls).
I think a combination of these two would be good eventually, it is just not a high priority at the moment.

Rad Improvements:
Eventually I want to introduce a whole set of 'ghoul' improvements that benefit from fallout, This will come at the faction specific development phase.

Roads/Routes:

Spoiler :

The point about adding in yield negatives, came from I think pie and his ancient mod, where it was thought that the negative food yield made the AI value them less and so spam them less, as it suddenly 'saw' a downside to having them (where as normally roads can go everywhere and have no downside) I think it was then observed that the AI made only 'routes' rather than spamming everywhere. With the addition of the new path types for this build, we are back to 'no-downside' route types, so we need to reintroduce the downside to then observe if it has the desired effect.

As to your 'explanation' about roads, you are in part missing what the roads in FTTW represent.
Eventually we will be introducing more 'powers' to roads, including jroute stuff like vehicles move fast on good roads, and people move normal.
What roads really represent though is safe passage for trade caravans etc. what a bandit is doing when he destroys a road is not running around with a sledge hammer breaking up bits of tarmac, he is raiding the travellers along that route and killing caravans. This then means the route is no longer 'safe' literally or perceptually, and it takes resources to restore those routes to being 'safe' and once again allowing the free flow of trade and 'stuff'.

Think of it as in a place like afghanistan, where the 'threat' of IEDs make most routes unusable, or involved elaborate logistics, massive convoys, and constantly changing methods, all because a route might contain a threat.
Routes in FTTW represent the knowledge that something is safe, as much as the functionality of having a bit of road to travel on.

I am still working through the ideas of techs or buildings having the power to 'unlock' trade across certain terrain types (like how coast and oceans work) which represents the logistics applications of travelling through 'unsafe' territory. (like heavily armed caravans and escorts). I just haven't got around to doing it yet. (or having someone create an underlying system to do it in a way that won't be over powered.)

Right now I am looking into ways to make the AI better at protecting itself from the new DOp threat, and using security forces to create a security 'net' over it's territory, so that it can also effectively make everything safe, just like the player can. Then there becomes the cat and mouse game of finding and exploiting weaknesses in each others 'net' or applying an overwhelming force to 'pierce' the net and reek havoc on an under protected area.

I want routes to be 'thin' (single lines, not spammed) so that enemies can move in and 'snip' them potentially knocking out whole cities or groups of cities, as they are cut off from crucial resources, and much of their economy and industry drops to a trickle. (I want this to be a big and viable strategy eventually, where bonus resources are almost more crucial to economy, than buildings. as the buildings will be next to useless without there attendant resources.)

One of the reasons the Union won over the Confederates in the American Civil war is that they were better at both breaking and fixing rail routes, so eventually a broken piece of rail was being fixed within days rather than weeks (I can't remember the actual ratio of times) and the important flow of information, resources and men were restored much sooner by the Union.

Raids:

Spoiler :

I am not sure what a lot of suggestions are specifically for?

As in, what effect does this give that is different to the game now?

Less Safety, More Safety? More/Less Aggressive Barbs? More Expansion Space?

The Vault and Ruin stuff I am confused about... They are features, they shouldn't be able to be pillaged?!?

A lot of the suggestions seem to just be a more complex spawning system, that would probably be very 'expensive' on memory, processing and performance. (I know things relating to tiles and movement can be very draining as they check all tiles every turn for every unit, or some such as this..) I am not sure what this system would net in terms of benefits to game play?

The AI just needs to be 'retrained' to go harder at barbarians and the like, especially when it comes to sending settler stacks and escorts into uncontrolled territory.

Specific unit balance for barbs should be an xml fix, like including barb only units with specific tech unlocks. I haven't got to that properly yet though, as right now I am more concerned with the player AI rather than barbs strength, I have seen some weaker nations overwhelmed by barbs, but not many. It is more about the player overwhelming the AI right now. Which I am almost 100% convinced comes down to the AIs irrational fear of barbs and lack of applied muscle to getting a settler where it needs to go.

One thing that used to happen in Fury Road was Barbs would spawn in uncontrolled or 'fogged' ruins, something like that might come back into play, right now we have lairs for animals. I want to expand this concept further to include ideas like bandit 'lairs' and robot 'lairs' a bit like what ciV has. At the moment Barbs are jsut a bit hap-hazard, eventually I want to mix a bit of logic into their dispersal and areas of influence/control. So more valuable areas (like blobs of ruins, animal bonuses, etc.) will have 'nastier' units and barb densities, as these are places where anarchic groups or large monsters could survive more readily. A lot of this will be about mapscripting I think. So the east will become progressively more deadly.
First the AI needs upgrading to make it more 'ready' to take on barbs effectively and fight to expand though.

I do like the idea of the highest scorers facing larger threats from barbs, that would give a nice scaling difficulty. Would need to figure that out though, maybe a new eventtrigger that relates to score, the higher your score the more likely you will be to trigger some nasty barb events.

Spread:

Spoiler :

There is a strategy that focuses on this exact form of play, where the goal is to just keep planting new cities constantly and how best to get around the constraints.
I don't think this requires making wider limits on city placement, as that just makes using the space you have less effective.. It might be the way to go.

I think it is more about getting the balance right (which will take a long time) between cities and improvements, and production/cost within cities.

Some system that makes having lots of tiny cities (landwise) more costly and at a point financially crippling, than having lots of larger well spaced and developed cities.

Maybe something like yield changes with techs that gradually make buildings have more upkeep and negatives to gold, while improvements (like towns+) get more gold yield, so in the late game if you spammed too many cities with 'base yield' outputs you find yourself in a financial sink hole...
That along with things like Number of Cities upkeep and such that you find yourself on the end of a losing policy if you go city spammy.

That will just take time to figure out exactly what the right math is for that outcome.

First I need to finish testing game speeds for quick and normal, so the game is playable at that speed, then need to fix the AI expansion and security issues.

That is probably going to be all my goals for the A5 release, so when they are done they will be a patch release and I will update the main A5 download as well.

I forgot to say, I like the idea of a Mr Raddy.

 

[Lark]

Raiding Redux:

The Vault and Ruin stuff I am confused about... They are features, they shouldn't be able to be pillaged?!?

I am certain I had some ruins get destroyed just after my borders had expanded to them. I was very annoyed. Ok, I bet I've figured it out: there are Ruins the feature which must be unpillageable as you say, and there are City Ruins the map improvement which uses the same graphic. This also explains why I couldn't build a Scavenger Camp the other day ...and I had it on a list of things to test - I was wondering if Decayed Urban Wasteland by a river made it fail. Well, that was amusing.
Ok, if they're already unpillageable I think it would be neat if:
4.) Raiders could set up Lairs on Vaults and Ruins in Civ territories which would give defense bonuses to defenders and have a chance of spawning more raiders.

I am not sure what a lot of suggestions are specifically for?

1.) Something to do for the first 100 turns. The need to defend occasionally and rewards if you want to actively seek out trouble.
2.) Something to do besides expansion in the second 100 turns. Also, the faster you expand, the harder this pushes back.
3.) Encouragement for the AIs to expand. It looks and is safer to settle, though new cities will need more defense.
4.) I read somewhere that 90% of the AI time was spent moving units. This removes 90% of the units so that almost every one is doing something. I would guess that the extra tile checks (once per game turn) in #5 and movement check (per unit moved, but after) in #6 cost much less in CPU cycles. But if my datum is wrong, my conclusion will be.

Basically, it's a non-AI system to make the game more fun for humans and easier for the AI civs. Right now it's too predictable and too hard, respectively.

Spoiler :

First 100 turns, you and the AIs build and research. You might pick up some places by optimizing the order, but you start tied, you end tied, and nothing happens in between.
Second 100, you start to make enough money for occasional improvements. You build Settlers and some Wanderers for escorts. You beat your neighbors to the easy spots. You tech for Wealth and Refuge and Scientific Study. You make a second worker, then more. You lose some Wanderers and start more settlements. You end with double the AI scores.
Turn 220, you see an Armed Survivor. Turn 230, you see a Securitron. If you have Raging on, they come after you and you find out if your defenses are good enough. Then they kill your neighbors. If it's off they mooch around and only occasionally send a probe. You're 3+x and rising by turn 250, while the AIs have mostly quit the sensible civics and quit researching to raise coins to rush and support and lose ever more low-quality units. You quit before you reach turn 300.

 

[Lark]

http://www.filedropper.com/showdownload.php/a5p1l03

Here is a rough playable version of a faster branch.
The goal is to play at Normal pace when Normal game speed is selected.

Back up your FTTW folder.
Put the contents of A5P1L0.3 into your test FTTW folder.
(Copy or move -> merge -> replace.)
Play.

What I have done so far:

Spoiler :

Reduced tech costs
Era tags changed: from Ancient (5) + Classical (5)
to Ancient (4) + Classical (3) + Medieval (3)
Cosmetic: changed all monasteries' prereqtech to NONE
Cosmetic: ditto missionaries (so both back to vanilla)
***Technologies\CIV4TechInfos.xml

Set Garden to have +2 Food, +3 at Gardening (from +4, +1 @CropRot)
Changed Farm from redundant +1 Food @CropRot to +2 @ Assault Vehicles
Changed Farmstead to carry irrigation
Changed Ghoul Farm to work on Rad Waste or Fallout (and nothing else)
Added upgrade tech prereq Prospecting to Scavenger Base
***Terrain\CIV4ImprovementInfos.xml

Moved Garden from Gardening to Gathering
Reduced build costs for early roads and producing improvements
Cosmetic: moved Sea Bridges from Electric Engines to Radio
***Units\CIV4BuildInfos.xml

Changed Wealth builds from 10-50% to 20/40/60/80/100
Changed all TXT_KEY_PROCESS_GOLD to TXT_KEY_PROCESS_WEALTH
Changed some wealth descriptions from explore to scavenge
Changed some buttons to ProcessResearch for science
Changed some buttons to ProcessCulture for safety
***GameInfo\Civ4ProcessInfo.xml
***Text\Fallout-TTW_Process_CIV4GameText.xml

Reduced building costs
Added 1 coin each to Foraging Hut, Snare Traps, Gathering Hut
(Vaulters already get 2+1 coins with Rat Cleanout)
Changed Library to gain 20 research from Books, not 200
Changed Caravan Market (Tech) to gain 10 Safety from Pre-War Tech, not 101
Changed Coal Plant to gain 5 research from Coal, not 50
Gave a prereqtech (Safe Haven) to Forbidden Palace
Added a Scrap Sorter building (a Junk Yard for Vault civs)
Changed resource buildings so Vault can't see Village and vice versa
Removed silly prereq from Gun Maker (essentially, required self)
Moved some resource buildings to r.b. sections of the .xml file
Added Vault Radio (an Eiffel Tower for Vault civs)
***Buildings\CIV4BuildingInfos.xml
***Buildings\CIV4BuildingClassInfos.xml
***Text\Fallout-TTW_Buildings_CIV4GameText.xml
***Text\K-Fallout_Buildings_CIV4GameText.xml

Reduced unit costs
Added Mister Raddy at Hydro Power to Scrub Fallout and Cover Rad Waste
Removed Scrub Fallout from other Worker units
Gave <FeatureImpassables> for Rad Waste and Fallout to all units except Rad Beasts, Ghoul Workers, Wanderers, Mister Raddy, Tactical Nukes, ICBMs, AutoTurrets and the Mega Robot
***Units\CIV4UnitInfos.xml
***Units\CIV4BuildInfos.xml
***Units\CIV4UnitClassInfos.xml
***Text\Fallout-TTW_Improvements_CIV4GameText.xml
***Text\K-Fallout_Improvements_CIV4GameText.xml
***Text\Fallout-TTW_Units_CIV4GameText.xml
***Text\K-Fallout_Units_CIV4GameText.xml


The earlier coins, Wealth and Garden is to speed up the early game (which is still slow). Cheaper Wanderers and more-expensive Stupid Mutant Fighters are to help with early balance. The Mister Raddy section adds management for radiated tiles. There are minor bug and cosmetic fixes.
Everything else is pacing increases.

What I have left to do:

Spoiler :

look again at what Lib said about Ghoul Farms and health
food is limiting, ask about adding some late in the tech tree
--probably not, if it's in the to-be-added techs
--farm resource discovery chance would increase food
--post my old Fallout resources & improvements file, maybe he'll see something
--particularly if he intends to copy C2C's buildings for resource combos
look at AI -brain hurts- no wonder he wants to change everything else instead
look at later techs (seem roughly okay but haven't gotten the AI there yet)
tinker & polish/get feedback/polish & tinker
do something else

The FTTW tech tree has very wide tech bars. I'd guess Platy's display automatically adjusted the one with the most icons - Capital City - to fit all its icons on its bar and set all the other bars to be that size. You're going to be making more civs - adding two icons per civ - and ending up with tech bars wider than my screen. The three cosmetics changes above are to help with that -- Capital City now only shows the generic monastery/campaign office and no missionaries and the second-longest bar is also shortened. Net: I now fit three bars where before I had not-quite two, soon-to-be one.

Vanilla Civ ground unit costs are ~8x their strength. I set FTTW units to be 40x before promotions. (Two exceptions: Stupid Mutant Fighter +50% because it's so strong so early and Wanderer -50% because it's needed to escort early Settlers.) This increased the cost of some units a little, but decreased them overall by half on average - more with later units. Also, I gave some units with a ranged attack some ranged strength, some units with ranged strength a ranged attack, removed "only defend" from two Vertis [no defensive bonuses, -50% city strength, +25% hills attack and could only defend??], and fixed the displayed spelling of Assassin and Dirigible Convoy.
I wasn't sure what to do with Robots. They have 10B/turn cost (except autoturrets) and get no promotions from experience. For the latter I thought I would take a third off the base value from my algorithm, but for the former... First, a flat 1000H reduction makes them too cheap and second, isn't the whole point of robots that, though they cost more to buy, they cost less in upkeep? But I'm trying to only change costs and a few bugs, so what I ended up doing was just halving the algorithm value except for the world unit. I would like to do the reverse: double the base cost but reimburse 1% of that in bottlecaps per turn as maintenance, training and oversight saved (except for the world unit). This would be good later on when the civics are balanced so civs don't have money pouring out their ears.

Rad tile management: the encouragement was slight, but this was something interesting I could do, so I did it.
Fury Road uses Python unitCannotMoveInto, but davidlallen said it made the game 5x as slow. He wanted to move to using FeatureImpassables on all but a few units, but the AI wouldn't make exploring units if he did that. In your mod the AI uses almost nothing but Wanderers and perhaps Spies for exploration, so if it's left off Wanderers the AI should explore just fine.
I didn't put in anything to make Mister Raddy faster later. I didn't see an easy way and it shouldn't change the game much.
You can decide later if <FeaturePassableTechs>, or a promo, or the purchase of potions like RadAway & Rad-X, or specialized units are worth adding -- or if Scrubbing is sufficient to get you through radiated tiles.
If you want radiation (from features or improvements or weapons) to hurt Wanderers and heal Ghouls, say, you'd probably have to add a Python check.

 

[Lib.Spi't]

Most of the content that we will be adding from now on will be .dll based, and will be whatever guys like archid, night and kailric can offer, but basically in almost every way .dll is better than python.

In terms of game speed, my goal overall for the game is make a grand strategy game rather than a fast strategy game. So my 'normal' speed will actually be 1000+ turns, likegames like Crusader Kings 2 where the game is played out over days/weeks/months rather than an evening.

The goal that I am trying to achieve right now is to get the quick and normal game speeds to play like fast civ players want, so 300ish and 500ish, these game speeds will than eventually be renamed to something like quick and casual.

These will be gamespeed file adjustments, more than base game adjustments as they are not the base game that I am heading for.

The information you are gathering is interesting though (although at times slightly confusing )

Especially as you seem to get through games much faster than I do, so you get much more examples of results. Your AI observations are also very helpful and interesting, as they help to pinpoint the areas where we really need to try and help the AI succeed.

On Robots:
The idea of robots is actually the opposite of what you are thinking (at least right now) they actually cost more in maintenance, because they do not cost population to build.

That is the goal I am moving towards, that robots will allow you to more easily maintain your high population economy and industry, but they will be a much bigger drain on your funds.

A lot of the accompanying content is still to be added though, or tweaked to fit and be right.

So units already have a min pop level, but I still need to do the same with buildings, and continue to populate the full roster of units and buildings.

Tech screen:
The box size is a ggod point and one I had not looked at or considered,it may be that at some point we can fix this so that each one scales to it's minimum required size or something, the tech tree is still not finished and some things will be more spread out, moved, etc. as more techs come along to fill in some of the missing concepts that will come along in the future.

Some stuff is just plain wrong, and was just carried over from vanilla tech tree, they need to be placed in their right and fitting places.

AI:
For AI changes I need to wait for codey guys like Night and Kail and Archid, to have some time, energy, or interest to look into it, as I cannot code and I can barely just sometimes read it.

So we need to first 'visually' ID an AI weakness, then locate all code connected to it in the .dll to see what it is actually thinking, then figure out how it needs to think differently, then figure out the right code to make it think differently.

I can only really do 2 of those things.

I have asked archid to make the expansion AI our next priority, so I am just waiting for him to be available to work on it with me. Right now he is focussed on more pressing issues for himself.

I believe(hope) the expansion issue is a relatively simple one, I think the AI just needs to know that it can put more of it's total unit strength into pushing out settlers into the waste, and that so long as settlers have that stack, it can risk being attacked by several units. This will hopefully make it more aggressive against the dangers that block it's expansion.

This one is my big priority for the next (final) patch for A5. Along with having a gamespeed file setting that gives the quick/casual gamer something to sink there teeth into.

Not enough to do:
On your point about the first part of the game being 'empty' and the middle part being all about 'settling' I do plan to try and increase the content in this part of the game. I am well aware that it is lacking.

My goal with it is mainly around the story event system (which I am still working the kinks out of).
So eventually we will have (hopefully) 30 or 50 or even 100 individual stories, so that is 50(or whatever ancient era events that you could possibly encounter. These then branch off into there own unique stories. (you can read more about the story event system on moddb.

Spoiler :

1. Story Events:
I have now managed to code a template for our brand new Wasteland Story Events.
(I hope, I still have to fully test it to make sure I did not screw it up somewhere!)

"What is a Wasteland Story Event?"
I hear you cry!
Well let me tell you!
It is our brilliant, innovative and unique
(at least I don't think I have seen it in any other Civ mod)
Event System!

One thing that Fallout and it's successors (and predecessors) have always been known for is:
Choices and Consequences.

Now very soon FTTW will (hopefully) be known for this too!

We will have oodles (currently almost 1!) of curious/bizarre/moral/ideological occurences and choices for you to make.

Something will happen in the lives of your little citizens, and they shall call upon you to decide what to do about it.

You will be presented with 3 choices, but you won't know the consequence of your choice until after you decide!

Not only that, but these seemingly inconsequential events will have unforeseen effects down the road, leading you down one of a number of unique storylines.

Each Story will consist of 3 parts
(we hope to make longer ones in the future but it is really hard!)
With almost 40 (39 + the starting event) unique choices/consequences in each!

However you will only get to see a handful of these each time you wander down the ever branching paths of fate and destiny!

So prepare to experience the grit and toil of the wasteland like never before with the

ALL NEW

WASTELAND STORY EVENTS!

That is a brief explanation.

But eventually my hope is that in that ancient era period, you will be facing probably 1 'choice' every few turns, lets say maybe 1/3rd of those 100 stories per game, so in the quick casual speed that be be spread across 50-100 turns.

Each of the choices you make in the beginning of your empire will lead you down ever branching paths in the future, so those small unimportant choices you made as a small town mayor, could have amazing or terrible consequence your future great nation!

So while you are grinding your way through surviving in your fledgling town, you will be constantly laying the path for the future of your empire.

That is the hope anyway!

 

[clanky4]

On Robots:
The idea of robots is actually the opposite of what you are thinking (at least right now) they actually cost more in maintenance, because they do not cost population to build.

That is the goal I am moving towards, that robots will allow you to more easily maintain your high population economy and industry, but they will be a much bigger drain on your funds.

A lot of the accompanying content is still to be added though, or tweaked to fit and be right.

So units already have a min pop level, but I still need to do the same with buildings, and continue to populate the full roster of units and buildings.

Robots fundamentally shouldn't cost anything to maintain. At least from a lore point of view there are numerous robots throughout fallout that have had little to no maintenance in over 200 years and still work perfectly, well to their programming anyway fine.

The main cost of a robot should be in getting it active. In other words reprogramming it to serve your purposes.

While you could potentially take on an already active robot you have to deal with the unfortunate actuality that its last standing orders were to defend some random facility that likely no longer exists. Therefore making them inherently hostile until they can be programmed otherwise, generally by deactivating and reactivating.

It would cost caps to get the needed supplies: fission batteries, conductors, power cells, possibly flamer fuel: as well as the needed expertise to repair and reprogram the robot. But from there it would have little to no cost.

If there is one thing to be said of fallout tech it is fairly reliable has an extremely long lifespan.

http://tinyurl.com/mrhouseknows

 

[Lib.Spi't]

I accept your point, although you do like to contradict yourself when it comes to robots!

You are constantly telling me that robots are too hard to build, then you argue here that they should be hard to build, the recipe you just laid out is more than the recipe I have in game that you have argued is too hard and should be easier!

In the future the maintanance system may change, right now though there is no way to add a gold cost to production (at least none that I know of).

Eventually when we have more resources and a more complex production system in place, the maintenance cost may be removed or shifted over to things like maintenance on the robot production facilities.

For now though we need a gameplay mechanic that stops everyone from spammming robots because they don't cost population, the maintenance cost seems like a good compromise in the short term.
As you said you need a strong economy to build robots, so it stands that the stronger your economy the more robots you can maintain and support.

 

[clanky4]

I accept your point, although you do like to contradict yourself when it comes to robots!

You are constantly telling me that robots are too hard to build, then you argue here that they should be hard to build, the recipe you just laid out is more than the recipe I have in game that you have argued is too hard and should be easier!

There should be a middle ground.

I do not know what the current setup for making robots in game is. Last I checked you needed pre-war tech and some other resource to make them, I think it was munitions. Some of the lower end robots could potentially use different raw materials, scrap metal and so forth.

Anyway main take away

Making Robot:
1.Beeline for Tech needed for unit and needed Resource
2.Get needed Resources
3.Build Robots
4.Get some money deducted from treasury when robot is built
5.Repeat step 3 and 4: &#8734; times
6.?
7.Profit

In the future the maintanance system may change, right now though there is no way to add a gold cost to production (at least none that I know of).

Eventually when we have more resources and a more complex production system in place, the maintenance cost may be removed or shifted over to things like maintenance on the robot production facilities.

For now though we need a gameplay mechanic that stops everyone from spammming robots because they don't cost population, the maintenance cost seems like a good compromise in the short term.
As you said you need a strong economy to build robots, so it stands that the stronger your economy the more robots you can maintain and support.

I distinctly remember seeing a mod somewhere which added a gold cost when a unit was built. That is when you build the unit the amount of gold is deducted. I cannot remember what the mod was.

 

[Lib.Spi't]

well, go find that mod again...

Otherwise it ain't much use to anyone!

I do plan to diversify the resources for construction requirements, not just for robots but other things too, I just need to do it. Right now I think almost everything is jsut munitions, I am not even sure if advanced munitions units use advanced munitions..

Right now I can't really be bothered to do anything until I can get the AI expanding like a pro, becasue it gets consistently outpaced by me mid game (and I am a bit crap) so until I can get it to keep up and punch hard, nothing else really matters as the game is over by the midpoint.. which is just rubbish!

 

[clanky4]

Right now I can't really be bothered to do anything until I can get the AI expanding like a pro, becasue it gets consistently outpaced by me mid game (and I am a bit crap) so until I can get it to keep up and punch hard, nothing else really matters as the game is over by the midpoint.. which is just rubbish!

So the AI are completely incompetent. Too many malices makes them afraid to do anything?

 

[Lib.Spi't]

hahahahaha!

 

[Lark]

I'm still stuck on getting Ghoul Farms to work. I don't think they're adding yields to Fallout the way you intended, and tile-worked health and happiness don't get added to a city. Below is what I've tried. Any more ideas?
Also, should I go through the four files (Bonus, Feature, Improvement, Terrain) and add some zero yields so they conform to the Ghoul Farm 5-yield format?
It would be tedious but hard to screw up.

The health think may be from our switch to health as yields, I may need to move the health boost to the yield section, as yield heatlh was added after the Gfarm. (you can try this yourself by adding in the 5th iyield tag (4th is happiness) if you feel more productive than me!).

Spoiler :

Fallout reduces yield by -3 -3 -3 to zero (it can't go below) and adds .5 ill-health to the nearest city if it's in its workable territory.
In my last check, Ghoul Farms should have added 9 2 2 0 4 to the tile (unirrigated, after Crop Rotation).
Oh, and 2 happy to the city for some reason - "we love parks" works.
Every kind of tile (land or water, peak or flat) with both Fallout and Ghoul Farms says it yields 9 2 3 0 4 (the 0 is just omitted, the 2->3 is from Financial). That is, instead of terrain+Fallout+GhoulFarm = 6 -1 0 0 4, it's giving (terrain+Fallout=0)+GhoulFarm.
And the health isn't being added to the city. Looking in the city screen the 9 2 3 is there but the 4 does nothing.

I tried 4 2 3 1 5 at game start:

	<YieldChanges>
		<iYieldChange>4</iYieldChange>
		<iYieldChange>2</iYieldChange>
		<iYieldChange>3</iYieldChange>
		<iYieldChange>1</iYieldChange>
		<iYieldChange>5</iYieldChange>
	</YieldChanges>

I got 4 2 3 1 5 displayed, 4 2 3 0 0 effective.

I tried going through TerrainInfos adding pairs of <iYield>0</iYield>, e.g.:

	<Yields>
		<iYield>2</iYield>
		<iYield>0</iYield>
		<iYield>0</iYield>
		<iYield>0</iYield>
		<iYield>0</iYield>
	</Yields>

and changed Fallout in FeatureInfos:

	<YieldChanges>
		<iYieldChange>-3</iYieldChange>
		<iYieldChange>-3</iYieldChange>
		<iYieldChange>-3</iYieldChange>
		<iYieldChange>0</iYieldChange>
		<iYieldChange>0</iYieldChange>
	</YieldChanges>

-but still see no health change in the city from Ghoul Farms, on Fallout or off.
Working a Ghoul Farm gives no change in health versus any other tile and in fact reduces health by 1 from not working it, since a Citizen gives 1 health.

Note 1: Mousing over happiness shows the contribution of specialists; mousing over health should too.

Note 2: If the city works the city tile and one more that also has 2 food, the city display says it's making 5 food, not 4. If the second tile worked is a Ghoul Farm with 7 food, it says 11 food, not 9. ??

Nice one, Clanky.

 

[Lib.Spi't]

You should not need to add blank yields to improvements (except to get the correct number of yields to count a value in the right place e.g. health is the 5th yield), everything is set so that it 'reads' a default if none are there.

I will get archid to take a look as at the health/happiness yield issue, when he has some time to see if he may have omitted something from his code.

I have added some of your other points to the bug thread.

 

[Lark]

I was briefly a Super Mutant today.
My old washcloth wore out, so before I walked in for groceries I dug out a new one. I happened to look in the mirror; I was green. Apparently this particular cloth, 2/3 of the way through the original stack, wasn't colorfast. Rather the opposite, in fact; even the towel end I used to re-wash with turned green. It would have been quite Podkaynian if I'd arrived at the store without checking my reflection.
The little used-to-be-white tag says Linencorp, made in China. Linencorp should dye.

 

[Lib.Spi't]

Lark Smash!

Good news (hopefully) on the settling/AI front, Archid found the code that hardcoded settler escorts to 2 units (I knew it was in there!) and increased it to 6.

He haas reported seeing a settler go as far as 10 tiles out of border to make a city.

I now just need to update my version and the patch folder to accomadate the new version, then I can see if it leads to significant improvements in the AI competitiveness, then I will be one step closer to releasing the last A5 patch and calling A5 'done' and move on to the next big move!

 

[Archid]

Mousing over happiness shows the contribution of specialists; mousing over health should too.

That was a copy'n'paste error where the specialist code ended up being inside the building code! Fixed in the code for the next release

Happiness & health yields for improvements were (are in A5) broken. Unfortunately they don't automagically work like other yields, they have to have specific code for every XML file they are used in. Looks like the original mod comp was hideously flawed

 

[Lib.Spi't]

Oh poop shoes!

One fix in! One fix out!

I also have to relearn trait point tags! it don't like the old ones no more!