View Full Version : New convert


Ahriman
Jun 29, 2009, 09:42 AM
Hey, just discovered this mod, and all I can say is wow!
Very awesome. I love the Dune books (well, the first 3 anyway) and I love how you've adapted some key mechanics; food scarcity, commerce through trade routes, inaccessible deserts, low-emphasis on teraforming (this makes the AI much tougher in a relative sense, since poor improvement construction is usually their downfall) etc.

Once I've played it more and have a better feel I'll probably come back with a more complete set of comments, ideas and suggestions, but just diving in now I'm very impressed.

I'm playing the 1v1 version, since I haven't yet updated to 3.19 as I'm worried it will break other mods (and we haven't updated Warhammer to 3.19 yet).

Initial things;
a) What does soil enrichment do? Seems to do nothing
b) Some buildings seem to have very little in the way of effects. Trade port just gives 10% culture? I presume the city wind trap provides fresh water to adjacent tiles? This doesn't show up anywhere.
c) It seems like light hover vehicles are king for a long long time - why would you build anything else? Maybe there should be a unit that gets a bonus vs them? Also, maybe they shouldn't get terrain bonuses? They're hover units after all....
d) Seems like all the factions are the same except for leader traits and starting units? Hopefully this will change. Or maybe I'm not far enough in yet?
e) What map scripts do you recommend? I started with Archipelago, which seems to be interesting but means no early game combat.

Also:
f) A sand storm captured my city? That doesn't seem right....
Can't you flag them as unable to attack cities, or -100% city attack strength, or something?

davidlallen
Jun 29, 2009, 10:58 AM
I'm playing the 1v1 version, since I haven't yet updated to 3.19 as I'm worried it will break other mods (and we haven't updated Warhammer to 3.19 yet).

I highly recommend you update to the 1.1.4 patch if you haven't. See this post (http://forums.civfanatics.com/showpost.php?p=8176701&postcount=342) for details. You can search in the thread for "posts by davidlallen containing the word patch" to find the changelists of 1.1.3 and 1.1.2.


a) What does soil enrichment do? Seems to do nothing
b) Some buildings seem to have very little in the way of effects. Trade port just gives 10% culture?

The improvements are rebalanced in 1.1.4. I agree the buildings could use more tuning. The best way to detect buildings which need tuning is to find buildings which users/AI rarely build. Which ones do you find like that?

I presume the city wind trap provides fresh water to adjacent tiles? This doesn't show up anywhere.

You may want to look at the FAQ post on the main thread in the forum, or get the same information from the hint tab in the 1.1.4 civilopedia. Both the building and the improvement provide fresh water, just like an oasis.

c) It seems like light hover vehicles are king for a long long time - why would you build anything else? Maybe there should be a unit that gets a bonus vs them? Also, maybe they shouldn't get terrain bonuses? They're hover units after all....

Tuned slightly in 1.1.4, but the unit progression needs more work. Please see the unit progression thread (http://forums.civfanatics.com/showthread.php?t=325388) for more discussion.

d) Seems like all the factions are the same except for leader traits and starting units? Hopefully this will change. Or maybe I'm not far enough in yet?

I agree the civs and leaders have "minimum" differentiation. There are some UU and UB which occur late in the tech tree, but we should probably improve that. Once the unit progression is better, then the right UUs will be more obvious.

e) What map scripts do you recommend? I started with Archipelago, which seems to be interesting but means no early game combat.

That's what I use. See also the Arrakis mapscript thread (http://forums.civfanatics.com/showthread.php?t=325293).

f) A sand storm captured my city? That doesn't seem right....

A bug. Please edit assets/xml/units/civ4unitinfos.xml. Find "SANDSTORM1", find underneath that <bNoCapture>0</bNoCapture> and change the zero to one. Do the same for the other three sandstorm units 2,3,4. I am sure that will be fixed in the next release or feature patch.

Ahriman
Jun 29, 2009, 12:22 PM
Will try these. Thanks again, this looks very cool.

Ahriman
Jun 29, 2009, 01:13 PM
Installed the patch... now neither soil enrichment OR drip farms do anything except on bonus resources?

davidlallen
Jun 29, 2009, 01:34 PM
In vanilla, there are improvements such as pasture and plantation which "activate" certain resources and have no other effect. In Dune Wars, the complex and the soil enricher are like this. For the drip farm and its upgrades growth facility and greenhouse, perhaps they should be buildable on "nothing" and allow larger food yield.

In most cases, upgrading the drip farm with the newer techs seems to provide enough food. But perhaps these improvements should be buildable on nothing. What do you think?

Ahriman
Jun 29, 2009, 01:55 PM
In vanilla, you can only build pastures on the bonus resources, its just confusing if they are available to build anywhere.

It seems you have two options.
a) Make drip farms and soil enrichment buildable *only* on the relevant bonus resources. This leaves wind traps as the only food resource improvement (is that your intention?).
Also, this makes fresh water access irrelevant; yes, you need fresh water to build these things except on bonus resources, but doing so is useless
Hence things like the city windtrap serve no purpose.
b) Give them a base yield of +1, a further +1 on the relevant bonus resource, and have the require fresh water access in order to build.
They're like a farm in vanilla; you either need fresh water, or you can still build them on a rice/wheat/corn bonus resource. And you eventually get a bonus food on the rice/wheat/corn if you connect it up to fresh water.

Another comment:
It may be because I'm now playing Atreides (financial trait), but I'm completely breezing through the techs in 5-6 turns each, and the game is going by way too fast.
In my first game, commerce was really hard to get; because cottages didn't grow (they were just a +2 commerce bonus), most commerce income came from trade routes. I really liked this mechanic; it made the mod play quite differently, and made a peaceful builder strategy as attractive as a conquest one because of the need for open borders and bonuses for sustained peace.
After the patch, there are cottages, and the spice blooms (I often have 9 or more spice resources with a total of 5 cities) and the spice corporation boost means my economy is pretty crazy.
Now though my tech advances so quickly that I only ever build a handful of units before they become obsolete and the number of buildings available increases faster than I can build them.
I've discovered 35 techs by turn 170. I'd think a new tech every 9-10 turns is more appropriate than a new tech every 5 turns.

So, I'd suggest significantly increasing a lot of the tech costs, or toning down the commerce levels.

Partly the problem is that commerce yields are outclassing hammers. In vanilla, I build mines on hills for hammers, and farms and cottages on flatlands.

Here, I'm building cottages on flatlands, and windtraps and mines on hills. So there is tons of commerce, but only a little hammers. Increasing the hammer yields of improvements might help too.

Ahriman
Jun 29, 2009, 02:56 PM
Another reason for too much commerce; cottages here give +2, hamlets +3 etc whereas in vanilla cottages give +1, hamlets +2.

So with financial trait, building a cottage is immediately a 3 commerce yield, rising to 4 commerce after 10 turns, and then further.

Drop the cottage back to vanilla would help, as would providing some more hammer yields from improvements (its very hard to get more than 1-2 hammers without a bonus resource).

Ahriman
Jun 29, 2009, 04:34 PM
You've done the same thing with mines; the mines are now useless except on bonus resources.

I don't understand the logic for this at all. At the moment, all terrain improvements either come from bonus resources, windtraps, or cottages. And you can get a handful of hammers mid-late game from turbines. You have to leave tubers most place because you need them for hammers.

Why not put surface mines back to +1 hammers and deep mines to +2 hammers, and put drip mines to 0 food + 1 food with fresh water, and growth facilities to 1 food +1food with fresh water?

I don't see why cities should only be viable around bonus resource clusters. They should be *better* with the bonus resources, but not useless without them.

Also
Bug: Hornet carrier can't actually have hornets loaded onto it.

davidlallen
Jun 29, 2009, 05:08 PM
This is excellent feedback. I do not claim to be an expert at balancing the bonus and improvement values. I am pretty sure 1.1.4 is better than 1.1, but I am sure it is not perfect. Please have a go at modifying the bonus and improvement files, playtest what you get, and upload it if you think yours is better. We can all work together to make the mod really great.

Ahriman
Jun 29, 2009, 05:41 PM
Sadly I'm no coder, I don't even have an xml editor installed and wouldn't know how to use it. I'm just a suggester :-)

Some other things that affect this:
a) The Archipelago mapscript means that most cities have lots of sand (commerce), and only 1-2 hills for food.
b) Financial trait
c) Lack of early game buildings with decent hammer bonuses or access to hammer-producing specialists.

There are also I think too many buildings in the mod; happiness is really never a problem, because there are *so* many happiness providing buildings.
I'd consider culling the total number of buildings significantly, by maybe 20% or so.

davidlallen
Jun 29, 2009, 11:42 PM
Changing the xml is really easy, and fun too! No tools required besides notepad. Open up file dune wars\assets\xml\terrain\civ4improvementinfos.xml in notepad. Use file explorer to make a backup copy, with a name like "myimprovements.txt". (It's safer if the backup does not have the same file extension.) In the xml file, search for COTTAGE, all upper case. The XML looks like this:


<ImprovementInfo>
<Type>IMPROVEMENT_COTTAGE1</Type>
<Description>TXT_KEY_IMPROVEMENT_COTTAGE</Description>
<Civilopedia>TXT_KEY_IMPROVEMENT_COTTAGE_PEDIA</Civilopedia>
<ArtDefineTag>ART_DEF_IMPROVEMENT_COTTAGE1</ArtDefineTag>
<PrereqNatureYields>
<iYield>0</iYield>
<iYield>0</iYield>
<iYield>0</iYield>
</PrereqNatureYields>
<YieldChanges>
<iYieldChange>0</iYieldChange>
<iYieldChange>0</iYieldChange>
<iYieldChange>2</iYieldChange>
</YieldChanges>


The yield change section is food, hammers, commerce. So this is the part that gives cottages 2 commerce. Change the 2 to a 1. Now find <Type>IMPROVEMENT_HAMLET2, and change the iYieldChange from 3 to 2, and decrease village and town the same way. Save, start a new game and voila! You can see if your change makes the game better.

If you make this change successfully, you can probably look around more in the same XML file and try out a bunch of your own suggestions. Changing the numbers in the xml files is easy, and it is hard to mess up the game. If you do mess up the file, and get syntax errors when the game loads, copy your backup file and try again. XML syntax is not so scary. Python syntax is a little more scary, that is an actual programming language.

koma13
Jun 30, 2009, 03:03 AM
Sadly I'm no coder, I don't even have an xml editor installed and wouldn't know how to use it. I'm just a suggester :-)


Only suggesting is fine too. It's what we need most right now. :)

Ahriman
Jun 30, 2009, 06:41 AM
Biggest other suggestion would be culling the number of units and buildings.

The build menus get very cluttered. It would help if a number of buildings were removed (particularly a lot of the weaker ones that will never get built - like the trade port (10% culture) and the reception call (+1 happy per 20% culture slider), and if some units were moved or at least became obsolete and so were shifted out of the build menu.

Another suggestion; make harvesters able to cross desert (not deep desert) tiles in the same way that settlers can.
This might make the AI a little better at expanding, particularly on archipelago map scripts, where it often only has 2-3 cities after a couple of hundred turns, whereas I have 8-9.

Ahriman
Jul 01, 2009, 03:41 PM
Another issue, and a big part of why the game runs so fast; you need to consider toning down each faction's spice guilds. Maybe make it maybe +3 commerce per spice, down from +5. Or reduce the spawn rate of spice blows.

At the moment, my capital (with my spice corp in it) runs basically my entire economy on the spice corp; its easy to get 15-20 spice resources, which can be 100 commerce right there, multiplied by all the bonuses from bank, university, etc. etc. can easily be 250+ beakers/gold right there.

Its also devastating for the AI performance since sometimes the AI doesn't even found its spice corp until turn 80+.
So my economy powers on past theirs.

I'd also suggest toning down (or removing) inflation.

davidlallen
Jul 01, 2009, 05:10 PM
That is interesting feedback. I have chosen the value of 5 commerce by comparing what the AI does using autoplay. I have found that in a vanilla archipelago map, the AI's have a typical per turn commerce around 250-350 by turn 250. Using the current Dune Wars commerce rates with a similar map, the per turn commerce is around 100-150 by turn 250. So the AI's, at least, grow slower than vanilla.

Since you find as a human player that you can get huge commerce rates, the problem must be that the AI is inefficient at growing and using this commerce. Maybe setting the iAIWeight of the spice corporation building would make the AI more interested in building it.

I will encourage you to change xml files on your own, one more time and then stop.

With an active playtester such as yourself, it is very valuable to make small changes to the xml on your own and try them out rather than waiting for a release team to make a patch. In this case the problem is slightly worse because any patch I make will be a 1.2.x which will require you to upgrade to BTS 3.19 to use it. Plus right now 1.2 reliably crashes to desktop around turn 300.

To try this out, I have explained somewhere recently about how to change the commerce yield but I do not have the game files here at work. Approximately, in xml/gameinfo/civ4corporationinfos.xml, find the "500" in the yield section. For a reason I don't understand, the yield per bonus is in units of 100, so 500 represents 5 commerce per spice bonus. Turn it down as low as you like to see how fast your own economy grows.

To make the AI more interested in the spice corporation, look in xml/buildings/civ4buildinginfos.xml and look at the City Windtrap. Just under that you will find a line "<iAIWeight>10</iAIWeight>". You will notice that all the other buildings have a weight of 0. This field is a weight factor which makes the AI build buildings even if they don't look "interesting". Find the unique corporation buildings such as CORPORATION_7 (the names are not so obvious) and change all their iAIWeight's to 10. Does this make the AI build these corporations sooner? If so does that help their economy grow faster?

Ahriman
Jul 01, 2009, 08:38 PM
I'm sorry, but I really don't really have the time to start learning xml fiddling. I realize its not that hard, but its not really time I have.
Its also a deliberate commitment device; as long as I don't install an xml editor, then I'll limit my modding time somewhat, and actually get my dissertation finished :-)

Since the guilds only cost 1 hammer, setting the AI building priority on the guild super high should be harmless.
Another idea: make the guild structure require the Palace, to make sure the AI builds it in their capital, which will probably be their best city, and will have the most gold/beaker/trade boosters and the benefits from idiomatic?

Unless the problem is they aren't building deep harvesters - or they're just using them to explore?
Maybe they aren't getting the spice resource early enough. The AI shoudl build the spice corp in their capital as soon as they get their first spice resource.
Another suggestion; maybe even just remove the requirement for a spice resource and set the AI build priority really high? That should make it build the thing as soon as they get the spice extraction tech. Does it have a gold maintenance cost if no resources are being harvested?

The Suggestion menu keeps suggestnig to build deep harvesters to me, so presumably they're building enough.
I also noticed a weird thing; if the deep harvesters were in a non-"coastal" city, sometimes even when automated, they wouldn't go off and start building refineries.
I wonder if this is because they're using workboat AI, and the workboat AI only looks for improvements to build in ocean terrain that they're connected to?
No obvious fix here, except maybe to stop the deep harvesters from being able to go on land, so they're always on sand or in a "coastal" city.

Partly the poor AI economy is because they aren't expanding enough; the AI still gets plenty of commerce from its spice guild once they build it, but they're not getting much else, partly because they don't realize that hills are the only non-bonus that give you food, and so choose city placement poorly.

AIs still have 2-4 weak cities when I have 8.

I think they also might be building drip farms and soil enrichment on base tiles; the worker AI seems to recommend this sometimes, even when it does nothing except spread fresh water.

Adopting my idea of putting drip farms back to +1 food with fresh water, and don't spread fresh water, and require fresh water (or bonus resource) to build, and making soil enrichment unbuildable except on bonus resources, would probably fix this too.

Final though: there is another corporation (spice merchant guild, maybe CHOAM too?) that also consumes spice. Since you can have dozens of spice resources (but only a handful of normal resources), this corporation is also waaay too powerful and can lead to an insane economy.

zup
Jul 02, 2009, 03:32 AM
Notably, these other spice corporations compete with the house spice corporations. Of course, that's just one city you cannot get them in. But then, I absolutely want to found all of my corporations in my capital (my pet city).

Ahriman
Jul 02, 2009, 06:26 AM
Try founding one of the spice guilds in the shrine city of a widespread religion.

Last game I had, my capital had 500 beaker output and my holy city had another 350.

These guilds are orders of magnitude more powerful than the other corporations.

Deliverator
Jul 02, 2009, 07:10 AM
Well, we want the spice to be the primary source of commerce, but clearly there is some balancing required. :)

davidlallen
Jul 02, 2009, 08:23 AM
I have just posted a conceptual problem with the "water for health" project, but I think there is some economic tweaking I can do easily. Here is what I plan as a 1.2.2 economy patch.

* Decrease commerce of cottage/village/town by one to match vanilla
* Replace harvester graphic with desert worker, give it same desert crossing capability as scout, soldier, settler (also fix sounds on these)
* Remove second spice corporation
* Using iAIWeight and/or python, force AI to build spice corporation ASAP
* Change drip farm to give +1 food on any tile which gives food and has fresh water
* Add "cannot capture city" to sandstorm

After the above changes, test commerce per turn capability of AIs after 250 turns. Presently the AI lags significantly behind an equivalent AI on vanilla in total commerce. If the above changes fix this, only then will I reduce the commerce per spice bonus. The underlying problem is that the AI is not using this efficiently; if I reduce it so that the player economy doesn't grow as crazy, and it completely cripples the AI, then we are not getting very far ahead.

Are there any other obvious economy related changes that I missed in the above list?

Ahriman
Jul 02, 2009, 08:37 AM
Well, we want the spice to be the primary source of commerce, but clearly there is some balancing required.

Spice as *a* major source of commerce is a cool mechanic (makes the mod different, adds flavor), but the problem is how it all gets concentrated into a single city, and so building a ton of % bonuses in that city skyrockets your economy insanely.
Maybe try to find some way of spreading the wealth through multiple cities? Cut down the bonus per spice resource, but make the corporation spreadable (or automatically spread) through all your cities?

@david
Make soil enrichers buildable only on bonus resource, so the AI doesn't build them uselessly to spread fresh water.
Make the drip farm require fresh water to be build (or bonus resource), like farms in vanilla.
Put mines back to +1 hammers, deep mines to +2 hammers, so mines built on hills give a hammer boost, OR make them buildable only on bonus resources, to make sure that the AI builds windtraps on hills and so gets fresh water for their farms.
Strongly recommend: make the spice guild require the palace, to make sure that the AI builds it in its capital.

Consider: removing fresh water spread from drip farms.

But also: what matters is not AI commerce relative to vanilla, what matters is AI commerce relative to human player commerce.
If the human player can easily exploit the system to gain an advantage over the AI, then thats what makes the mod too easy and breaks the system.

If the human and AI economy growth are similar, but lower than vanilla, you could easily fix this by lowering or removing inflation.
I hate inflation in civ; its just a drain on your resources, and there is absolutely nothing you can do about it. I was finding by turn ~200-250 that inflation was getting near 100%. Ie all unit and civic and city upkeep costs were doubled. Thats a huge economy drag.

In your simulations, try to note what turn number the AI builds its spice guild - it should be recorded in the noteworthy events log (cities built, wonders built, corporations and religions founded, etc), right?

davidlallen
Jul 02, 2009, 11:08 AM
Spice as *a* major source of commerce is a cool mechanic [...] Maybe try to find some way of spreading the wealth through multiple cities?

I will try some experiments to see how the commerce bonus gets allocated among multiple corporation buildings. If each building gets the closest ones, then having the buildings automatically appear in every city will distribute the wealth. At least that will defeat the strategy of filling one city with +% bonuses.

Make soil enrichers buildable only on bonus resource, so the AI doesn't build them uselessly to spread fresh water. Make the drip farm require fresh water to be build (or bonus resource), like farms in vanilla. Put mines back to +1 hammers, deep mines to +2 hammers, so mines built on hills give a hammer boost. Consider: removing fresh water spread from drip farms.

Soil enrichers buildable anywhere is a bug, which I did not intend. I will make the above changes.

But also: what matters is not AI commerce relative to vanilla, what matters is AI commerce relative to human player commerce. If the human player can easily exploit the system to gain an advantage over the AI, then thats what makes the mod too easy and breaks the system.

I agree. The measurement I described is one I can get easily by letting my computer autoplay several games while I do other things. Your measurement requires me to play the game carefully, which is more time consuming. Getting this information from playtesters via suggestions is also helpful, of course, but this loop is also time consuming.

I hate inflation in civ; its just a drain on your resources, and there is absolutely nothing you can do about it. I was finding by turn ~200-250 that inflation was getting near 100%. Ie all unit and civic and city upkeep costs were doubled. Thats a huge economy drag.

I never understood the motivation for this either. But as far as I know, the game engine does not expose any "hooks" to allow changing this. At least it is the same drag on AI players as human players.

Ahriman
Jul 02, 2009, 12:00 PM
Agree that corporation mechanics need testing, I don't know what the formula for corp maintenance is or how it works. But definitely spreading the bonus would be very useful.

Understood that automated testing is much easier, and it tests at least some of what you want. It definitely helps test the effects of various design changes; run 10 games with one set of terrain improvements, implement some changes (like tweaking farm values), run 10 more games.

Re: inflation
I know its possible to eliminate entirely; FFH (or Fall Further at least?) has a game option that sets an inflation multiplier to zero somewhere.

Also, I know that in the Warhammer mod and in FFH there are some events that reduce the current inflation multiplier for your civ.
So in a worst-case fudge you could at least make an event that occurred every x turns that reduced inflation by y%. Hardly optimal though.
The other thing I hate about inflation is that it disguises the true impacts on gold from various things; if you have a city with an upkeep cost of 6 and are considering building a structure that reduces city maintenance by 50%, the player normally evaluates the value of that building by thinking that it will save 3 gold per turn. But if inflation is at 100%, the building will actually save you 6 gold per turn.
Similarly for civic upkeep costs and unit maintenance costs. It messes up evaluating the value of different options.

koma13
Jul 02, 2009, 02:26 PM
Here is a modified CvGameCore.dll without inflation. I also removed inflation display from finance advisor:

http://www.mediafire.com/download.php?q2ygcn35utj

davidlallen
Jul 02, 2009, 02:36 PM
Here is a modified CvGameCore.dll without inflation. I also removed inflation display from finance advisor:

http://www.mediafire.com/download.php?q2ygcn35utj

Is that compatible with BTS 3.19 and DW 1.2.x, or with BTS 3.17 and DW 1.1.4?

koma13
Jul 02, 2009, 02:38 PM
BTS 3.19 and DW 1.2.x

CvPlayer::calculateInflationRate() now returns always 0. I hope this is enough to disable it. I didn't touched the inflation offset.

Pfeffersack
Jul 11, 2009, 11:17 AM
(I hope it is ok to hijack this old feedback thread, as I don't have seen one especially dedicated to give feedback for patch 1.3)

First of all I want to say that I enjoyed my first game of Dune Wars a lot :) The grphics do a nice job of capturing the atmosphere of a desert planet and I like the differences to the vanilla game - like a spice based economy.

I haven't got a real feeling for the mod yet, so I don't have a feeling for balance yet, but here are some (mostly cosmetic) issues I discovered during my first 100 turns:

- Right at the beginning I get notified that the altroot.py is in the wrong direction (I still have it copied over in the BTS directory from my attempt to play with Dune wars 1.2)
- I found the groundwater ressource nearly impossible to spot without the ressource overlay
- Some events behave weird. Often the choice you got offered consist only of a single word, which is the same for all. Then one time my scout passed a desert and caused the "found black oil" event - but the second option of paying 10G for discovering all spice wasn't available (despite having enough Gold). Perhaps the save helps to track that issue down (if it is one)
- Since that event I get the two python errors every few turns (see pics)
- You still get "Gold" (isn't the currency on Arrakis "Solarii"?) from tribal huts, also a computer opponent "circumnavigated the globe"
- The hoover help for the Space Port tech talks of improving chop yield by 50% - is there anything left to chop after the removal of tubers?
- The tech Xeno Botany allows "Desalinating a Salt Pan" - is that just a worker command for clearing tiles covered with the feature "Salt"?

Deliverator
Jul 11, 2009, 11:24 AM
Hi Pfeffersack. Glad you had fun with your first game, and thanks for the great feedback. I can't comment on all of these issues, but I can tell you that I am working on a 3d graphic for groundwater right now. At the moment, it is literally invisible so there is no wonder you can't see it!

I'll probably be doing a graphics only patch at some point this weekend.

davidlallen
Jul 11, 2009, 11:52 AM
@ Deliverator, thanks for creating the feedback thread, you beat me by a few minutes.

@ Pfeffersack, please see my replies (in a few minutes) on that feedback thread.