[Religion and Revolution]: Bugs and Todos

[Tigranes]

There seems to be a big problem with valuable wood/furniture. All my European competitors are literally doing nothing else than collecting VW and making furniture from it.

Saved game can be provided, if needed.

I can absolutely second it, actually I have reported about amazing affection with WV earlier. New France would not sell me 100 WV for more than 4000 gold

Europeans are indeed very weak, and quite unreasonable during wars. One time Danes had a huge Army in remote village. I captured their capital and all the prime cities with few units being sure that Army will come back and reclaime lost cities. Instead they sued for peace, it really felt like cheating...

 

[dmalleus]

In all of my games (release 1, faerie weather mapscript) I am finding the Inuit natives in a really off position: the equator. Equatorial Inuits is... odd, to say the least. The oddness doubles when you combine their map position with their (very excellent, might I add) civilization graphics. Just wanted to mention this, though it may be a rare problem that only appears on my end?

I'm playing a New World map, and I also met "Equatorial Inuits" in the middle of the jungle, lol.

I also noticed that AI trades with valuable wood/furniture a lot. Not exclusevely, but mostly.

In addition, ff Malinche and Jacques Marquette are usually able too late to be fully effective (map is already mostly explored by then). Still, an event with naked women dancing in the trees helped me to get Malinche when it's trait "see all burial grounds" was still usefull.

 

[raystuttgart]

I'm playing a New World map, and I also met "Equatorial Inuits" in the middle of the jungle, lol.

The placement of Natives uses a "Bubble Sort Algorithm".
But it does not influence which Nations are chosen.

So if you are playing on a small map where for example many "North American Natives" are chosen,
it might happen, that some of these are pushed to the central of the map by the other "North American Natives".

The bigger the map and the more Natives chosen, the better the algorithm will work.

The algorithm will simply try to place the Natives randomly chosen by the game according to their XML setting for Northern, Central, Southern.
But it still needs to ensure an equal distribuiton of these Natives over the complete map.

All the algorith ensures is:
Northern American Natives further North than Central American Natives and Southern American Natives.
Central American Natives further North than Southern American Natives.

So yes, it is possible that Inuits appear in the central of the map, if they are pushed down there by other Northern Natives.
(Depends on the map structure and which Natives are actually chosen by the game.)

The best approach to dealing with bugs/fixes is to be open and honest.

Where have I not been open and honest ?

I said I will take a look.
I also said, that decisions of AI are not simple but include several factors, so I will need a bit of time.
Then I found the time and took a look and found a bug.
Then I told in the forum that there is a bug.
Then I told in the forum that I fixed the bug.

What is not open and honest about that ?

If I do not know there is a bug, I cannot say that.
Things are often not that simple that I can know without checking.
Also I am wrong sometimes, too.

Also please excuse if I am currently to busy to do long explanations about AI.
I am already happy if I find the time to analyse and fix and take care of all the other things in this project.

Last but not least:
All team members and partners have access to SVN.
They can see everything I do.

-----------

Guys, the problem with Valuable Wood / Furniture is already fixed.
(And even more improvements to AI decision about Yields have been done.)

On Sunday you will get the new Release 1.1.

The complete Founding Fathers are in work.
(New ones and their order.)

 

[raystuttgart]

As for the pink screen error I pointed out earlier, it still exists. I even installed a 1Gig DDR3 Nvidia 9800 GT EVga video card which makes the game run maximally smooth. The pink screens are still there.

You are playing the same mod as the others which do not have that problem.

The only thing I can think about (and which I have already told you here):

Maybe it is about the screen resolution you are playing with.
(TAC and Religion and Revolution are not meant to be played with extreme wide screen resolutions.)

Did you check that ?

 

[orlanth]

I have only been able to test a little bit, but can confirm that the problem with VW AI yield balance seems much improved in the latest revision. I'm thinking this could really improve AI competitiveness.

Europeans are indeed very weak, and quite unreasonable during wars. One time Danes had a huge Army in remote village. I captured their capital and all the prime cities with few units being sure that Army will come back and reclaime lost cities. Instead they sued for peace, it really felt like cheating...

Hmm that's too bad.. I haven't played the latest version mod much yet with colonial wars, do you think the AI war planning has any worse or different behaviors compared to TAC? I'm hoping some of the AI's problems may have been due to the VW / yield planning error which is now getting fixed; if the AI war strategy is still poor maybe we could try something such as having them keep extra Cannon Garrisons in cities.

ff Malinche and Jacques Marquette are usually able too late to be fully effective (map is already mostly explored by then

That's a good point.. I think it would be good if exploration in general becomes a bigger challenge; in vanilla it's really far too trivial and uninteresting to explore everything quickly and easily with a single explorer (not to mention using Treasures to explore in vanilla, which now are more vulnerable in this mod). In my opinion this was a significant problem with the vanilla game : exploring = boring! Adding some Wild Animals will be a big benefit for adding interest and challenge to exploration, especially if killing them can harvest some goods for a nearby city. Possibly in addition to Animals we could occasionally encounter some hostile Native units controlled by neutral AI, to represent some local minor Native tribes that aren't affiliated with one of the main Native civs.

On the other hand, if Wild Animals are too powerful people may get frustrated with exploration. I wonder if we can use some already existing unitinfos.xml features like <withdrawalprob> or even <EvasionBuildings> for units like Expert Explorer and Ranger, so if they're defeated by Animals they can retreat and wait a few turns to heal, or possibly even withdraw to the nearest settlement with the Evasion feature. This will make exploration more of a challenge where you'll need to work at it with more than just 1 Explorer, while still making it achievable and not too frustrating.

 

[raystuttgart]

AI war planning has any worse or different behaviors compared to TAC ?

AI logic considering wars has not been touched.

We do have a different balancing and totally new nations though.
The Danish leaders for example were extremely peaceful. (Already changed a bit.)

This aspect Aggressiveness / Passiveness is not about bugs.
It is simply about balancing.

And since I do not have time for playing, I can adjust balancing only by the feedback I get.

----------

But AI economy had a big problem.

Release 1 had a major bug here, that heavily threw back AI.
(And I have done more improvements than simply fixing that bug.)

----------

Let us see how things behave with the next Release 1.1.
(Everything else would be pure speculations ...)

 

[orlanth]

thanks. (btw I was editing my post above while you were responding - do you have any thoughts re ideas to make exploring more challenging using Animals, and using Withdrawal or Evasion for Explorers / Rangers?)

 

[raystuttgart]

do you have any thoughts re ideas to make exploring more challenging using Animals, and using Withdrawal or Evasion for Explorers / Rangers?

Well I guess we should first of all have Wild Animals at all.
(It is planned, but we do not have a modder working on it.)

Then we could see how it feels, how it effects exploration and what additionally should be implemented.

Also discussions like that should probably take place in the thread for Mod Development.

 

[Commander Bello]

That's a good point.. I think it would be good if exploration in general becomes a bigger challenge; in vanilla it's really far too trivial and uninteresting to explore everything quickly and easily with a single explorer

That's the reason why in my modmod for TAC I have limited access to peaks to only pioneers.

This has quite some effects: you now can face bottleneck situations, which make placing cities interesting AND exploration becomes more interesting.

Currently, an explorer is only slowed down by peaks. On the other hand after having climbed that peak, he has even an increased view. The result is that the map is revealed in almost no time.

Don't allow the explorer to enter peaks, and mountain ranges not only become an obstacle, but you don't learn what's behind. You have to move around. And this may take some time depending on the map size.

 

[raystuttgart]

Currently, an explorer is only slowed down by peaks.

Seasoned Scouts can enter Peaks.
Normal Scouts (simply colonist in the profession) cannot.

That is how we the feature was planned and implemented.
I really do like it that way.

But you have already changed it to the way you prefer.
(So everybody is happy. )

Play a map on "Gigantic".
You will not complain about map being discovered to early anymore.

 

[Commander Bello]

Well, I already maintained your kind of "gigantic maps" to be really gigantic. And it was good that I forbid scouts to enter peaks.

But as you said, the feature was implemented as you like it.

 

[frontiersman]

Isabelxxx posted the following in the 'Dawn of a New Era' subforum, thread 'The Next Goals', post #34:

Fixed most of the bugs and assert messages I was having while using a Debug DLL.

which reminded me that I ran into similar 'Assert failed' messages when compiling a debug version of 'Religion and Revolution', which prevented me from playing a game with the debug DLL. Note that I can successfully compile a 'release' version DLL using the supplied source files, and play the game with the release DLL, but for development purposes, it's good to use the debug version to help solve bugs for any code changes. Here are the two assertion errors I ran into trying to launch with the compiled debug DLL:

Spoiler :

Assert Failed
File: CvXMLLoadUtilitySet.cpp
Line: 1373
Expression: bSuccess
Message:
________________________________

Assert Failed
File: CvDLLButtonPopup.cpp
Line: 917
Expression: false
Message:
_________________________

The first assert happens when launching the game as the initial screen appears. Choosing "Ignore All" allows the game to proceed to the menu. I can choose a custom game with all the options. Then, launching the game from there, it freezes. Minimizing and maxing the game a few times (and clicking on the game in the task manager) can move it along (sometimes) to the opening scene on the ocean, where it crashes. The second assert message can (sometimes) be seen at this point.

I hope you can investigate and fix this problem with the debug DLL asserts for the new 1.1 release which we are all waiting eagerly for.

 

[raystuttgart]

... but for development purposes, it's good to use the debug version to help solve bugs for any code changes.

I know.
I am working with Debug-DLLs all the time.

I hope you can investigate and fix this problem with the debug DLL asserts for the new 1.1 release which we are all waiting eagerly for.

I will check.
(But Assert messages do not necessarily mean that it is a real bug.)

 

[Isabelxxx]

(But Assert messages do not necessarily mean that it is a real bug.)

That's true. Be aware I said that in DoaNE because they were real bugs, or better said: non optimal code. But don't extrapolate my words to all, although we should make the new things compatible with the existing assert checks and don't ignore them...

They didn't affected the game play at all, but those asserts tell you most times that you are trying to do things that are not supposed to happen.

For ex. in DoaNE the AI was trying to use the ships in the Ships Market for trading, but since those are not supposed to be able to do anything but wait until the player bought them the DLL was showing messages every turn due to that. As said, most times there is not a bug in the sense that something is broken, but in the sense that even if it's not harmful what you are trying to do with the code is probably wrong.

Hope this clarifies the topic, but sure Ray already knows well what to do.


What frontiersman shows clearly appears to be something to consider, because that is an error in SetGlobalClassInfo, which is not natural at all...

 

[raystuttgart]

What frontiersman shows clearly appears to be something to consider, because that is an error in SetGlobalClassInfo, which is not natural at all...

I will see what I can do and when I find the time to do it.
(But maybe this will be taken care of in Release 1.2. ... )

 

[dmalleus]

That's a good point.. I think it would be good if exploration in general becomes a bigger challenge; in vanilla it's really far too trivial and uninteresting to explore everything quickly and easily with a single explorer (not to mention using Treasures to explore in vanilla, which now are more vulnerable in this mod). In my opinion this was a significant problem with the vanilla game : exploring = boring! Adding some Wild Animals will be a big benefit for adding interest and challenge to exploration, especially if killing them can harvest some goods for a nearby city. Possibly in addition to Animals we could occasionally encounter some hostile Native units controlled by neutral AI, to represent some local minor Native tribes that aren't affiliated with one of the main Native civs.

On the other hand, if Wild Animals are too powerful people may get frustrated with exploration. I wonder if we can use some already existing unitinfos.xml features like <withdrawalprob> or even <EvasionBuildings> for units like Expert Explorer and Ranger, so if they're defeated by Animals they can retreat and wait a few turns to heal, or possibly even withdraw to the nearest settlement with the Evasion feature. This will make exploration more of a challenge where you'll need to work at it with more than just 1 Explorer, while still making it achievable and not too frustrating.

Hmmm, nicely said... Imho, the whole concept of exploration and finding treasures are way too easy to exploit, and gain huge adventage over AI colonists... Think of it: you enter a native tribe, and you gain a money - in every single tribe, just because. Ok, sometimes is not money, sometimes is an ancient map (that's a good thing in reality, but in game is not very useful) or a unit expirience. Exploring native burial grounds is even more exploitable - if your luck is just average, you'll gather a dozen treasures on standard or large map just like that, gaining cca 10000 or 15000 gold for free, which is completely absurd und unrealistic. And on the top of all that: who is transporting all that treasure, a local indians? Imho, a simple feature could make all this far more challenging: when a scout unit finds a treasure, it should be responsible to transport it to the city. That way it's not only more realistic, but also prevents a player to gain unfair adventage over AI; it forces a player to gather several scout units, and makes the exploration slower. In addition, it makes Malinche and Marquette a much more useful ff.

 

[dmalleus]

The placement of Natives uses a "Bubble Sort Algorithm".
But it does not influence which Nations are chosen.

So if you are playing on a small map where for example many "North American Natives" are chosen,
it might happen, that some of these are pushed to the central of the map by the other "North American Natives".

The bigger the map and the more Natives chosen, the better the algorithm will work.

The algorithm will simply try to place the Natives randomly chosen by the game according to their XML setting for Northern, Central, Southern.
But it still needs to ensure an equal distribuiton of these Natives over the complete map.

All the algorith ensures is:
Northern American Natives further North than Central American Natives and Southern American Natives.
Central American Natives further North than Southern American Natives.

So yes, it is possible that Inuits appear in the central of the map, if they are pushed down there by other Northern Natives.
(Depends on the map structure and which Natives are actually chosen by the game.)

Thanks for such detailed info ray. It is funny to meet an Inuits in the middle of the jungle, but it is certainly not something truly problematic.
P.S. Maybe there is a way to put that specific tribe always on the northest (or southest) indian AI position, via algorithms?

 

[raystuttgart]

Maybe there is a way to put that specific tribe always on the northest (or southest) indian AI position, via algorithms ?

It is possible, but I will personally not spend any more time on "Placement of Natives".

After implementing the "Random Placement of Europeans" and later on "Sorted Placement of Natives"
("Historical Correct ..." is not really a good name for the feature) I am fed up with that.
Both features were not important for me. I basically implemented them, because my former colleagues from TAC wanted them.

I would not mind, if somebody else really wants to spend the time though.

But please let us have discussions like that in the thread Mod Development.

 

[raystuttgart]

The first Assert failed is fixed.

The game was simply complaining about a comment in an XML.
(CivCol is a little picky about the way where you should write comments and how you should do it.)

So it was totally harmless.

Edit:

The second Assert failed is fixed, too.
It was also totally harmless.

The game was simply complaining, that in a switch a special case was not defined.
(Which was not necessary because the default was the right thing to do.)

----------

@frontiersman:

Thanks for reporting.
It really is a good idea to clean up small stuff like that.
(And it is usually really easy to do.)

 

[frontiersman]

Hi Ray,
My pleasure. I'll be on the lookout for any other issues.
And congratulations on getting 1.1 released!