Single Player bugs and crashes v37 plus (SVN) - After the 24th of December 2016

[Thunderbrd]

In 9620 and in some earlier builds that I've tested the game crashes to desktop, somewhere at the end of opponents turn.

I'm sorry but I cannot work with this save. You renamed your mod file and furthermore added a /trunk/ extension. I'm not sure how you can even run the mod this way but since you figured out a means to do so, it may work on your end but it doesn't work on mine.

 

[Thunderbrd]

You really need to look at this savegame T-brd.

I'll take a look. Sounds like some weird stuff taking place. As for the barb flag LE units... very strange indeed. It almost sounds like stock v37 was recently in use on this game because some of the overconcentration of LE units in the capital is the kind of thing it used to do. I'm not sure what can be done here yet.

I have a question for you, if I were to add the iGoldmodifer we have in GameSpeedInfo.xml to the EraInfos.xml would there need to be a schema change?

Yeah, along with a significant amount of programming to make it work. I would've thought Era infos has something somewhat like that already... I'll have to look into it soon.

But more will be needed as the whole Civics categories and city maint costs are also out of whack. Too many civics give too many breaks for unit upkeep and City maint modifiers.

This is very probable. A big way to control balance is here in civics and a lot of that has been less respected than perhaps it should've been in the effort of making ongoing improvements to civic selections as techs go on. I'm not saying they shouldn't improve some but we've gotta be careful not to make things too easy on upkeep. It's tough to design right.

Part of the problem with overflows is the totals, not just the round by round incomes. The totals hoarded will overflow long before incomes will. I know we did a lot to protect against culture levels doing this but there's other matters to address. The most important thing is not so much to try to keep it from happening but to track exactly what happened. I need to know what overflows need to be addressed specifically.

 

[JosEPh_II]

Yeah, along with a significant amount of programming to make it work. I would've thought Era infos has something somewhat like that already... I'll have to look into it soon.

Crap! No EraInfo has nothing for managing gold like IGoldModifier. We can tweak research but not the money.

EDIT: Players gonna hate me when they scrap by on next to nothing in the Preh era so that by the time they get to med Era they can have a few hundred gold/turn. And I got to pinpoint what it is in Ancient era that opens up the money bag. I'm almost certain it's a Civic or 2.

 

[JosEPh_II]

Yeah, along with a significant amount of programming to make it work. I would've thought Era infos has something somewhat like that already... I'll have to look into it soon.

Why would it require "a significant amount of programming". And iGoldModifier is already in the schema too. I plugged it into EraInfos set at <iGoldModifier>100</iGoldmodifer> for each Era and then started a new game. It loaded fine and I've played 11 turns so far.

Next step will be to up the amount to 150 and see what happens>

 

[Thunderbrd]

Why would it require "a significant amount of programming". And iGoldModifier is already in the schema too. I plugged it into EraInfos set at <iGoldModifier>100</iGoldmodifer> for each Era and then started a new game. It loaded fine and I've played 11 turns so far.

Next step will be to up the amount to 150 and see what happens>

If it's already programmed and it wasn't in the schema, that's very irritating, not from you but to think something would be programmed but not be included in the schema... smh.

Just putting a tag in the schema does nothing but allow you to use it in the XML without an error coming up. You can do all that and have it have absolutely no effect because the DLL isn't using it. That's where the rubber really meets the road.


I've been looking at this savegame... wow.

So far my diagnosis is that in the late era, a little crime can go too far. The AI can adapt to the need to control its properties pretty well unless it starts getting out of hand too quickly. At which point, other priorities conflict and the slow ability to react lets it get out of hand too quickly. There may also need to be some reluctance programmed into the AI to build heavy crime creating buildings for those apparent gold benefits. This may infuse too much crime for them to address in an appropriate enough time frame if they suddenly build them all over, which it looks in many cases like they have or are in the process of doing.

Another issue may have been that police stations were obsoleting which was causing trouble for the police precincts and that's another major issue to resolve further than was recently resolved. This was leading to the AI having no access to building effective policing units to counter the era's crime waves.

The HUGE amounts of unhappiness and unhealth are coming from these unchecked crime waves. The strength of crime buildings in this era is tremendous... up to +30 unhappiness in all cities from some crimes. When many cities get this, you can see how it compounds and compounds and compounds. You simply cannot build enough to counter this kind of pain.

Part of what's happening is that later crime and LE units have stronger effects, thus in the later game, getting crime levels up to 6000+ is not all that out of line with a growing imbalance of power with criminals getting the upper hand. When the worst of the worst stuff is happening at 2000, it doesn't take much of that imbalance of power in the later game to trigger every crime in the game in every city they have AND there's SO many of those horrific buildings that cause cumulative national pain.

If as the game goes on we have crime buildings more and more prepared for there to be a wider potential range of crime then we give nations a fighting chance.

We may also have to take another look at the 'operational range' setting because perhaps the AI is giving up on the 'hopeless' situation when they really need to be freaking out seriously about it and doing all they can to address it. I might also need to figure out how to get them to start building more LE units locally when they start having a national problem with crime like this... they're trying to limit themselves to building LE units only where they are going to get the best LE units trained. This elitism works great for war and to keep them from overtraining some units but in this case, it's causing them to react a little too slowly.

For a lot of these nations, the problem got so severe that they have lost access to any decent crime control measures and it's just become a serious runaway forestfire for them where there's so much national unhappiness everywhere that people have starved out... it's really an apocalyptic nightmare and in many cases it even shows that with the crowbar guards that they're trying to train to respond to this. In some cases, only canines are the best selection left.

I'm thinking the barbarian flag units that are showing are actually criminals but its possible that crowbar guards are HN as well? I dunno.

I don't see an overflow here... this is just a multipronged problem with excessively painful crime design and some display for the need for some further AI work. That AI work is much harder than letting the assumed operational range of crime to expand as the game progresses though.

It would've been good to track the progress. Another thing I'm seeing, slavery should've been left long ago by a number of civs. Their hanging onto it too late is a big problem for them. I'm not sure there's effective enough AI for triggering the desire to train a judge who can remove the worldview that becomes poor to continue to maintain, and to use it to do so.

 

[Dancing Hoskuld]

If it's already programmed and it wasn't in the schema, that's very irritating, not from you but to think something would be programmed but not be included in the schema... smh.

The tag bGraphicalOnly is programmed in vanilla BtS for but does not appear in the vanilla schemas. It is referenced in the Python. I just had to add it to the C2C schema files to get the expected results.

 

[Thunderbrd]

Nobody said Firaxis was perfect

I just checked the code... sorry Joe but iGoldModifier was only ever made for GameSpeedInfos. It's in the same schema as the one that controls EraInfos but it's not programmed in the code for EraInfos soooooo you can use it there all you want but it's not going to do anything until it's added in the DLL. That's the core of tag building.

I can do that and it probably would be a lot more beneficial in Era Infos. However, see if you can tackle this through civics and upkeep perhaps. It's going to be a little while before I can get to this. Maybe Alberts can address it faster than I can, hard to say.

 

[strategyonly]

I went back to an earlier game (Ren Era) and everything is ok, so it has to be after that that everything went haywire . . . .

 

[olduser]

I played his game for 2 turns. The cities dropping to 1 are not his. But AI civs that are in seriously bad Unhappiness revolts and have starved down to 1 pop. 1 of those cities Aiken I think, in his screen shot has 624 red faces vs 124 And Crime is over +1600. Another had +2300 Crime and No LE units at all but was instead filled with 20+healer type units.

Can anyone check if LE units lost their promotions or can't get any promotions or anything like that. I made some changes in that area and even knowing that i really tested them doesn't mean I didn't miss something. The same would go for unit combats.

 

[JosEPh_II]

I just checked the code... sorry Joe but iGoldModifier was only ever made for GameSpeedInfos. It's in the same schema as the one that controls EraInfos but it's not programmed in the code for EraInfos soooooo you can use it there all you want but it's not going to do anything until it's added in the DLL. That's the core of tag building.

Drats! It would've given such a fine control on Gold balancing.

Sooo.......The Individual Crimes and the Crime Diffusion property needs attention. As well as some of the other Properties that give crime when they go south of 0. Tourism has to be a big problem in this area with it's 10 levels of crime adding. They are too strong imho.

But like I warned when I 1st started adjusting crime that if you let it get too high Hydro had some "killer" penalties. The AI has to keep it's crime below 800 as does the player. You let it hit even the 500-600 range and it ramps up fast. And by the time it hits 800, doomsville. You better stop everthing else you are doing and start fighting the Crime. There is no way, with the way Crime and the Individual Crimes and not even taking into account the Criminal elements, that the AI should Not be in full crime stopping Mode when Crime hits 1000 level.

I can space Individual Crimes out to 2000 or even 3000 Crime level. It would give a longer time frame to work on battling Crime for all Civs. And/Or the Anti-Crime early buildings need a Big Buff. This can be worked on in a 2 pronged attack.

And this confirms what I said long ago that No Crime should start at 0 or 1 or even 10, same for Disease. Bad Idea to put these elements at this low an entry level early in the game.

I think the 1st thing I will do is lower the Diffusion rate (set it below 10%). Then adjust the Ind. Crimes entrance levels, followed up with going over some of the Crime giving buildings and anti-crime Bldgs, starting in the early eras. This will slow the pace and the already reported rollercoaster crime level ride that many players have complained about for sometime now. But the mid to late game crime giving buildings can't be overlooked either especially any that give crime in % values. This skyrockets the level when they come into play later in the game. Finally all the crime giving and fighting buildings in these new Eras. Lots of digging to do.

Leaders that give large Crime values will also need those reduced.

I also have a preliminary set of base values reset for the GS to commit. It will Only be the 1st of many steps to regain a semblance of balance overall back.

 

[Shoobs]

Why would it require "a significant amount of programming". And iGoldModifier is already in the schema too. I plugged it into EraInfos set at <iGoldModifier>100</iGoldmodifer> for each Era and then started a new game. It loaded fine and I've played 11 turns so far.

Next step will be to up the amount to 150 and see what happens>

Access to any jewelry or precious metal tends to really -really- skyrocket gold income at metal casting. My current game on Immortal was having issues until I hit metal casting. Early game gave me no food resources... but when I hit scavenging, I saw I had Copper, Silver and then Gold somehow came into the picture. I also had marble close but not within the holy cross. Now there's no problems ( even though flamability is HIGH ). My nation doesn't really have much food bonuses except for the sea, but it doesn't need it. I have gotten all my population from captured civilians ( wiping out the nearby rival and multiple barbarian cities. It's also the one where I have a level 30+ Wanderer ).

 

[JosEPh_II]

@T-brd,

Do we really want adjacent tiles diffusing crime to other adj. tiles? (I bet you say yes)

<PropertyPropagator>
                     <PropertyPropagatorType>PROPERTYPROPAGATOR_DIFFUSE</PropertyPropagatorType>
                     <PropertyType>PROPERTY_CRIME</PropertyType>
                     <GameObjectType>GAMEOBJECT_PLOT</GameObjectType>
                     <TargetObjectType>GAMEOBJECT_PLOT</TargetObjectType>
                     <TargetRelationType>RELATION_NEAR</TargetRelationType>
                     <iTargetDistance>1</iTargetDistance>
                     <iPercent>4</iPercent>
                     <Active>
                       <Is>TAG_OWNED</Is>
                     </Active>
               </PropertyPropagator>

We have Main city tile diffusing to adjacent 8 tiles, and then they in turn diffuse a % back to the Main. But this section allows diffusing from the initial 8 to the 7 tiles adj to it . So we have a process of adj to adj which is governed by If the tile is owned TAG_OWNED. But this is a multiplicative diffusion as 1 adj tile diffuses to Any tile that meets the <iTargetDistance>1</iTargetDistance> and >GAMEOBJECT_PLOT< conditions. Now multiply by it by up to a max of 8 adj tiles(by the time your city has it's Fat X shape and can be less if Realistic Culture Spread is On). This ripple effect out and then back becaome much more potent as the Crime level increases.

I can see why you want this but 2 things need attention in this area: 1. the Decay rate for these is only set to 4%. At a Crime level of 100 it's not really noticeable but by the time Crime level hits 300-400 it begins to be noticed, and by level 1000 the Decay is overwhelmed easily and the "sudden" Rises in Crime take on significant importance. And 2. the diffusion rate of the initial 8 back to the city is at 12% currently.

So I will be doing 3 things here:
1. the Decay rate for Adj tiles >GAMEOBJECT_PLOT< I will increase to 6% from it's current 4%
2. the Diffusion rate Back to Main city tile from adj 8 will be decreased from 12% to 10%
3. the Main city tile Decay will be raised from 4% to 5%.

This will give the AI and player a bit more time to react to rising Crime.

After this adjustment then it will be necessary to chart When each Individual Crime appears in the Game and at what level to see if there is further need for entry level adjustments. I believe there will be such a need.

Our current Range to work with is 0 to 2000. The highest entry level iirc was 1500. With Sectarian Violence at 1350 level. This is one of those "killer" Individual crimes. Where "killer" means it was a BIG unhappiness hit.

If you have no problem with my suggested changes I well commit them once you say Okay.

EDIT: I did more math on these new % I proposed and they will help. If the diffusion turns out to still be too high the Decay (both places) is the place I suggest would get further change.

Now I need to fire up SO's save game again with these new values and see if any change is noticeable. I don't think Properties get locked in at game start like other portions of the Mod does.

I'm going to start my commits soon.

 

[Thunderbrd]

I don't feel diffusion or decay really has a lot to do with the overwhelming amounts of crime these cities are experiencing. It's not a sudden or unexpected surge they're getting slammed by so much as the fact that as the game progresses, buildings and criminals bring higher and higher crime values. Although this is compensated by higher and higher crime control values on LE units and buildings to match, when the balance tips, it tips harder the later you are in the game.

Thus, a range of penalties up to 1200 may not seem like much if you're basing that on an understanding of normal crime ebb and flow up through the Classical Era, when you consider that crime can swing wildly into the 4-6000 range with ease in the Modern Era, and by that time ALL the crimes are operating and some of them coming with not just local but tremendous national penalties, even one city going out of whack on crime for a bit, which would be totally normal under numerous causes, could then cause the sudden collapse of another city because suddenly all its citizens refuse to work (including specialists helping to counter crime) then that causes the collapse of every other city in the same way and now the ability to react is pretty much gone too because there's very little production to work with and much of that is immediately dedicated to trying to deal with unhappiness that's past the point of hope. Before you know it, local criminals are spawning everywhere and it all goes to hell.

When this happens, diffusion is a very very small factor in this picture. It CAN be a trigger to start this process by creating a surge but I haven't seen huge surges from that like we were seeing with trade. It's important to have some tile diffusion to allow the Ruffian units to be capable of spreading crime to cities without having to be IN the cities.

If, as the game proceeds, we consider the operational range of crime swings to be wider and wider, we can limit how prepared the crime system is to cause this much damage.


@alberts2 :
I don't think they're having any trouble with promotions or unitcombats. They seem to have some great promotion selections at least... it's just not enough to compensate.

The game could use some under the hood AI response analysis for when things get out of control this fast and violently. Perhaps we may need to insert some emergency LOCAL responses rather than relying entirely on the contract broker - many cities should be training LE units to respond but it's only happening from the primary training center and that's too slow for this much need. While you're in the code, perhaps you could add a little something for that for us.

I think what created the problem for this game was some messed up building prerequisite stuff that was recently noted and solved - just too late to save these AI nations from the fires that problem initiated.


We haven't spent a lot of time with players reaching this late in the game. Some imbalances and insufficient AI responses are to be expected. @strategyonly: good job on reaching this far in the first place!

 

[JosEPh_II]

I don't feel diffusion or decay really has a lot to do with the overwhelming amounts of crime these cities are experiencing. It's not a sudden or unexpected surge they're getting slammed by so much as the fact that as the game progresses, buildings and criminals bring higher and higher crime values. Although this is compensated by higher and higher crime control values on LE units and buildings to match, when the balance tips, it tips harder the later you are in the game.

Fully understood but these small changes will slow the early advance allowing the AI and lower level players a chance to cope longer. They are not huge changes, but subtle changes.

4-6000 range with ease in the Modern Era, and by that time ALL the crimes are operating and some of them coming with not just local but tremendous national penalties, even one city going out of whack on crime for a bit, which would be totally normal under numerous causes, could then cause the sudden collapse of another city because suddenly all its citizens refuse to work (including specialists helping to counter crime) then that causes the collapse of every other city in the same way and now the ability to react is pretty much gone too because there's very little production to work with and much of that is immediately dedicated to trying to deal with unhappiness that's past the point of hope. Before you know it, local criminals are spawning everywhere and it all goes to hell.

Which why I think your Crime buildings you added need Higher crime level entry points. I have looked at them in the CIV4PropertyInfos and feel all should be raised by 100 - 300 points:

Proposed Example just for starters:

    <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_BANDITS_HIDEOUT1</BuildingType>
                       <iMinValue>300</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_BANDITS_HIDEOUT2</BuildingType>
                       <iMinValue>500</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_BANDITS_HIDEOUT3</BuildingType>
                       <iMinValue>800</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_ASSASSINS_DEN</BuildingType>
                       <iMinValue>600</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_LOANSHARK</BuildingType>
                       <iMinValue>500</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_CONSPIRATORS_HIDEOUT</BuildingType>
                       <iMinValue>600</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_MOB_STOREFRONT</BuildingType>
                       <iMinValue>900</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_MOTORCYCLE_CLUB</BuildingType>
                       <iMinValue>800</iMinValue>
                       <iMaxValue>100000</iMaxValue>
                   </PropertyBuilding>
               <PropertyBuilding>
                       <BuildingType>BUILDING_UNSANCTIONED_BLACK_HAT_HIDEOUT</BuildingType>
                       <iMinValue>1100</iMinValue>
                       <iMaxValue>100000</iMaxValue>
[code]

 

[Thunderbrd]

Go for it. You can go for even higher for the later era stuff if you like. Up to 1500 as a trigger point for criminal spawning in the latest eras seems appropriate.

 

[JosEPh_II]

Go for it. You can go for even higher for the later era stuff if you like. Up to 1500 as a trigger point for criminal spawning in the latest eras seems appropriate.

Went thru the Individual Crimes and lowered many that had Health - 10, -20, -40 to lower levels. generally cut in half. But not all.
Did same with High Unhappiness and if they had Globalhappiness -3 or above lowered most by at least 1 and some by half.
Also adjusted the ones that had very high Crime Property like +20, +25, +50 down to +10, +15, +20.

You might want to look at your Insidious levels as I would have no clue what to base them on.

I'm leaning more now to adjusting the Range from 2000 to 3000 and some of the later Ind Crimes that have high penalties making their entry level much higher. Like Time Tampering and Asimov (Robot) Tampering as 2 examples. Sectarian Violence, Gang Warfare, and Rioting need pushed higher too. I want to do some more checking on SO game with these changes to see if it drops the citizens and even a little. If it does then games in progress should have a decent response.

 

[Thunderbrd]

That sounds all like a set of very appropriate responses, for now, then we'll see how it goes again. SO said he has a much earlier save so I wonder if he'll play forward from there and see if he runs across the same problem once he gets these updates.

I wouldn't worry about insidiousness too much. Just by adjusting the trigger points of the crimes they come in on will help a lot. One thing I noticed - I've gotta make it so that if a unit doesn't have any innate insidiousness, building sources won't add to the insidiousness for that unit. I thought I'd already fixed that a while back...

I wouldn't expect that there's any amount of response that can save this particular game though. It's kinda like real world global warming... once its too late, just pulling back on emissions isn't going to solve the issue, it's going to require forces that actually reverse the process and heal the damage once the fit hits the shan in full glory.

 

[strategyonly]

SO said he has a much earlier save so I wonder if he'll play forward from there and see if he runs across the same problem once he gets these updates.

Sure just let me know when to proceed and i will give it a go . . . but it will take me awhile to get back to the Inform Era . .

 

[JosEPh_II]

Sure just let me know when to proceed and i will give it a go . . . but it will take me awhile to get back to the Inform Era . .

I made 2 commits already SO. It should give you a decent place to start.

How have you been feeling of late? Better I hope.

 

[strategyonly]

I made 2 commits already SO. It should give you a decent place to start.

How have you been feeling of late?

I made 2 commits already SO. It should give you a decent place to start.

But is that from the now game or the earlier game i mentioned??

OFF TOPIC: Well a few weeks back i was admitted to the Emergency Room with COPD that Exacerbated (they said it was because my brother died a few weeks before) then to top that off i had Phenomena with it . . so they said lucky i came in, i only had like 2-3 days left if i didnt see them then, but better now, I guess / / / /??