Single Player bugs and crashes v38 plus (SVN) - After the 20th of February 2018

[Thunderbrd]

T-brd did alberts2 not say he had disabled or removed multi-threading?
EDIT:


Different Multi-threading? Or related to all Multi-threading that koshling added to the Mod? Just trying to understand.

Yes, different multi-threading. The multi-threading that's causing me some issues is in the city build queue multi-threading. But I don't think it's the problem, specifically, just making it hard for me to read through the next steps to find the problem. I've gotta figure out how the queue picks up the units called by the brokerage and actually puts them in the build queue. I THINK I've figured that out before but the thing is that it does it after the call for the unit is logged. Either the storage for logging the call is somehow broken, which it may be but I don't know how that would've happened and it didn't LOOK to be broken, though that might just be my vague understanding of how that information storage is supposed to work exaclty, or the spot where the queue picks up unit training is broken.

 

[Dancing Hoskuld]

Yeah a long time ago I argued that where they come into the game in relation to trackers was not good. And why not just remove them Or make them the unit you get instead of Wanderers at Nomadic Lifestyle. Nothing became of it. Imho the mod could lose them and no one would care.

Wanderers are unlocked on Nomadic Lifestyle (X=1).
They can be upgraded to Scouts on Tracking (X=5) or Chasers on Persistence Hunting (X=4).
Chasers can be upgraded to Trackers on Tracking (X=5), without unlimited national units option you can have only 5 trackers at once.
If you go over limit, then you can train Chasers only.
Trackers upgrade to Hunter on Hunting Tactics (X = 13), column before Tribalism. You can have max 5 hunters without unlimited national units option.

Basically Chasers are buffer for Trackers and Hunters, that can be used as pest control within your borders.
You can have 5 Trackers or any other hunting units, but Chasers are unlimited.

But if you have master hunting units, cultural hunting units and WW Neanderthal hunting units running, then chasers are useless.

Originally we only had Trackers. Wanderers and Chasers were added in to back fill to the start of the game. Unfortunately animals have only been balanced for Trackers. It was never decided how we should balance them for the other hunters. Should the animals be weaker at the start and stronger over time? Is it the individual animals that become stronger or do we make them groups of animals? Should only weak animals spawn at the start? Many questions.

If you are playing the game with the normal limited number of hunter units then you should not be able to build earlier ones when you have the full complement of the currently available ones. The tag that forces obsolete should be on used to stop this happening.

 

[Toffer90]

Yeah a long time ago I argued that where they come into the game in relation to trackers was not good. And why not just remove them Or make them the unit you get instead of Wanderers at Nomadic Lifestyle. Nothing became of it. Imho the mod could lose them and no one would care.

I made the same statement about it once too.
In the "My take on stuff" modmod I changed it up quite a bit so that the Tracking tech is quite a bit later in the tech tree and persistence hunting is a few steps earlier in the tech tree.
I gave chaser a 100% combat bonus against fleeing animals, those that has the unitCombatType UNITCOMBAT_ANIMAL_FLEEING, so that Chasers have a good chance against (in addition to the weakest animals) a certain group of animals that usually have too much strength for them to handle. I find I use the chasers quite a bit, even on the faster gamespeeds, in my modmod.

I guess I'm saying that I think chasers have a place in C2C if we want them to, but some stuff needs to be changed for it to be.

 

[Dancing Hoskuld]

In the "My take on stuff" modmod I changed it up quite a bit so that the Tracking tech is quite a bit later in the tech tree and persistence hunting is a few steps earlier in the tech tree.

The problem is that the tracking tech is about making trails, trail markers and notes to others about the lay of the land, where to find water and game. This happened quite early in human development most likely before fire and thus before homo sapiens sapiens.

 

[T C]

Perahps the solution is to move the tracker unit to another tech a little later on, maybe Hunting?

 

[Noriad2]

Chasers become obsolete partially because the animals become too strong. And I don't recommend slowing down animal spawns as it is already odd that you start in a world without animals as animals exist much longer than humans.

 

[Toffer90]

The problem is that the tracking tech is about making trails, trail markers and notes to others about the lay of the land, where to find water and game.

The tracking tech is not about making trails in my modmod. ^^

 

[Thunderbrd]

Trackers are probably a bit too early. But that said, a faster gamespeed beckons the skipping of the use of the chaser. Even on Epic I tend to try to avoid their use.

 

[tmv]

Originally we only had Trackers. Wanderers and Chasers were added in to back fill to the start of the game. Unfortunately animals have only been balanced for Trackers. It was never decided how we should balance them for the other hunters. Should the animals be weaker at the start and stronger over time? Is it the individual animals that become stronger or do we make them groups of animals? Should only weak animals spawn at the start? Many questions.

If you are playing the game with the normal limited number of hunter units then you should not be able to build earlier ones when you have the full complement of the currently available ones. The tag that forces obsolete should be on used to stop this happening.

Trackers (and their replacements) used to require the Master Hunter building, which was very difficult to obtain without dedicated hunter units. So there was a very good reason for Chasers to exist.

Should animals really be worthwile enemies in Ancient+? Animals ambushing single people is one thing, but fighting military units in - more or less - open warfare? And the reward goes down in the meantime, with myths and stories obsoleting and the reward for studying the animal becoming less and less with respect to tech costs.

 

[KaTiON_PT]

Should animals really be worthwile enemies in Ancient+? Animals ambushing single people is one thing, but fighting military units in - more or less - open warfare?

Let's not forget of the Great Emu War.

Though animals do have a place in later eras for their bonuses to labs, circuses and zoos.

 

[tmv]

Let's not forget of the Great Emu War.

Yes, that's a counter-example. Only one country has wildlife so dangerous that they send the military against them - and lose.

 

[Toffer90]

@Toffer90
There are bugs in python,
Traceback (most recent call last):

File "C2C_Terra", line 115, in normalizeAddExtras

NameError: global name 'balancer' is not defined
ERR: Python function normalizeAddExtras failed, module C2C_Terra

Should be fixed now.

 

[olduser]

Yes, different multi-threading. The multi-threading that's causing me some issues is in the city build queue multi-threading. But I don't think it's the problem, specifically, just making it hard for me to read through the next steps to find the problem. I've gotta figure out how the queue picks up the units called by the brokerage and actually puts them in the build queue. I THINK I've figured that out before but the thing is that it does it after the call for the unit is logged. Either the storage for logging the call is somehow broken, which it may be but I don't know how that would've happened and it didn't LOOK to be broken, though that might just be my vague understanding of how that information storage is supposed to work exaclty, or the spot where the queue picks up unit training is broken.

In case of the hunter escort it might not be broken that much but instead the AI has lots of other things to build first.
This has somehow changed at some point because the AI has to build more buildings at the start of the game now as they did then I set this up and the priority of the hunters request for an escort is now too low.

At the moment it is HIGH_PRIORITY_ESCORT_PRIORITY and it seems this has to be moved up to HIGHEST_PRIORITY_ESCORT_PRIORITY, this should accelerate the training of that escort.

EDIT:
It seems that the AI builds lots of Food giving buildings at the start of the game has anything changed here in terms of balance???

 

[Thunderbrd]

In case of the hunter escort it might not be broken that much but instead the AI has lots of other things to build first.
This has somehow changed at some point because the AI has to build more buildings at the start of the game now as they did then I set this up and the priority of the hunters request for an escort is now too low.

At the moment it is HIGH_PRIORITY_ESCORT_PRIORITY and it seems this has to be moved up to HIGHEST_PRIORITY_ESCORT_PRIORITY, this should accelerate the training of that escort.

I was wondering if this was along the lines of the problem. If a unit is evaluated to be trained and a broker call is made, if that call is made from a city evaluation during chooseProduction (or whatever its called), shouldn't the priority of that be set higher automatically than any building that's later evaluated as a construction selection? Shouldn't the priority be established by the timing of when that determination was made in that list? It doesn't seem connected, which seems like a major flaw in itself. The first hunter won't train because a bunch of buildings discovered as items to be queued long after the hunter was determined should be trained are getting in the way. IMO, we need to do something to enforce that as soon as an item, be it a building or a unit, is determined to be placed in the build queue, that it BE the NEXT thing to be queued, not waiting until another evaluation later down the line can take the next queued position.

 

[olduser]

I was wondering if this was along the lines of the problem. If a unit is evaluated to be trained and a broker call is made, if that call is made from a city evaluation during chooseProduction (or whatever its called), shouldn't the priority of that be set higher automatically than any building that's later evaluated as a construction selection? Shouldn't the priority be established by the timing of when that determination was made in that list? It doesn't seem connected, which seems like a major flaw in itself. The first hunter won't train because a bunch of buildings discovered as items to be queued long after the hunter was determined should be trained are getting in the way. IMO, we need to do something to enforce that as soon as an item, be it a building or a unit, is determined to be placed in the build queue, that it BE the NEXT thing to be queued, not waiting until another evaluation later down the line can take the next queued position.

If the AI decides to build a hunter it builds a hunter but if it has other things to build it doesn't even reach the point where it checks if it needs an hunter and builds something else instead. Something must have changed here in terms of balance just start a autoplay for 50 turns and check which buildings the AI builds.

 

[Thunderbrd]

If the AI decides to build a hunter it builds a hunter but if it has other things to build it doesn't even reach the point where it checks if it needs an hunter and builds something else instead. Something must have changed here in terms of balance just start a autoplay for 50 turns and check which buildings the AI builds.

You are instructing me to do what I have done to determine what I just stated. Seeing hardly any units being trained during the prehistoric era at all, I then watched under the attached debug and put a stop at the first hunter evaluation in chooseproduction and watched it go through all the steps to issue a broker call for a hunter only to find a later building selection take precedence over that hunter, which, based on the previous observations of building selections, had been overriding the unit training for some time before I stopped to watch it happen.

 

[olduser]

You are instructing me to do what I have done to determine what I just stated. Seeing hardly any units being trained during the prehistoric era at all, I then watched under the attached debug and put a stop at the first hunter evaluation in chooseproduction and watched it go through all the steps to issue a broker call for a hunter only to find a later building selection take precedence over that hunter, which, based on the previous observations of building selections, had been overriding the unit training for some time before I stopped to watch it happen.

That code is just
If that happens the building is more important as the hunter in that moment for that city.

 

[raxo2222]

That code is just
If that happens the building is more important as the hunter in that moment for that city.

I changed tech cost to 350 and production cost to 70.
I test Normal/Standard/Noble with AI autorun.
Some AIs are wasting time building lesser research/gold....
They come back to building buildings though.

 

[olduser]

I changed tech cost to 350 and production cost to 70.
I test Normal/Standard/Noble.
Some AIs are wasting time building lesser research/gold....

The production code was mostly setup a long time ago and it wasn't programmed to handle the current balance setup. If radical balance changes happen it might be necessary to adjust that code.

 

[raxo2222]

The production code was mostly setup a long time ago and it wasn't programmed to handle the current balance setup. If radical balance changes happen it might be necessary to adjust that code.

They also are too lazy to build units.
I'll change units global to 1 to see what happens.