Caveman 2 Cosmos

@alberts2: I'm having some trouble with Settler evaluations. The AI seems to be running up against an inability to see obvious good settler sites. During bool CvCityAI::AI_choosePropertyControlUnit(int iTriggerPercentOfPropertyOpRange, PropertyTypes pProperty), this function:
int CvPlayerAI::AI_getNumAreaCitySites(int iAreaID, int& iBestValue) const
suspiciously comes up with 0 in cases where I'm pretty sure that's not correct. The caching is making it a little difficult to determine for sure. I'm looking deeper but do you have any immediate theories as to why perfectly good city sites aren't being determined here?
 
@alberts2: I'm having some trouble with Settler evaluations. The AI seems to be running up against an inability to see obvious good settler sites. During bool CvCityAI::AI_choosePropertyControlUnit(int iTriggerPercentOfPropertyOpRange, PropertyTypes pProperty), this function:
int CvPlayerAI::AI_getNumAreaCitySites(int iAreaID, int& iBestValue) const
suspiciously comes up with 0 in cases where I'm pretty sure that's not correct. The caching is making it a little difficult to determine for sure. I'm looking deeper but do you have any immediate theories as to why perfectly good city sites aren't being determined here?

No obvious issue comes to mind but the code to calculate the found value is complex so there could be some hidden issues. The found values should get logged in the BBAI.log every turn maybe that log can be helpful.
 
No obvious issue comes to mind but the code to calculate the found value is complex so there could be some hidden issues. The found values should get logged in the BBAI.log every turn maybe that log can be helpful.
I'm suspicious of the caching system. Just testing some runthroughs of the function that gets the found values shows that there are plots coming up with value, which means there shouldn't be a failure to find some additional city sites. I don't know... maybe I'm jumping at nothing here. More testing is needed in more regions of the code. But it seems like the city sites determiner is not clearing old data properly. Then again, nothing really should have changed here recently and just a bit ago, they were growing like weeds. So I may just not be understanding something critical still.
 
I'm suspicious of the caching system. Just testing some runthroughs of the function that gets the found values shows that there are plots coming up with value, which means there shouldn't be a failure to find some additional city sites. I don't know... maybe I'm jumping at nothing here. More testing is needed in more regions of the code. But it seems like the city sites determiner is not clearing old data properly. Then again, nothing really should have changed here recently and just a bit ago, they were growing like weeds. So I may just not be understanding something critical still.
Posing a question; If you are suspicious of the caching you then cleared the cache to see if there was a difference? Especially so with the bug in missions recently discovered and fixed.

Quite frankly with every "coding" commit for sometime now I flush the cache folder (delete contents). Maybe it's an over reaction, then again maybe it's actually prudent. I'll keep doing it to lean towards the prudence of the situation .
 
Posing a question; If you are suspicious of the caching you then cleared the cache to see if there was a difference? Especially so with the bug in missions recently discovered and fixed.

Quite frankly with every "coding" commit for sometime now I flush the cache folder (delete contents). Maybe it's an over reaction, then again maybe it's actually prudent. I'll keep doing it to lean towards the prudence of the situation .
Not the kind of caching I'm suspicious of. Caching is not always a matter of saving data on files. In this case it's a matter of saving data in ram to keep from having to continuously reprocess the determination of where the best next cities would be. It appears to just be holding on to the answer... none. When there are actually quite a few. To derive the answer 'none' must have taken place a long long time before this point. But then how settlers ever work at all if this is happening is as much the question.
 
Bugs continue to be found. Do you want us to serve you more bugs?

My text holds clear wishes and hopes for their destruction, not the opposite which you certainly must know but choose to provoke instead. Quallity and it's control has been so bad in this product recently that you really should project your apparent frustration towards that problem, not the other people.
 
Re-release was scheduled for this weekend. And these discussions are an aside to it's work being done. As for your opinion over the use or not of the SVN, we will still give the better recommendation, which is use the SVN.

Great for you and the mod itself if it's comnig but I play a better version now and am not the one in a hurry here. Once again, your problem is not me and the svn. It is all those new players that only know the broken main downloadable file as an option. None of your real problems will be solved if I start to use the svn.

By all means, you don't have to reply to messages if they annoy you and consume so much of your time. I thought you quit already from the team already anyway?
 
Great for you and the mod itself if it's comnig but I play a better version now and am not the one in a hurry here. Once again, your problem is not me and the svn. It is all those new players that only know the broken main downloadable file as an option. None of your real problems will be solved if I start to use the svn.

By all means, you don't have to reply to messages if they annoy you and consume so much of your time. I thought you quit already from the team already anyway?

Why are you getting snippy with me? I find that is uncalled for. Nothing in my post you quoted is derogatory to anyone, especially to you.

I was forced out of balancing the mod. My way was deemed not good enough. I have not Officially resigned in any post Public or Private. My role is diminished that is true.

Look, I just lost almost 2 years of hard work. Thrown out for a new way. Am I happy and all giddy about it? Of course not! Especially when the work was starting to get a good reception from the users, you the players. But I am not in charge of this mod. And a leadership decision was made. Now we are on a new course, and unfortunately a bad release was put out.

We understand the users frustrations. The Whole team is as frustrated as well. But constant badgering and demands are Not helping the situation either. Understand this, there are only 2 modders on this team that can devote large chunks of time to it, because we are retired. All the rest have Real Life Jobs, Families, and other responsibilities. And I have No C++ or Python skills or training. So all I can do is test and tweak the xml.

I have replied to you as a courtesy. Please if you need to respond do so in kind.
 
My text holds clear wishes and hopes for their destruction, not the opposite which you certainly must know but choose to provoke instead. Quallity and it's control has been so bad in this product recently that you really should project your apparent frustration towards that problem, not the other people.
I'm working on the bugs as quickly as is possible. There is also, due to being a full time employee, a bit more time to discuss some things than there is to actually work on them. Some discussions take place on the side - that doesn't mean we aren't dedicating time to the problem. I might be missing some of what you're trying to say here but I hope you aren't getting the impression that fixing every problem we find isn't the highest priority here. I thought for a moment we were near a rerelease and we might actually be pretty close.

I'm going to have some things coming at me here for the next month in RL that are going to severely hamper my time to work on the mod. So I'm hoping we can pretty much find that the AI is playing 'sufficiently well enough' for release after just a few tweaks here because if it isn't, it'll be all on @alberts2 to find out why for a month or so from here as I won't have time for it. I do think we need to get at least what we have out there because it seems to be getting some decent game feedback now. Still have a few things I want to look at but I'm not finding that many direct bugs in the code as much as just stuff that we might want to tweak, like spawn rates and such.
 
I've started streaming.

Link to your stream? I don't recognize your name. Only streamers I know are on Twitch (IdioticUlt1mara, occasionally NinjaZombie0, and MajinBuu who posted a 19 hour long C2C stream yesterday). IdioticUlt1mara and NinjaZombie0 post copies to youtube.
 
Do animals have AI for avoiding hunters? I often see situations where a scout or other non-hunter finds large groups of animals, but when I move a nearby hunter in they're gone. There are no predators or other units that could have killed them, and random wandering couldn't have dispersed them that fast.
 
Do animals have AI for avoiding hunters? I often see situations where a scout or other non-hunter finds large groups of animals, but when I move a nearby hunter in they're gone. There are no predators or other units that could have killed them, and random wandering couldn't have dispersed them that fast.
Animals de-spawn as well as spawn. If understand the coding for them, and the explanation given from Dancing Hoskuld (the author of the subdued animals modmod) they have a limited amount of time they will stay on the map.
 
Animals de-spawn as well as spawn. If understand the coding for them, and the explanation given from Dancing Hoskuld (the author of the subdued animals modmod) they have a limited amount of time they will stay on the map.
I don't think animals de-spawn.... I've never noticed it happening, and I've never seen any coding for it either.

@amjh I think defensive only animals do have an AI that makes them avoid dangers. At least that is my experience, that non-attack capable animal flee from obvious dangers.
A wounded animal will if the attacking unit has good invisibility stop to heal on the plot it withdrew to, but if the attacking unit is not invisible to the wounded animal it will flee further away before healing up.
 
Last edited:
I don't think animals de-spawn.... I've never noticed it happening, and I've never seen any coding for it either.
Just repeating what DH told me. So if he has changed that I'm not aware of it.
 
Maybe I've just been having bad luck, but it feels like the animals are a bit too good at avoiding hunters in some situations as huge groups just disappear when one approaches.
 
Maybe I've just been having bad luck, but it feels like the animals are a bit too good at avoiding hunters in some situations as huge groups just disappear when one approaches.
Do you have Peace Among NPCs On or Off. If Off the animals will prey upon each other as will the Barbs and Neanders. This can cause the situation you are describing.

Also are you manually hunting or are you using the AutoHunt feature from the BUG options? AutoHunt is less micro management but it is also not that good in doing what it's intended purpose is. Manual while more involving is the better way to hunt.
 
Do you have Peace Among NPCs On or Off. If Off the animals will prey upon each other as will the Barbs and Neanders. This can cause the situation you are describing.

Also are you manually hunting or are you using the AutoHunt feature from the BUG options? AutoHunt is less micro management but it is also not that good in doing what it's intended purpose is. Manual while more involving is the better way to hunt.
Auto hunt is not as bad as it used to be.
 
Back
Top Bottom