v20 bugs and crashes

This happens if you start on the part of the world that equates to the Americas south of Mexico. We just don't have enough of the animals from there.

On my GEM map the entire south of Amercia east of the Andes (which is isolated by mountains from any civ start point) had an animal on EVERY tile by the time I got there from north America with canoes.
 
This happens if you start on the part of the world that equates to the Americas south of Mexico. We just don't have enough of the animals from there.

But i opened up the WB and looked at the "whole" map and nothing on it besides Eagles and Neanderthals?? Now that shouldn't happen on a NEW map just formed?? What am i missing here??:crazyeye:

So that means then that PerfectWorld2f is not a good mapscript then? It used to be?

EDIT: OK i tried two different maps, one a mapscript, Mongoose one, and GEM, and it still has NO animals, attached pic is from South America, bottom portion, so that isn't the correctly theory, do you think it might have been the dll portion that Koshling has changed that might be affecting this by any chance??

Something has changed in the last 10 days for this to happen, never did before????? Its NOT even fun to play withOUT them:mad: Shows you how much they are a PART of C2C, IMHO.
 
This happens if you start on the part of the world that equates to the Americas south of Mexico. We just don't have enough of the animals from there.

What animals do you have on that list? I would be interested in seeing a list of what animals you have where. I know all about the animals I converted for you.
 
But i opened up the WB and looked at the "whole" map and nothing on it besides Eagles and Neanderthals?? Now that shouldn't happen on a NEW map just formed?? What am i missing here??:crazyeye:

So that means then that PerfectWorld2f is not a good mapscript then? It used to be?

EDIT: OK i tried two different maps, one a mapscript, Mongoose one, and GEM, and it still has NO animals, attached pic is from South America, bottom portion, so that isn't the correctly theory, do you think it might have been the dll portion that Koshling has changed that might be affecting this by any chance??

Something has changed in the last 10 days for this to happen, never did before????? Its NOT even fun to play withOUT them:mad: Shows you how much they are a PART of C2C, IMHO.
What is your animal spawn modifier set to?
 
I can confirm that the holes in the fog of war at sea are caused by battles where the ship has extended range of sight. I only let one ship have that promo. Unfortunately getting a save just before is proving difficult. I keep getting distracted.
 
I can confirm that the holes in the fog of war at sea are caused by battles where the ship has extended range of sight. I only let one ship have that promo. Unfortunately getting a save just before is proving difficult. I keep getting distracted.

If your using the SVN then it automatically saves in the Max Comp style already, i did that just for that reason.;)
 
OK here is another slave moving into enemy paths 1 complete turn before it happens.


EDIT: OK here is a different one, now even the settlers of the enemy are doing the same thing????? 1390

EDIT EDIT: OK now a non-attacking unit (TownWatchman) walking right up to a city and just standing there, with the enemy with a 10 str as an opponent? 1490
 
In the attached savegame, when giving the explore order to the selected Caravel(s) (if not selected, go to Delhi, there are 2 non moved Caravels) the game stops for about 10 secs. Not serious, but annoying when sending them mass exploring, I haven't checked yet if this impedes turn times.
Revision 1759 (may be already fixed, IDK).

View attachment 313907


Edit: Seems the effect on ones own turn is significant, till all Caravels are moved the game is practically unresponsive.

Anyone took a look at this ? On maps with large bodies of oceans any sea automation is unusable because it increases turn times considerably and makes diplo/map/city screens choppy and slow until all automated units have made their move.
 
OK here is another slave moving into enemy paths 1 complete turn before it happens.


EDIT: OK here is a different one, now even the settlers of the enemy are doing the same thing????? 1390

EDIT EDIT: OK now a non-attacking unit (TownWatchman) walking right up to a city and just standing there, with the enemy with a 10 str as an opponent? 1490

Unlikely to have time to look at this before Monday. However, recent chnages applied only to workers (and settlers for some action), which use different routines in the AI. Nothing should have changed the behaviour of other units in this regard recently. So the town watchman issue (and possibly the slave issue - I'm not too sure about slaves off hand) are not related to the previous worker issues. The settler case might be. The town watchman copuld also conceivably be related tio the attack stack behaviour chnage - it might be rendevousing with an attack stack that got cold feet or something (should be possible to check, though I might need a save from a couple of turns earlier)
 
Unlikely to have time to look at this before Monday. However, recent chnages applied only to workers (and settlers for some action), which use different routines in the AI. Nothing should have changed the behaviour of other units in this regard recently. So the town watchman issue (and possibly the slave issue - I'm not too sure about slaves off hand) are not related to the previous worker issues. The settler case might be. The town watchman copuld also conceivably be related tio the attack stack behaviour chnage - it might be rendevousing with an attack stack that got cold feet or something (should be possible to check, though I might need a save from a couple of turns earlier)

The saved games above are all the same game, so you need just to back track or proceed from there, BUT only if you have time, if not no biggy.
 
The saved games above are all the same game, so you need just to back track or proceed from there, BUT only if you have time, if not no biggy.

I'll look as I'm able. So far I have looked at the slave example. Slaves do use the same AI as regular workers, but the reason this one behaves as it does is (semi) deliberate. It's moving next to a foreign unit inside its own territory. Because it is within its own territory it doesn't require an escort unless some actual enemy action has occured (like attack or pillage) since attacking would be an act of war, and it basically assumes you are not going to cause a war over a single worker (otherwise the presnce of explorer units liek this one would cause the AI to stop improving its tiles out of paranoi).

The flaw in this case is that you are ALREADY at war, sio the logic doesn't really apply. However, it's not so much a bug as a wekaness in the AI in thi regard, which is not new. As such I don't think it's urgent to fix. However, I will add it to my list of AI wekanesses to address, as it's probably low-hanging fruit and easy to do (so expect it sometime next week - it shouldn't prevent v21 however).

I'll edit this post as I get time to examine the other two cases...

Edit - I've loaded the 1390 save and I don' see the settler - can you load it up and find it, then post a descriptiion or image to tell me wheer to look? (probably going to have the same question for the 1490 case too when I get to that - in general please try to post a brief description, or map coordinates or soemthing to tell me where on the map to look)

Edit 2 - ok - the 1490 case I can see the problem unit and will analyse. Still can't find the settler in the 1390 case...

Edit 3 - ok, understand the townwatchman issue (1490). It's not new, or strictly a bug. Just an unintended stupidity which again I will try to address sometime soonish. Non-urgent and should not stop v21. Basically the town watchman is answering a request from a worker for an escort, but to reach said worker it has to pass through danger itself. Currently the code for units heading off to escort workers doesn't look for safe paths (since usually it is just traversing it own territory to go take up escort duties). In this case the nearest defensive unit able to answer the escort request has to cross your territory (or go around via a longer path) to do so, so it is putting itself in (severe!) danger in answering the request this way. I'll change the code to handle thi case, but it'll be next week sometime.

Both the save games are valuable test cases. I expect the 1390 one will be too, but I'll need more info on that one to identify where the problem is occurring.
 
I had a big paragraph, but that darn server cut me off, anyways heres a NEW settler one turn before moving into enemy territory

This one will be obvious right, since no description with it ;) ?

Edit - ok this one was obvious and it turns out to be an original BBAI bug. Again not new, nor super-urgent. I'll fix it next week sometime. It shouldn't hold up v21 release.

The bug occurs when a settler (witgh escorts) is headed to a city founding site, and on the way there finds there is danger nearby. At that point it unconditionally gives up on the founding , separates itself from it escorts (!!) and retreats to the nearest safe city. Couple of things wrong with this:

1) Separating from the escort before it reaches owned territory is dumb.

2) Giving up on the founding without evaluating how bad the danger is relative to the strength of its escort seem wrong.

3) Giving up on the founding without determining how close it is to completing its mission is sub-optimal. In your case it was actually ON the tile it wanted to found on!!
 
This one will be obvious right, since no description with it ;) ?

Like i said above i have a big paragraph prepared but that dan server keeps going out and when i hit submit it erased it all, so i did take any changes that time and just send what i could, sorry.

But here is another worker, but this time with an escort, BUT why are they even IN my territory, nothing to the right at all??
 
Back
Top Bottom