v 25, Bugs/Crashes reporting thread

[krushyn]

......

@Krushyn: what rev are you on with that game? I need to know when I go to take a look at debugging it.

It was 3519......
I am just updating to 21, let see.....

Edit: Same thing.... straight CTD with 3521 (in all my attempts I try with viewports on and off no change in behavior)

 

[strategyonly]

I am experiencing the same condition as SO reported (visibility of hidden units in my territory), but this answer confuses me. If every tile I own lets me see hidden units, why would I ever build dogs?

I don't remember the exact previous versions but I do recall in the past that hidden units remained hidden in your territory. You had to guard workers and generals that were located "safely" within your empire, and it was wise to have pickets along your perimeter that spotted hidden units. I think those were good challenges for the player to overcome. It may have been as recent as V24 that required dogs or special units to see hidden units in my territory.

EDIT: I am playing without viewports turned on.

On the topic, are there any buildings that grant visibility on hidden units? I think the game Rise of Nations had that and I did like that feature. There are buildings that extend a city's visibility, but I have not noticed (or played C2C into later eras) to find a building that let's me see hidden units.

In regard to picture 2, you have visibility from every tile you own, so this is probably correct.

How do i turn that off?? It defeats the purpose of the camo/invis

 

[ls612]

It was 3519......
I am just updating to 21, let see.....

Edit: Same thing.... straight CTD with 3521 (in all my attempts I try with viewports on and off no change in behavior)

It's crashing when someone is trying to build a new city.

 

[krushyn]

It's crashing when someone is trying to build a new city.

When you reveal the map you will see that all continents and islands are full of civs and cities.... taking on account that ity is a gigantic map.... did I reach some number of cities limit?

 

[Thunderbrd]

Camel Archers are broken, when you attack with one, it will withdraw even though uninjured. You then cannot attack at all with ANY other camel archer anywhere.

I was curious about this since I've been working on that area of coding so I checked the camel archer xml. Apparently it has a combat limit of 60 so it can't take an enemy below 60% off of its full hp. This will mean once it reaches that point, it will automatically withdraw.

As to the rest of them not being able to attack... that's a bit strange and I can't say for sure why that's taking place. However, theoretically, you could've only been trying to attack where units were already under the 60% hp limit.

Generally this is a setting reserved for collateral units. I get the concept as to why it was put on the camel archer, but I should also mention that for cases like the camel archer, a new system is about to be put in place that will represent an improvement of function for this kind of dynamic.

 

[strategyonly]

When you reveal the map you will see that all continents and islands are full of civs and cities.... taking on account that it is a gigantic map.... did I reach some number of cities limit?

If its Gigantic, what are your spec on the PC, and sorry if anyone has asked this already??

 

[calvitix]

When you reveal the map you will see that all continents and islands are full of civs and cities.... taking on account that ity is a gigantic map.... did I reach some number of cities limit?

Yeah, there is a problem with AI player 13 (Israel), trying to build a city on plot 165,58. For any reason, the city is not built, but the game try to access to the city pointer, to choose production.
It appends in CvUnit::found(), ln8210.
I manage to skip that line, and the game goes on. Here a save for next turn. Try if you can continue : there are a lot of AI settlers that are on their way to found cities. If you reached a limit, you'll know it quickly.

 

[ls612]

Yeah, there is a problem with AI player 13 (Israel), trying to build a city on plot 165,58. For any reason, the city is not built, but the game try to access to the city pointer, to choose production.
It appends in CvUnit::found(), ln8210.
I manage to skip that line, and the game goes on. Here a save for next turn. Try if you can continue : there are a lot of AI settlers that are on their way to found cities. If you reached a limit, you'll know it quickly.

I'll point this out to Koshling so he can look at it when he gets back.

 

[Koshling]

I'll point this out to Koshling so he can look at it when he gets back.

Yeh, crashes and wfocs will be my top priority when I get back later this week. This is the third save that needs looking a right? (two crashes, one WFoC)

 

[ls612]

Yeh, crashes and wfocs will be my top priority when I get back later this week. This is the third save that needs looking a right? (two crashes, one WFoC)

Correct.

 

[krushyn]

Yeah, there is a problem with AI player 13 (Israel), trying to build a city on plot 165,58. For any reason, the city is not built, but the game try to access to the city pointer, to choose production.
It appends in CvUnit::found(), ln8210.
I manage to skip that line, and the game goes on. Here a save for next turn. Try if you can continue : there are a lot of AI settlers that are on their way to found cities. If you reached a limit, you'll know it quickly.

Muchas gracias herr General .... I'm much obliged ....

Edit: So far so good..... after cleanup I did a few turns and no problem so far...

 

[Thunderbrd]

I'm pushing a workaround 'fix' to the svn now but I can't say I was able to find the root of the problem.

For those who wish to review my changes, simply do a full project search for "TB Debug". You'll find where the issues were and what I did about them.

Takes a while to push changes... so be patient. I see you've already got a workaround though.

 

[strategyonly]

With that last dll change i get this:

Asset Failed

File: CvCity.cpp
Expression: isInViewport()
Message . . . . . . . ... .. . ....

EDIT:

Asset Failed

CvArea.cpp
Line 660
Expression: getPower(eIndex)>=0

 

[krushyn]

With that last dll change i get this:

Asset Failed

File: CvCity.cpp
Expression: isInViewport()
Message . . . . . . . ... .. . ....

I just updated and tried to load. The first one :

Assert Failed

File: CvGlobals.cpp
Line: 4884
Expression: stricmp(szType, "NONE")==0 || strcmp(szType, "")==0
Message: info type BUILDING_TRENCHTACTICS not found, Current XML file is: modules\CIVPlayer8\Buildings\CP8_CIV4BuildingInfos.xml

----------------------------------------------------------

Edit1:
.... and than few more .....after map showed :
Assert Failed

File: CvCity.cpp
Line: 16212
Expression: getReligionInfluence(eIndex) >= 0
Message:

----------------------------------------------------------
..... and still recalculating .....

... and I am giving up waiting .... revert

Edit2:
I am back to '22 and playing

 

[calvitix]

With that last dll change i get this:

Asset Failed

File: CvCity.cpp
Expression: isInViewport()
Message . . . . . . . ... .. . ....

EDIT:

Asset Failed

CvArea.cpp
Line 660
Expression: getPower(eIndex)>=0

Those Assertion are not new : the DLL uploaded is in Debug mod (not in release). I'll compile it again to avoid those displaying assertions.

 

[krushyn]

weird thing happened... I was about to conquer some city with about 10-12 healthy units and only one opponent unit.... I killed it got a pink info pop circle like when some event happened in that square and lost all but 2 unit in an instant.... anybody had that?

Edit: It is the city Cahokia of Native America ..... tried few times and almost all lost units but two disappeared (Bermuda?) on one obsolete archer with that weird pink pop

 

[Koshling]

Those Assertion are not new : the DLL uploaded is in Debug mod (not in release). I'll compile it again to avoid those displaying assertions.

More to the point, the DEBUG version is approximately 10-20 times slowest than the final release version. The one uploaded should always be 'final release' ( along with the matchng PDB). Note that 'release' is a halfway house, since it includeds profiling, which also slows things down (though no where near as much), so build 'final release' and do a final check test run with that before your final push to SVN each time.

 

[Thunderbrd]

Yeah, I forgot to put the right dll in place. Thought I'd already done that before pushing. Sorry. Won't let that happen again.

Also... I tried to create a fix for those asserts but they didn't work like I thought they would. In fact, as I've been thinking about it, those fixes should probably be reverted. I tried to establish that 0 should be the minimum amount possible on those but in retrospect, this isn't the way it should actually be done as it can throw off the value when the offending negative modifier is somehow removed. After some thought on the matter, however, I think I have a better fix in mind. So I'll address those and mark them with the TB Debug notes so y'all can review.

Of course, its only due to my mistake that y'all were seeing them come up at all...

Mostly, they stem from xml values that weren't expected to ever create those situations. For example, some water plots are coming up with negative food values as a result of -5 food per plot from the Garbage Docks. There were some there though that simply confused me a bit as to why they were evaluating to the values they were getting - in particular unit counts by civilization often ended up being 1 less than the count should have been.

 

[strategyonly]

Is there anyway you'll can re-name the dll Description from "Game core DLL for Rise of Mankind: A New Dawn to "Caveman2Cosmos."

 

[strategyonly]

Was using Viewports for the first time, and i tried and tried to get to another screen but it wouldn't let me, so i clicked on the minimap flag and it took me to the Capital city where i clicked on a unit there and POOF CTD