Current (SVN) development discussion thread

[Leoreth]

Okay, thanks for the saves so far. I haven't been able to fix the reported problem but instead discovered and fixed two related issues

 

[Lew09]

I've already got another one, playing as egypt. This time the game just crashes to desktop, even after i quit and restarted civ. It happened some turns before already, but only once, after reloading everything seemed fine until the second crash.
This happened with revision 583.

autosaves from the turns right before the crashes (powered by blizzrd):

 

[Lew09]

I almost forgot that in the same game the first UHV had the green checkmark (not sure if this word exists), but still read "not acomplished yet" next to it.

 

[Leoreth]

Are you playing with Python exceptions enabled? Otherwise I cannot explain the "not yet" situation. Or did it start after loading the game?

I also didn't experience a crash after ending the turn in your saves. Did you do anything special to trigger it?

 

[Bocator]

Hi,

I have noticed a little problem with the Unit Naming option in BUG Mod, which I use with GEM, here are my settings :

CombatAIR = ^cntc[o]^ ^ut^ Squadron
CombatARCHER = ^cntc[o]^ Light Infantry Brigade
CombatARMOR = ^cntc[o]^ Heavy Armored Brigade
CombatGUN = ^cntc[o]^ ^ut^ Brigade
CombatHELICOPTER = ^cntc[o]^ ^ut^ Squadron
CombatMELEE = ^cntc[o]^ Heavy Infantry Brigade
CombatMOUNTED = ^cntc[o]^ ^ut^ Brigade
CombatNAVAL = ^ut^ - ^cntu[r]^
CombatRECON = ^cntc[p]^ Recon
CombatSIEGE = ^cntc[o]^ ^ut^ Brigade

As you can see, I use cntc most of time, and cntu for naval units :

## - ^cntc[f]^ - count across same combat type (increments based on combat type)
## - ^cntu[f]^ - count across same unit (increments based on unit)
http://forums.civfanatics.com/showthread.php?t=241801

In my GEM games it works as intended so I get something like :

1st Cannon Brigade
2nd Cannon Brigade
1st Rifleman Brigade
Battleship - I
Destroyer - I
...

But in my RFC games it just counts accros all units (like if I just use cnt), so I get something like that :

1st Rifleman Brigade
2nd Cavalry Brigade
3rd Cannon Brigade
Battleship - IV
...

I know it's not that important but at least this is signaled.

I haven't test but I suspect that all of these don't work too

## - ^cntct[f]^ - count across same city (increments based on unit)
## - ^cntuct[f]^ - count across same unit / city (increments based on unit)
## - ^cntd[f]^ - count across same domain (increments based on domain)
## - ^tt1[f][x:y]^ - total where the total is a random number between x and y (number)
## - ^tt2[f][x]^ - total count (starts at x, incremented by 1 each time ^tt is reset to 1)

(never really understood how the last two work)

 

[blizzrd]

With BUG now enabled, I am seeing some random city names pop up when the Independents found a scripted city. Sana'a has been variously "Independents found Chinook" and this sort of thing.

The correct name appears next to the city, but the BUG message popup appears to be using a random list of independent city names. Which then seem to be overwritten by the scripted code with the correct name.

 

[Leoreth]

As for the automatic unit names, I have no idea how they work but I'll give it a look.

Is the city name announcement only wrong for independents but also for other civs?

 

[Lew09]

I wasn't sure what python exceptions enabled means (looked it up, will enable them), so i guess it's that. I'm sorry if you were chasing something that wasn't there because of this.

No, I don't think i did anything special in these turns, but what exactly would be special here?

 

[Leoreth]

Well, did the crash happen after you did something, or did it happen during the AI turns?

 

[Lew09]

I see. No, it happened during autoplay.

 

[blizzrd]

Is the city name announcement only wrong for independents but also for other civs?

Announcement said Korea founded Hanseong as their first city, which I think is correct?
I didn't notice anything odd about the AI civ announcements, just independents.

 

[BenZL43]

Yes Hanseong is how Seoul is founded.

 

[Leoreth]

Okay, Hanseong is definitely not the first name on the Korean standard list. I'm wondering what the difference is for independents.

 

[Malchar]

I had a fatal error on 583. I was playing Prussia 1700 AD, and when I capture and keep Belgrade from the Ottomans, it either crashes to the desktop immediately, or I get this:

MemoryError
File RFCUtils line 621 in convertPlotCulture
File RFCUtils line 1698 in cityConquestCulture
File CvRFCEventHandler line 273 in onCityAcquiredAndKept
File BugEventManager line 401 in _handleDefaultEvent

If I get the error above, I can continue the game, but a few turns later I get a popup that looks like this:

FATAL_ERROR
Memory allocation failure - exiting program
Reason:bad allocation

The attached save is from before capturing Belgrade.

 

[merijn_v1]

On what machine are your running?

Bad Memory Allocations do occur on 32bit systems. If you switch to a 64bit system, you don't get them anymore. Unfortunately, that means you need to get a new machine.

I had the same problem some time ago. But recently, I got a new machine. Now I don't get these messages anymore.

 

[Leoreth]

No there's definitely something wrong here. Conquering Belgrad froze my computer.

 

[Lew09]

Are you playing with Python exceptions enabled? Otherwise I cannot explain the "not yet" situation. Or did it start after loading the game?

I also didn't experience a crash after ending the turn in your saves. Did you do anything special to trigger it?

Hm, i just got a similar thing again in a new attempt as egypt, some turns later this time. Again no hints what could be causing this, and only in egyptian games. Could it be the greek conquerer event?( iirc, these are not in the exact same turn every time?)

 

[Leoreth]

In the same game, they will always happen in the same turn. But different games have different turns, yes.

 

[Lone Wolf]

Yeah, MAFs usually occur on all mods with large maps and a lot of graphics due to Civ4 memory management sucking. 64-bit machines allow programs to use more memory.

 

[Malchar]

I am on Windows XP, and I have had some memory problems in the past. However, I can usually complete the games with this mod because I've noticed that this mod isn't too intense. At worst, sometimes I have to restart the application if I've been playing for a while, and then it will work after that.

With this bug, however, it seems like something else is going wrong. I tried replaying from an autosave in 1884, which was before I even had Budapest, but I had the same problem when I captured and kept Belgrade.