Fimbulvetr
Emperor
0.83d if memory serves me right. BBAI.log last modified date is 28.12.2009 at least.
Specific Bug Reports
- Thread starter Iustus
- Start date
phungus420
Deity
- Joined
- Mar 1, 2003
- Messages
- 6,296
Personally I thought the missionary behavior in 0.82 was really good. Previously the AI did not spread it's state religion, and thus gain the civic advantages from having the state religion, unless the AI had the holy city. Since 0.82 the AI does spread it's state religion, which helps immensely for calming evolutionary sentiment, and helps the AI gain it's civic bonuses. It may have been a little overzealous, but nothing to write home about, and was certainly much better then previous behavior. I also don't think having a missionary to spare in case you capture or found a new city is a bad idea either, which the AI would do if it built an extra one.
Fimbulvetr
Emperor
Nvm, I was wrong, the Player had open boarders - with me. But he was on the other side of the pangaea and I believe I was blocked off by my neighbour really soon, made me unreachable over land, so the missionaries just sat there in his captial until the open borders with player 6 (which was not my neighbour but still opened up 3 cities to spread the religion to).
Player 5, the Dutch. Amsterdam:
2nd city, Utrecht:
T86(!): Utrecht is founded -> builds Granary food 2)
T96: city grows to 2
T98: 1 pop rush (bad tiles) to finish Granary -> builds Missionary (3) (abandoned to T107)
Open Borders:
T56 with Player 0
T76 with Player 6
T81 with Player 1
T99 with Player 3
..T118 with Player 2
Tech: starts with Fishing and Agriculture
Turn 100 score comparison (Player 0 is human)
2 cities, 4 pop and 1 Settler at turn 100.: that's where I say something went wrong. But I admit, it looked worse on the map than it does here as text, and maybe that's all normal for a religious zealot (I always have random personalities on)
Am I the only one who has troubles with the forum today?
Player 5, the Dutch. Amsterdam:
Spoiler :
2nd city, Utrecht:
T86(!): Utrecht is founded -> builds Granary food 2)
T96: city grows to 2
T98: 1 pop rush (bad tiles) to finish Granary -> builds Missionary (3) (abandoned to T107)
Open Borders:
T56 with Player 0
T76 with Player 6
T81 with Player 1
T99 with Player 3
..T118 with Player 2
Tech: starts with Fishing and Agriculture
Spoiler :
Turn 100 score comparison (Player 0 is human)
Spoiler :
2 cities, 4 pop and 1 Settler at turn 100.: that's where I say something went wrong. But I admit, it looked worse on the map than it does here as text, and maybe that's all normal for a religious zealot (I always have random personalities on)
Am I the only one who has troubles with the forum today?
Afforess
The White Wizard
No, the forums have been really spotty today. Lots of database errors and slow pages. There server must have some issues.
jdog5000
Revolutionary
It certainly is a lot of time spent building missionaries early in the game for this player ... multiple missionaries and Stonehenge before the second city is not a good strategy. I'll think about what can be done to shift the missionary pump back a little bit for those that do go that route.
I'm playing BTS on linux, wine. Version 0.82 doesn;t have this bug,
so this is something introduced in 0.83. Huge map, 18 civs,
when there are lot of units with developed civs, game crashes on same place
always (sometime I can revert back few moves and it continues)
. Reverted to version 0.82 dll same save, all other
files are from 0.83, game continued to
work without problem till the end.
I use mod, (changed only fields in xml files, nothing special), game lasts
more then 3000 turns, around turn 1400 start problems with 0.83 dll,
hope this helps.
Problem is in 0.83 dll since with 0.82 dll version game works
without any problems.
so this is something introduced in 0.83. Huge map, 18 civs,
when there are lot of units with developed civs, game crashes on same place
always (sometime I can revert back few moves and it continues)
. Reverted to version 0.82 dll same save, all other
files are from 0.83, game continued to
work without problem till the end.
I use mod, (changed only fields in xml files, nothing special), game lasts
more then 3000 turns, around turn 1400 start problems with 0.83 dll,
hope this helps.
Problem is in 0.83 dll since with 0.82 dll version game works
without any problems.
phungus420
Deity
- Joined
- Mar 1, 2003
- Messages
- 6,296
That's probably the same crash I have reported here:
http://forums.civfanatics.com/showpost.php?p=8827247&postcount=718
If you could provide a save game where it crashes that would help debug it. My crash report and save game is for the current LoR SVN, which would require jdog to debug from LoR's source. Your crash occurring on pure BetterAI should be easier to debug. So please upload a save
http://forums.civfanatics.com/showpost.php?p=8827247&postcount=718
If you could provide a save game where it crashes that would help debug it. My crash report and save game is for the current LoR SVN, which would require jdog to debug from LoR's source. Your crash occurring on pure BetterAI should be easier to debug. So please upload a save
I think that it is easily reproducible without any mod. Marathon speed, 18 civs,
custom_continents, around tank era, bug will show up,
just autoplay. I think it has to do with units,
as when there are lot of units crash happens during wait for other
civilizations turn.
Yes, that is exact bug you have, no segfault, process just
silently ends.
I think that savegame would not help since this is modded, too,
but I think everyone that play huge maps and have lot of
units will hit this bug.
custom_continents, around tank era, bug will show up,
just autoplay. I think it has to do with units,
as when there are lot of units crash happens during wait for other
civilizations turn.
Yes, that is exact bug you have, no segfault, process just
silently ends.
I think that savegame would not help since this is modded, too,
but I think everyone that play huge maps and have lot of
units will hit this bug.
This is error:
err:seh:setup_exception_record stack overflow 848 bytes in thread 003b eip 7bc486d7 esp 00240fe0 stack 0x240000-0x241000-0x340000
After I increased stack size to 128mb, I got:
err:seh:setup_exception_record stack overflow 832 bytes in thread 003b eip 7bc486d7 esp 01010ff0 stack 0x1010000-0x1011000-0x8a30000
In second case it just needs more time to overflow. So it's endless loop, precisely
infinite recursion, that eventually overflows stack space.
While loop is executing, program is unresponsive, until stack overflow, than it
crashes.
So in version 0.83 there is condition that triggers infinite recursion somewhere
in new code, probably.
Second option is that something allocates too much stack, (windows program
stack is 1mb usually), but then again with 128mb stack size it executes
5 mins before overflow, I could increase stack more but I think it's infinite...
edit: in 3000 turns I got bug only twice, but reverting to 0.82 dll for couple of turns,
then bringing back 0.83, makes game playable, no problem. So it's some
rare specific situation when bug triggers, it's not game breaking at all.
err:seh:setup_exception_record stack overflow 848 bytes in thread 003b eip 7bc486d7 esp 00240fe0 stack 0x240000-0x241000-0x340000
After I increased stack size to 128mb, I got:
err:seh:setup_exception_record stack overflow 832 bytes in thread 003b eip 7bc486d7 esp 01010ff0 stack 0x1010000-0x1011000-0x8a30000
In second case it just needs more time to overflow. So it's endless loop, precisely
infinite recursion, that eventually overflows stack space.
While loop is executing, program is unresponsive, until stack overflow, than it
crashes.
So in version 0.83 there is condition that triggers infinite recursion somewhere
in new code, probably.
Second option is that something allocates too much stack, (windows program
stack is 1mb usually), but then again with 128mb stack size it executes
5 mins before overflow, I could increase stack more but I think it's infinite...
edit: in 3000 turns I got bug only twice, but reverting to 0.82 dll for couple of turns,
then bringing back 0.83, makes game playable, no problem. So it's some
rare specific situation when bug triggers, it's not game breaking at all.
Still_Asleep
Warlord
Same here :/
I am using Fuyu's Better Bug AI from 18.1.2010, so no clean Better AI, but almost clean
Save is attached: End the turn, and the game will eventually quit and go back to desktop without any error message.
It is a huge fractal map, 18civs, epic speed, immortal, gametime is 1786. Tanks, bombers and generally a lot of stuff crawling around, somehow killing my game.
Hope Grandmaster JDog or some other coder can debug this
I am using Fuyu's Better Bug AI from 18.1.2010, so no clean Better AI, but almost clean
Save is attached: End the turn, and the game will eventually quit and go back to desktop without any error message.
It is a huge fractal map, 18civs, epic speed, immortal, gametime is 1786. Tanks, bombers and generally a lot of stuff crawling around, somehow killing my game.
Hope Grandmaster JDog or some other coder can debug this
Fimbulvetr
Emperor
Replace your dll with the one from the official Better AI and see if the CTD remains. My dll is not clean BBAI, it has all of BULL and a tiny bit of original code too (even if that shouldn't do any harm).
Trying to load the savegame always results in "requires mod loaded: NONE" message. How I can deal with that I still have no clue but then again I wouldn't be able to help either way.
Trying to load the savegame always results in "requires mod loaded: NONE" message. How I can deal with that I still have no clue but then again I wouldn't be able to help either way.
Still_Asleep
Warlord
I couldnt load the save after replacing my dll, got a runtime error the second a clicked on the file. Changed the complete assets folder, same result. Switched back to your better bug AI and it works like before (quits in opponents turn).
I didnt install better bug AI as mod, just replaced the standard files - that might explain the error message you are getting.
Looks like I need to roll back to 0.82 and start a new game until someone figures out what's wrong.
Edit: In fact I switched to the latest Better Bug AI, that made no difference.
I didnt install better bug AI as mod, just replaced the standard files - that might explain the error message you are getting.
Looks like I need to roll back to 0.82 and start a new game until someone figures out what's wrong.
Edit: In fact I switched to the latest Better Bug AI, that made no difference.
Fimbulvetr
Emperor
For some reason I was thinking BULL didn't wouldn't affect save game compatibility .. I guess I was wrong.
What you can do is enter the worldbuilder and create a worldbuildersave. Everyone should be able to load that as a scenario iirc. The real question then would be if the error remains if you load from scenario or if it's gone after that. In the former case you could simply roll back to an earlier BBAI version and resume your game from the scenario file.
Latest Better BUG AI has almost no changes to what you were using so that's no surprise it wouldn't help.
edit: If it's the same thing bmaxa and phungus reported (BBAI + lots of units = CTD) then we'll just have to wait for jdog to figure that one out.
What you can do is enter the worldbuilder and create a worldbuildersave. Everyone should be able to load that as a scenario iirc. The real question then would be if the error remains if you load from scenario or if it's gone after that. In the former case you could simply roll back to an earlier BBAI version and resume your game from the scenario file.
Latest Better BUG AI has almost no changes to what you were using so that's no surprise it wouldn't help.
edit: If it's the same thing bmaxa and phungus reported (BBAI + lots of units = CTD) then we'll just have to wait for jdog to figure that one out.
Still_Asleep
Warlord
I tried it with a worldbuildersave, the error is indeed gone.
Unless someone is able to load my initial savegame I guess I got nothing that can be debugged :/
Unless someone is able to load my initial savegame I guess I got nothing that can be debugged :/
jdog5000
Revolutionary
bmaxa, phungus -
Sounds like an infinite loop to me too ... fun fun fun. In the past these have always been because of changes to how units group, but there weren't any changes like that in 0.83. I'll see if I can reproduce this and what I can figure out, but I'm currently traveling so it may take me a while to track it down.
Sounds like an infinite loop to me too ... fun fun fun. In the past these have always been because of changes to how units group, but there weren't any changes like that in 0.83. I'll see if I can reproduce this and what I can figure out, but I'm currently traveling so it may take me a while to track it down.
Hello Yakk...if you are referring to my post, the game crashes in 1562AD...I have a fair to middling sized empire, and QSH declares war on me in this turn (and always does, even with random seed reload), and then the game crashes on the next turn....one possibility I thought of is that I'm using an 18 Civ DLL, and for this game I used the maximum 18 civs...perhaps one of my competitors is trying to release a colony somewhere else on the globe? Just a guess.
Regards,
Jeff
Afforess
The White Wizard
I just grabbed a copy of BBAI to see if I could take a shot at fixing the crash issue... but here's the fun part. I compiled a new debug dll so I could see if I could pin-point the cause of the crash for Jdog, so he can get back to work quicker, but the Debug DLL didn't crash. The regular dll did. So this is one of those crashes that debug dll's avoid... making it 10x harder to trace down.
I just grabbed a copy of BBAI to see if I could take a shot at fixing the crash issue... but here's the fun part. I compiled a new debug dll so I could see if I could pin-point the cause of the crash for Jdog, so he can get back to work quicker, but the Debug DLL didn't crash. The regular dll did. So this is one of those crashes that debug dll's avoid... making it 10x harder to trace down.
:suicide:
Fun times ahead.
I'll hold out on using 0.83 for the moment, I guess.
Similar threads
