Bug Reports

[jdog5000]

That's true, 1.51 does fix a barb civ bug that caused the game to crash so give that a try.

 

[VolkoV]

Thanks, the problem was solved when I returned to the 34-Civs-DLL. So it probably was my fault since I just wanted to use all without reading about XML-Modifications.

Great Mod by the way. Enhanced the game for me a lot. Maybe I can help with the German translation in return?

 

[jdog5000]

That would be wonderful! Some pieces (like BarbarianCiv) have been partially translated into German ... look through the files in Assets/XML/Text and when you get a chance to translate some, send me a PM.

 

[malikahsen]

Hi there... i'm facing a problem here... the problem is for example if i start playing game as an arabian leader... then later on i change it to inca's or anyother civ... then when i save it i'm unable to load it... like i can not load the game...??? i don know what going on???

 

[StormLord-711-]

Do you mean when you change to controlling a different player? Or just changing the civ/leader type for the player you already control?

With changing which player you control, there's not much I can do about that ... it might be possible to change the max range of the espionage weights so it's a little easier, not sure.

Well, it happened to me once when revolutions in my empire forced me to cede control to another leader. When I got my civ back, all the espionage sliders were on 99 or close to it. But it's also happened to be when I switch to controlling a different civ.

It's probably just how the AI handles its espionage since the AI can work easily with numbers that big. Maybe the espionage sliders could be reset to 0 after you change civs or take control back from an automated leader.

 

[glider1]

like i can not load the game...???

There seems to be a bug in the latest build of Rev 1.51. Espionage data and other data relating to contact with civs is not being saved to disk. It's ok in the current session but not from a save game.

To test it, do this:
1) Start a standard game
2) Auto play until contact with a few civs
3) Record down on paper the espionage ratio next to a civ's name
4) Save and reload from disk
5) Note the espionage ratio. The rival civs will be zero but your own civ's espionage rate was recorded.

Other "symptoms" are that it forgets initial contact with a civ from a save game. You contact a civ then reload from sav and you will have to meet him for the first time again which is unpleasant if the leader has bad breath.

Suggest that it's data streaming to disk issue that shouldn't be too hard to track down. I will work on it in the background.

Other comments:
The good news is that barbarian civ does seem fixed (initial testing)

These sources have uncommented changes probably for good reason:
CvTeamAI.cpp
CvTeam.cpp
CvPlayerAI.cpp eg (AI_reset(bool bConstructor))
CvPlayer.cpp eg (AI_reset(false))

Cheers

 

[jdog5000]

Dernit ... alright, fix will be up tonight.

 

[glider1]

Dernit ... alright, fix will be up tonight.

Thank you oh great Jdog.

You have fixed all that was wrong and gave us even more. May I ask you oh master, but how did you fix it? I see the changes before my eye's, however I am not sure whether I am mystified, perplexed or confused for I cannot see how it was fixed. I see this little line of code:
isRebel() const;
But this cannot be the solution to the fault that was exhibited. Could you please release my burden and explain this riddle to me your humble servant before the god of Revolution.

Yours forever indebted.

 

[jdog5000]

I moved a chunk of code from CvTeam::reset to CvTeam::init (the code has minors and barbs re-declare war on a re-initialized team). Now that you mention it though the poor code placement doesn't seem to appear in the 1.51 sources I uploaded ... can't figure out what happened there, cause I know that's what I changed yesterday and that's what the problem was.

 

[glider1]

Yeah it appears cause and effect are seperated by deep fog. I've looked through 1.51 to 1.52 again:

I think I can rule out these changes:
1) CvPlayer.cpp -> (ctrl-shift-p funtionality)
2) DynamicCivName.py -> def resetName( self, iPlayer, bVerbose = True ) :

These are the only possibilities! (I think)
1) CvTeam.cpp -> (bool CvTeam::isRebel() const)
2) CvTeam.h -> DllExport void reset(TeamTypes eID = NO_TEAM etc)

That there was something about assigning teams in the initial game stages that knocked out civ data. I simply cannot find anything else that could come vaguely close. All I know is that 1.52 is great so far and 1.51 could not save (and/or restore) civ data to/from disk.

This could be one of the best examples of the seperation of cause and effect I have witnessed in a while. This civ engine is pretty complex.

Cheers

 

[TheLastOne36]

Noticed something: When you change your civ (by ctrl-shift-x not ctrl-shift-p) your preference aren't moved to your new civ. Let me explain more clearly:

Options like "Stack Attack" and "Quick Moves" Reset when you change civ. i tried going to options and reenabling it, but it is still checked in ( [x] ) and if you recheck it, it still doesn't change. When i go back to my original civ, the options are back to normal and working again. It's frustrating because i'm in a war and i have to use my SoD against my enemies SoD and i have to click through each battle instead of having stack attack.

I hope i'm being clear enough. if you don't understand i'll try to reword it.

 

[jdog5000]

Noticed something: When you change your civ (by ctrl-shift-x not ctrl-shift-p) your preference aren't moved to your new civ. Let me explain more clearly:

Options like "Stack Attack" and "Quick Moves" Reset when you change civ. i tried going to options and reenabling it, but it is still checked in ( [x] ) and if you recheck it, it still doesn't change. When i go back to my original civ, the options are back to normal and working again. It's frustrating because i'm in a war and i have to use my SoD against my enemies SoD and i have to click through each battle instead of having stack attack.

I hope i'm being clear enough. if you don't understand i'll try to reword it.

Alright, I'll see what I can do ... there was some other setting which I noticed before not sticking, I believe it was workers leaving improvements or forests so I know what you mean.

Doing a quick search, seems like those settings are controlled by various PlayerOptionTypes in the DLL and are on a per player basis (makes total sense), it should be simple to have them copy over when you take control of a new player. The option screen check boxes not working after switching players is a little puzzling, I'll look into that as well.

 

[Cybah]

start as minor civs is bugged (or at least the screen with red/green). even when enabled, the screen shows a red button - except you play with all mods enabled (then everything is green).

and when you start playing with "start as minor civs" only enabled, it's not working at all most of the time (sometimes it works ).

please take a look at this.

 

[jdog5000]

Yeah, wow ... that was a blatant copy and paste error. If you want to fix it, open Revolution\Assets\Python\Revolution\RevolutionInit.py and on line 162 change GameOptionTypes.GAMEOPTION_DYNAMIC_CIV_NAMES to GameOptionTypes.GAMEOPTION_START_AS_MINORS.

Without that fix StartAsMinors will only work if both it and DynamicCivNames are enabled. DCN being on turns the SAM text green but there's a second check for the SAM option later so it doesn't actually run without being enabled.

Thanks for the report!

 

[Cybah]

ok thanks.


; Read Game options from XML, not .ini
ForceGameOptions = 0

is not working either. civ doesnt remember the settings and reads options from xml.

 

[jdog5000]

Yeah, we'll have to wait for the next patch for that one ...

 

[Kalimakhus]

If you form a permanent alliance with an AI and it happens that a BarbarianCiv is spawned on the same turn you get a CTD. The CTD is triggered by Python code on line #417 in BarbarianCiv.py.

I got this reported on the thread for RevolutionDCM Add-On for Rise of Mankind. The post is here

It is of course a very rare situation but I thought it should be reported anyway.

 

[StormLord-711-]

I was playing a game with the "choose religion" setting off, and I had the game ask me what religion I wanted when I unintentionally founded one (I had thought that someone already had the tech, but it was a different tech with a chosen religion). Later, I founded another religion but wasn't given the choice (which I should have had since someone founded taoism later).

While I'm talking about strange bugs, I also had someone found the United Nations (and another wonder, I forget) in ancient times on the second turn.

These issues might have been fixed though; I've been using version 1.50 still.

 

[jdog5000]

Kalimakhus:

Thanks for the report, I'll check it out and see what I can figure out. A Permanent Alliance basically adds one player to the other's team, looks like the crash is somehow related to teams.

StormLord:

Definitely haven't seen anything like that ... the game options were moved to the end of the list since 1.50, there were a bunch of difficult to track errors surrounding that earlier but I'll be sure to test Choose Religions.

 

[0100010]

I've been playing the Mod (glider1's merged RevolutionDCM, also merged with Rise of ManKind by Kalimakus) I consistently get a CTD on the turn after a Barbarian city spawns a new Civ. (34 civ DLL) (I get the message pop indicating one has spawned (if it was close enough), I take my turn, and end it, then CTD.) Python Debug log shows the Civ being crated "change complete", and it get a few debug lines for the CTD turn (usually a few combat debug messages) before crashing. Python Error log doesn't produce anything. Python exceptions are turned on, but don't display anything. I assume its crashing when the new Civ is attempting to take its 1st turn, but I'm not sure.

Note that one of the things I tried, was using world builder to delete all units that could attack each other just before the CTD turn so that no combat would occur. The CTD happened anyway, so combat actions were not the cause.

Can you give me any tips on what to check or where to insert debug print statements that can help me narrow down the problem? You fixed a BarbarianCiv CTD within the last version or two right? What files where those changes in and what were they? Maybe I can double check to make sure those fixes were including in this merged version.

I've also noticed a CTD when newly formed revolution city end its uprising period and is able to take its first turn, however this is not quite as consistent.