Current (SVN) development discussion thread

[Leoreth]

New commit:
- Mexico starts with a Catholic Monastery
- civ switches now work properly even if multiple civs spawn in the same turn
- national wonders are properly removed when a city changes hands
- excluded the northern coast of Anatolia from the Italian UHV goal area
- AI civilizations cannot collapse anymore when a human player declares war on them
- fixed the unit graphics to correctly advance through the eras
- fixed the Colombian UHV so the third goal doesn't fail when Astronomy is discovered
- Triumphal Arc information is properly displayed without abusing the latitude tag
- changed the civic production modifier display to mention all affected buildings in one line

 

[Leoreth]

I loaded a savegame from prior to the latest Argentina change, in which I was playing as Russia. Every new turn I'm getting this:

Traceback (most recent call last):
  File "BugEventManager", line 400, in _handleDefaultEvent
  File "CvRFCEventHandler", line 672, in onBeginPlayerTurn
  File "Victory", line 2406, in checkPlayerTurn
  File "StoredData", line 305, in getArgentineGoldenAgeTurns
KeyError: 'iArgentineGoldenAgeTurns'

Not sure what's going on there.

Sorry, didn't read the entire post before. The Argentina update breaks savegame compatibility, I should have mentioned that. Same goes for the last update too, by the way.

Also about Russia, their UHV display is problematic. The first one is to settle 7 cities in Siberia by 1700 and have a railway line from Moscow to a Siberian city with a harbor in 1920. The display does not keep the count of how many cities in Siberia you've settled, it always appears as 0/7. Once you get 7, the goal is checked as completed, despite the lack of a railway line, but for that play session only. After loading the savegame with it completed, it will appear as uncompleted again, with 0/7 cities in Siberia and red crossmarks for both goals, even if the railway line has been made. I don't know if by 1920 the second part will appear as completed, or if that first goal can be completed satisfactorily at all.

Also quoting this for future reference.

 

[migck]

Yeah should have figured savegames will have trouble with updates, it's not like I will miss them anyways.
The thing with Russia has been discussed previously I've seen. More battles that await you...

- Triumphal Arc information is properly displayed without abusing the latitude tag

Now it displays a "can only be built at a minimum latitude of 90º", it should have been set to 0 like all other buildings. Oh well, who cares anyways, triumphal arches are just as effective at the equator than at any other latitude

 

[Leoreth]

Now it displays a "can only be built at a minimum latitude of 90º", it should have been set to 0 like all other buildings. Oh well, who cares anyways, triumphal arches are just as effective at the equator than at any other latitude

... not after the next commit.

 

[Leoreth]

New commit:
- civic requirements for buildings are now displayed in civilopedia and the building tooltip
- buildings whose civic requirements are not met now are displayed as grayed-out buttons instead of not at all
- changed the river and coast requirement display to match requirements

 

[Imperator Knoedel]

- AI civilizations cannot collapse anymore when a human player declares war on them

So you can keep a given civ perpetually alive by declaring war on them? What?

 

[Leoreth]

No. A declaration of war triggers a stability check that can but doesn't have to lead to collapse. The civilization can still collapse from any other stability check caused by a different reason, if its situation does not improve.

The new rule is only to prevent the frustrating experience of having your war targets collapse just the moment where you attack them.

 

[migck]

I get this one often when a civ resurrects, most often China and Vikings:

Traceback (most recent call last):
  File "BugEventManager", line 400, in _handleDefaultEvent
  File "CvRFCEventHandler", line 641, in onBeginGameTurn
  File "RiseAndFall", line 1294, in checkTurn
  File "Stability", line 1903, in checkResurrection
  File "Stability", line 2057, in doResurrection
RuntimeError: unidentifiable C++ exception

usually pops when the autoplay is running as well. No idea what could be wrong, all there is at that line is a 'bCapital = city.isCapital()' assignment. And the resurrection seems to carry on fine regardless, so, non-issue?

I also forgot to document something quite exceptional: in a 3000BC Brazil start, the Khmer had managed to build Wat Preah Pishnulok at Angkor, control Pagan, Hanoi and Singapore, and remain stable throughout, so much that the Thai didn't spawn. But later on the Chinese took control of Hanoi, eventually collapsed, and Hanoi was reassigned to... the Thai. If anything it's the Khmer the ones who end up relegated north of Thailand as Vietnam/Burma, not the other way round. Shouldn't that be checked so that if the Khmer manage to perform so well, the only chance for the Thai to appear is as a Khmer resurrection later on?

Also, I can still do great works with prophets while controlling the Sphinx and having Drama. Or maybe you're fixing that later for reasons and I'm just being a pest

 

[morff]

I think Arabia UHV is broken:
1. goal (control) everything was OK, except controlling spain... i had 4 of 5 cities owned by vassals, but it didn't work out
2. goal (most advanced) i was in technological lead (according to the graph), but i failed the goal anyhow

 

[Leoreth]

I get this one often when a civ resurrects, most often China and Vikings:

Traceback (most recent call last):
  File "BugEventManager", line 400, in _handleDefaultEvent
  File "CvRFCEventHandler", line 641, in onBeginGameTurn
  File "RiseAndFall", line 1294, in checkTurn
  File "Stability", line 1903, in checkResurrection
  File "Stability", line 2057, in doResurrection
RuntimeError: unidentifiable C++ exception

usually pops when the autoplay is running as well. No idea what could be wrong, all there is at that line is a 'bCapital = city.isCapital()' assignment. And the resurrection seems to carry on fine regardless, so, non-issue?

Not sure, I'll have to look at the code in more detail.

I also forgot to document something quite exceptional: in a 3000BC Brazil start, the Khmer had managed to build Wat Preah Pishnulok at Angkor, control Pagan, Hanoi and Singapore, and remain stable throughout, so much that the Thai didn't spawn. But later on the Chinese took control of Hanoi, eventually collapsed, and Hanoi was reassigned to... the Thai. If anything it's the Khmer the ones who end up relegated north of Thailand as Vietnam/Burma, not the other way round. Shouldn't that be checked so that if the Khmer manage to perform so well, the only chance for the Thai to appear is as a Khmer resurrection later on?

So am I following correctly?

1. Khmer are stable and the Thai do not spawn
2. China conquers Hanoi (from the Khmer?)
3. China collapses (while the Khmer are still alive?)
4. Hanoi goes to Thailand instead of the Khmer, although Thailand has never been alive?

I'd prefer a save to find out the actual reason of the assignment but I suppose you haven't kept it.

Also, I can still do great works with prophets while controlling the Sphinx and having Drama. Or maybe you're fixing that later for reasons and I'm just being a pest

No, it's just hard to find all instances of this problem because they're scattered all across the source code.

I think Arabia UHV is broken:
1. goal (control) everything was OK, except controlling spain... i had 4 of 5 cities owned by vassals, but it didn't work out
2. goal (most advanced) i was in technological lead (according to the graph), but i failed the goal anyhow

1. What about the fifth city?
2. I'd need a save to see what was wrong.

 

[migck]

So am I following correctly?

1. Khmer are stable and the Thai do not spawn
2. China conquers Hanoi (from the Khmer?)
3. China collapses (while the Khmer are still alive?)
4. Hanoi goes to Thailand instead of the Khmer, although Thailand has never been alive?

I'd prefer a save to find out the actual reason of the assignment but I suppose you haven't kept it.

Sorry I didn't, I purge loads of savegames every day, makes me wonder why I make them in the first place. Though I did keep a screenshot:

Spoiler :

kinda useless, I should also have retired from the game and see the replay to learn exactly what happened there, as you can see Hanoi has neither Hinduism nor Buddhism, strange.
Mongols were extra lucky at avoiding collapse, they made a second conquest of China!

 

[FinalLegendZero]

I think Arabia UHV is broken:
1. goal (control) everything was OK, except controlling spain... i had 4 of 5 cities owned by vassals, but it didn't work out
2. goal (most advanced) i was in technological lead (according to the graph), but i failed the goal anyhow

I just checked, and the UHV is working normally.

1. You need to control ALL cities in Spain. 4 out of 5 isn't enough.

2. It's possible a civ you have not yet encountered, or just recently encountered, is more technologically advanced; the graph only shows civs you are in contact with, and only if you have enough Espionage Points against them.

 

[morff]

OK i will try again the Arabia... i expected control = control more cities than any civilization in that area.

 

[citis]

Sorry I didn't, I purge loads of savegames every day, makes me wonder why I make them in the first place. Though I did keep a screenshot:

Spoiler :

kinda useless, I should also have retired from the game and see the replay to learn exactly what happened there, as you can see Hanoi has neither Hinduism nor Buddhism, strange.
Mongols were extra lucky at avoiding collapse, they made a second conquest of China!

Civs with conditional spawns are still in the game. They may don't spawn on the first place, but their core area is assigned and they may respawn later on.

 

[citis]

New commit:
- Mexico starts with a Catholic Monastery
- civ switches now work properly even if multiple civs spawn in the same turn
- national wonders are properly removed when a city changes hands
- excluded the northern coast of Anatolia from the Italian UHV goal area
- AI civilizations cannot collapse anymore when a human player declares war on them
- fixed the unit graphics to correctly advance through the eras
- fixed the Colombian UHV so the third goal doesn't fail when Astronomy is discovered
- Triumphal Arc information is properly displayed without abusing the latitude tag
- changed the civic production modifier display to mention all affected buildings in one line

 

[morff]

I noticed the AI has a lot of problems being stable in most recent SVN versions. They are all (and i mean all of them) being unstable / collapsing... collapsing and resurrecting and collapsing again... I think there is something fishy in the new stability.

 

[Hippo8085]

^ I've been noticing that many civilizations which are "meant" to collapse are staying stable and many that are not do this cycle... more on this in the Suggestions thread.

Sort of on the subject of the Thai situation, America has a similar problem because their early rebirth (hope I got the terminology right there, been a while since I last worried about it) date is 1770, before they spawn, though in my opinion this is not really the problem here.
Rarely (but I have seen it happen), a civilization in their flip zone collapses and they are born early and for some reason the AI is very weak, which is the problem. Regular spawn America does well in most of my games, so this is a one digit change in consts to 1775 as the early rebirth year.

 

[Qiu]

Does the Brazilian UP count the number of sugars？ I tried to seize 5-6 sugars，but the oil company still produces poor 3 gold. That made me so frustrated...

 

[Leoreth]

New commit:
- corrected the Russian UHV display
- fixed a bug that caused a crash when an optional civilization is disabled

 

[Leoreth]

New commit:
- various bugfixes
- spelling fixes