RFC Europe playtesting feedback thread

[sedna17]

EDIT: Ok, I removed the call to choosePurgeCity() in AI_update() meaning that prosecutors now dont do anything, recompiled the dll, and now the game goes past the turn where it previously always crashed.

This is great news, thanks ecv. This is quite a plausible cause for everyone's late game crashes -- I'm not sure the AI normally builds many prosecutors at all. I don't have the ability to compile new versions of the dll (3Miro does), but for now, would you mind posting your recompiled dll here (or in the RFC Europe "files" thread) so that I can distribute it with the next test version and see if it fixes everyone's problems? If so, we'll fix the AI code for the prosecutors as soon as possible.

 

[micbic]

@jessiecat: My game as Kievan experiences problems. First of all, at first load, no interface had appeared (see screenshot). So I saved and reloaded. The interface appeared, but next turn...all civs contacted me (some of then had minus or zero score-see screenshot) and I had to vote someone for UN!!!. After that, every reload crashes at 1248 AD. No matter, I will restart from scratch.

@ecv: That might be a cause to that, but AFAIK prosecutors were added much before Early Renaissance crashes begun.

 

[sedna17]

@jessiecat: My game as Kievan experiences problems. First of all, at first load, no interface had appeared (see screenshot). So I saved and reloaded. The interface appeared, but next turn...all civs contacted me (some of then had minus or zero score-see screenshot) and I had to vote someone for UN!!!. After that, every reload crashes at 1248 AD. No matter, I will restart from scratch.

That sounds like a corrupted save game file. This can happen sometimes and might be unrelated to our mod. If you have previous autosaves you could try playing on from one of them (but you might have overwritten them playing on from the bad load).

 

[ecv]

Hello,
Here's som feedback. In a game as the germans (with the dec. 21st version), it seemed to me that they are in a fairly weak position. In my game atleast, both Frankia and Burgundy had armies that the starting german would be unable to defend against. I think they need a few more defensive units from the start.

The swedish spawn map in the dec. 27th version should imo not include scania. I suggest removing the last 3 rows from it. The population in Scania were danish from before the year 800 till many years after the swedish conquest in 1658, and it was through a loong series of wars (11 I think), that the province became part of Sweden instead of Denmark. You can read more about it here: http://en.wikipedia.org/wiki/Skåneland

 

[sedna17]

Thanks, I did not know that (despite growing up in a region dominated by mock-serious feuding between Swedes and Norwegians). I can fix it as you suggest. I can also add that area to the Norse spawn area, which would have minimal direct implications.

 

[merijn_v1]

When I played Frankia, Protestantism was founded in Bastia (Genoa, Corsica). The St. Peter's Basilica was placed in the Holy-City and not All Saints' Church.

 

[Arwon]

Playing as Cordoba, I captured a Spanish city called Guadalajara and it got renamed as Burgos. That seemed a little odd.

Also, I built a city where I think Zaragoza should be, on the wheat tile on the lowland on the Ebro. Instead, it became the Arabic equivalent of Pamplona - seemed too far south for that.

 

[Arwon]

Also in that Cordoba game, Arabia began as Orthodox...

 

[jessiecat]

Also in that Cordoba game, Arabia began as Orthodox...

Hi Arwon. Usually see you on OT. What you've found is largely because we've tended to tag more than 1 tile for the important cities, up to 4 in some cases. We could have included places like Logrono, Vitoria etc. between Zaragoza and Pamplona and made it 1 tile per city but most of the names would be unfamiliar to most players. That could be changed though.
I did the Spain and the Cordoba maps and possibility they don't coincide 100%. I'll take a look at that again. I need to make revisions on the Spain map anyway.
There's no way Arabia should be Orthodox as Islam is founded automatically when Damascus is built. (Or so so it should.) I'd like to see a saved game of that. Could you post one here?

 

[3Miro]

Just posting, will not be able to work for another couple of weeks.

There seems to be some reports on religious related bugs. I have no way of testing/fixing anything right now, but it sounds like a consts.py problem. All changes to XML must be synchronized with consts.py (also some of the WB such as map size and independent cities and order of the players), otherwise prosecutors will act weird, wrong churches would be build and so on.

Also is anyone doing C++ editing or just python + XML + WB.

 

[3Miro]

I read the files tread and I see that someone is doing C++. That is fine, however, we need to fix the prosecutor AI problem and not ignore it. I will need to look at exactly what the crash is.

- I will (when I get back) need the source code and/or exactly what changes have been made.
- If anyone knows exactly why the crash is happening, please share the technicality (will save me time).
- A savegame right before the crash would be useful.
- before anything, I guess make sure consts.py has properly set up the number of religions and the index of the prosecutor unit is set correctly (in RFCEBalance, there was a function in the "don't change this" part of the code that sets the prosecutor, number of religions and I don't remember the third one).

 

[Arwon]

I've uploaded a save from 1000ish, it was Orthodox from when I started in 700 I think.

 

[sedna17]

Also in that Cordoba game, Arabia began as Orthodox...

I've sometimes seen Arabia switch to Orthodoxy, probably because Arabia takes over a whole bunch of Orthodox cities early on, and so has roughly equal number of Orthodox/Islam cities. Still, having the holy city should normally tip the balance... This plus the wrong protestant shrine plus the array crash for the AI prosecutor code clearly speaks to a building index problem between XML and Python. I'll double-check everything again.

 

[jessiecat]

I've sometimes seen Arabia switch to Orthodoxy, probably because Arabia takes over a whole bunch of Orthodox cities early on, and so has roughly equal number of Orthodox/Islam cities. Still, having the holy city should normally tip the balance... This plus the wrong protestant shrine plus the array crash for the AI prosecutor code clearly speaks to a building index problem between XML and Python. I'll double-check everything again.

I agree. I've seen Arabia become a vassal and switch religion. I've looked at Arwon's game and it looks like Islam was founded in Dimashq normally but only spread to Syroz but no further. And any new city founded by the AI became Orthodox. Which means the automatic adoption of Islam with a pre-built mosque just didn't happen. Something wrong there. At least Al Quds should have become Muslim from the flip.

EDIT I've posted the revised Spain/Portugal CityName map to the files thread.

 

[jessiecat]

I've uploaded a save from 1000ish, it was Orthodox from when I started in 700 I think.

Thanks for that. I've had a look but I can't tell when it it switched.Though I can see why. I've also looked at your city which became the equivalent of Pamplona instead of Zaragoza. If it was founded on either tile south it would have been Zaragoza but Pamplona is marked on the 4 tile square so anything on those tiles would be Pamplona, for the reasons I described earlier. I'm revising the Spain map today to make it more compatible with Cordoba so Guadalajara won't become Burgos again. And I'll look at the Zaragoza area again too. Thanks for your observations. Hope to see more of you on this project. Your help would be greatly appreciated.

 

[merijn_v1]

I agree. I've seen Arabia become a vassal and switch religion. I've looked at Arwon's game and it looks like Islam was founded in Dimashq normally but only spread to Syroz but no further. And any new city founded by the AI became Orthodox. Which means the automatic adoption of Islam with a pre-built mosque just didn't happen. Something wrong there. At least Al Quds should have become Muslim from the flip.

The UP of Arabia is: STATE Religion spreads to every city with a TEMPLE. If Arabia was Orthodox, than should spread Ortodoxy to all of his cities. And if that happened, there is nothing wrong I thingk. And there only comes a pre-built temple and no pre-built mosque in the cities. So I think that should be o.k. too.

 

[st.lucifer]

A couple of notes on late spawns:

Started a game as Sweden, where the Norse had settled much of Sweden proper (including the area around Stockholm - I ended up being pushed off of the normal starting area).

I waited to found a capital, as I knew I'd flip Vasteras and another city, and moved workers, settlers, and units to future city sites through open borders. The flip came, and the Norse DOW'd. While all of the units in the two cities that flipped came with them, all of my units within Norse borders but not in cities disappeared. Checking the log, they weren't captured or killed, and they didn't get auto-moved into my new cities - they were simply wiped out. I thought this might be due to my lack of a pre-existing capital, but when I built a capital the turn before the flip, there was no DOW, so I couldn't really test it properly.

The second issue is minor, relating to mercenaries - in the second attempt, with the capital founded before the flip, I ended up with more units than I needed, and put some up as mercenaries. They immediately were taken off of my maintenance rolls, and were hired by Austria in the next turn. I received the cash hiring bonus, but no GPT for the units which had been hired. Having achieved that early success, I tried hiring out more units, but found the mercenary screen empty - I couldn't hire or hire out units. I'm not sure if this is done in standard RFC, but when my units were hired, they were renamed as barbarian units - Jutes and Angles, I think. Further attempts at interaction with the mercenary screen yielded the same result.

 

[jessiecat]

Just finished a great game as Genoa. Good news that I got to 1700 with only crashing once (reloaded autosave and carried on). Bad news is that its now stuck in 1700 with "waiting for other civs". Got all the UHV conditions too including 11 OBs. So I'm counting that as a UHV victory anyway. Couple of questions though.
1. Why is there only 60 turns to go in 1700? Does the mod end in 1760 instead of 1800?
2. Why do some obsolete units remain on the build list, like Macemen, Knights, Trebuchets and Catapults, when you can already build Line Infantry, Grenadiers, Hussars, Cuirassiers and Cannons?
3. A lot of the late techs have no function. I thought you could build hospitals once you researched Arabic Medicine and Medicine. Same thing with Combined Arms and Civil Engineering.
4. Can't the AI civs change civics? I'm sick of being asked to adopt Paganism, esp. when I'm running Merchant Republics, Common Law, etc.

EDIT 5. The human player can win a UHV victory easily as long as they get a lot of OBs fairly quickly esp. Byzantium. I could have had all the conditions by 1550 if I really tried and still had them by 1600. I'd suggest bumping everything back by 100 years you can win at 1600AD.

Anyway here's some screenshots and the saved game stuck at 1700AD.

 

[merijn_v1]

When I played the Dutch, I saw that the spawn-date isn't correct. The civilclopedia says: 1580, but the actually spawn-date is 1540.

 

[3Miro]

When I played the Dutch, I saw that the spawn-date isn't correct. The civilclopedia says: 1580, but the actually spawn-date is 1540.

There is probably a problem with the translation of the year to specific turns. I will look at it (eventually). It is probably related to the 1760 bug Jessicat reports. Everything is in a game speed XML file and the WB. The numbers from those files need to be in sync with the spawn dates from python.