Unofficial Patch for 3.19

[ruff_hi]

here is a bug report for v319 ... http://forums.civfanatics.com/showthread.php?p=8173407#post8173407

It involves PBEM and the use of the altroot command line.

 

[GravityWave]

Could the knights in the stack have attacked, and then the remaining musketmen and pikeman been killed in a counter attack? This is all possible during the inter-turn.

According to the event log they died attacking my grenadier (screenshots attached) they also appear to attack from the forest but that could just be screwy graphics? Not sure what you mean by counter attack(?)...

 

[EmperorFool]

Regarding the Korean SoD, did they perhaps lose Open Borders with Russia and get kicked out . . . next to your city?

 

[r_rolo1]

I gave a somewhat detailed look to GravityWave save and IMHO there is definitely a bug in here....

First things first: This is a game with 18 civs, all of them alive ATM besides Babylon, that was player #6 ( you can check this easily ). The human player is Bizantium, that is player #4. The players before #4 already played ( including England, the master to be of America )....

When the human finishes its turn in 1714, the game goes to player #5 ( Russia ), skips #6 ( Babylon ) because it is dead, passes to #7 ( Egypt ) and so on ( I had checked friendly and enemy moves to keep track of all ).... all is normal until the ( BUG ) warning that England founded Dover and the diplo window of Lincoln ( that gets to be player #6 instead of Hammurabi of Babylon ), followed by the Rifling tech window....

After that the game shows Egypt, ... , moving the units again and it is in this phase that the flying pikeman attacks the city and the korean Speedy Gonzalez gets to that forest. This can be checked by the Event log, that shows the mongol pike attacking in 1716, before the human move , a thing that shouldn't happen because Mongolia is suposed to play after Bizantium by the player order.

Seeing this I decided to mark some units of various civs to check for double movement. All of the tested civs showed it, except Russia ( player #5 )....

The conclusion I can take from this ( that will require obviously code check ) is that the game started the year 1716 playing with the newlyfounded colony ( player #6 ) instead of starting with player #0 ( England ) and that continued the order until player #17, resuming to #0,1... until the player #4 , the human, giving a effective 2 turn period in the IBT to all the AI except for poor Russia

 

[GravityWave]

Regarding the Korean SoD, did they perhaps lose Open Borders with Russia and get kicked out . . . next to your city?

According to the diplo screen they have Open Borders - I guess it is possible for Peter to of cancelled the agreement and then Wang Kong resigned it, and done the same with Genghis which kicked out all the units in to the same square, enabling the Mongols to attack Adrianople.... I have no idea how to check/test this though...

Just read r_rolo1's post, and I'm sure that's the problem. Thanks for looking in to it... at least it's an improvement over the living dead civs mucking up any chance of a diplo victory... though I can see a lot of times when this could be a game ruining, if not breaking bug.

 

[r_rolo1]

It was not that what happened, because in that case the mongols would had attacked in 1714 and not in 1716 as stated by the events log.

 

[MadmanAtW]

So it sounds like there's a problem in 3.19's fix for spawning new colonies in games that have all the AI slots filled?

 

[r_rolo1]

I would definitely say so ......

GravityWave, maybe it would be better to post something in the bugs forum. The firaxians are scanning that forum regularly ( even now alexman is seeing it )

 

[EmperorFool]

Did 1716 happen twice for all the AIs #6 through #18 or did #0 through #5 get skipped in 1717? My guess is that when spawning the colony there is some code that allows the new colony to move in that turn that assumes they will get the last slot. When they take an existing slot, everyone after them gets another turn.

 

[r_rolo1]

I haven't tested what happened to all the civs .... But I think that I seen a portuguese unit ( portugal plays before bizantium ) moving.... I'll test it as soon as possible.

P.S That can't be... if that had happened, the mongols would had attacked in 1714 and not in 1716, I think....

 

[r_rolo1]

I rechecked and indeed it looks that the players #0-3 ( England, Netherlands, Portugal and rome, by that order ) have only played 1 turn in the IBT as expected.....

So a chronology:

-1714, end of turn for player #4 . Players 0-3 already played.....
-1714, IBT, usual moves for the alive AI from #5 to 17
-1714, Colony is created and it has slot #6 assigned
-1716 - Colony plays once ( queues confirm that .... btw didn't knew that colonies also received the initial 10 hammers as the other AI in the start )
-1716 - AI from #7 to 17 play, then the ball passes to #0 and from there to #3. In this period the Mongols attack Adrianopole and the koreans get to that forest. This makes the AI that ocuppy the slots between 7 and 17 to have played two turns in the human IBT. Players #0-3 and 5 only played once ( players 0-3 probably already in 1716 and 5 in 1714 )

 

[EmperorFool]

So does the year advance twice, skipping players 0-4? Or do players 6-17 get two moves in 1716? Towards the end of your chronology, you stop showing the year so it's unclear.

Also, what player # is England who spawns the colony? If they aren't #17 (last player), do the AIs between England and the barbs get two turns or do they get skipped during the first normal set of turns?

Finally, do the barbs get two rounds during this hiccup? They are player #18 here, right?

 

[r_rolo1]

Ok.....

England is player 0 and already played in 1714 when the human finishes it's turn. Player 6 ( the colony ) only plays once ( not sure if in 1714 or 1716 ( turns 752 and 753, respectively ... this is marathon ) ). Players 7-17 play once in 1714 ( after the human ) and another in 1716, this one before the human turn ( that is player #4, remember ) instead of being after, making them to play 2 turns in the human IBT . Players 0-3 play one turn in the human IBT ( as they should ), as well as player 5 ( also as it should be ). Haven't checked barbs....

Between 1716 and 1718 ( the next turns ), everyone plays only one turn in the human IBT, btw.... I would say that the colony creation messes up the play order...

I might do some more testing later, but there is a unrelated detail that stroke me: the colony has the artstyle and the flag of the dead civ ( babylon ), but not the color, that is the regular american blue. Wierd.....

 

[MadmanAtW]

Wait, so, if you keep playing out the game from there, in the end everyone will have taken the same total number of turns- only the turn order was rearranged?

 

[r_rolo1]

I would say that it looks that way....

 

[EmperorFool]

Since England is player #0, the colony had to be created in 1716--not 1714. It sounds like when England split the colony, the "current player" jumped from #0 to #5--skipping #1-#5.

After players #6-#17 get their turn in 1716--the second turns for them during the single IBT for the human, it goes to 1718 where players #0-#3 play and then it's the human's (#4) turn.

It sounds like players #1-#5 got skipped in 1716, so not everyone gets the same number of turns. I wonder if this happens when the new colony isn't taking a previously-played slot.

 

[r_rolo1]

The players 0-5 are not skipped in 1716. Players 0-3 already played in 1714 and they have a extra played turn in the IBT ( in 1716, most surely ). Player 5 plays ( in 1714 ) after the human and only plays once in the IBT and has one turn played in the 1716-1718 IBT ( meaning that it played in 1716 ). The Human ( player #4 ) obviously play in both turns.

 

[EmperorFool]

The players 0-5 are not skipped in 1716.

Player #0 (England) surely plays in 1716 because they spawned the colony (#5), right?

Players 0-3 already played in 1714 and they have a extra played turn in the IBT ( in 1716, most surely ).

They are supposed to play in 1716, so that's not extra. They're turns during 1714 are before the human's--not in the IBT.

Player 5 plays ( in 1714 ) after the human and only plays once in the IBT.

Since player #5 is created by player #0 in 1716, surely it plays its turn in 1716 as well.

The Human ( player #4 ) obviously play in both turns.

Which turns do you mean by "both" given that there seem to be three turns involved here: 1714, 1716, and 1718.

This is what I'm thinking to be the timeline, but it's getting quite confusing.

1. 1714 starts
2. Players 0-3 play
3. Player 4 plays -- HUMAN TURN
4. Dead player 5 skipped
5. Players 6-17 play
6. Player 18 plays -- BARB TURN
7. 1716 starts
8. Player 0 frees a colony; player 5 created anew
9. Players 1-3 play -- DOES THIS HAPPEN?
10. Player 4 skipped -- HUMAN SKIPPED
11. Players 5-17 play
12. Player 18 plays -- BARB TURN
13. 1718 starts
14. Players 0-3 play
15. Player 4 plays -- HUMAN TURN

The bold parts are the ones in question, and the second one (player skipped) is the problem. Is this timeline accurate?

 

[r_rolo1]

Player #0 (England) surely plays in 1716 because they spawned the colony (#5), right?

Colony is #6 . Besides that it is correct ( England even enters in anarchy )

They are supposed to play in 1716, so that's not extra. They're turns during 1714 are before the human's--not in the IBT.

I never said they were extra turns ( in fact , I'm getting increasingly convinced that technically there are no extra turns played by anyone ) And technically the IBT means "in between (human) turns"

Since player #5 is created by player #0 in 1716, surely it plays its turn in 1716 as well.

Like I said above the colony is player #6. Player #5 is Russia

Which turns do you mean by "both" given that there seem to be three turns involved here: 1714, 1716, and 1718.

I meant 1714 and 1716 in this particular regard. Sorry for the confusion. Human plays all the turns involved.

This is what I'm thinking to be the timeline, but it's getting quite confusing.

1. 1714 starts
2. Players 0-3 play
3. Player 4 plays -- HUMAN TURN ..... here is where we are in the posted save
4. Player #5 ( Russia ) makes his move
5. Dead player 6 skipped
6. Players 7-17 play
7. Player 18 plays -- BARB TURN
8. 1716 starts
9. Player 0 frees a colony; player 6 created anew Not sure.... Player 6 has 1 turn of prod in queues before the human play in this turn
10. Players 1-3 play -- DOES THIS HAPPEN? YES, but not necessary in this point. We have no way of accurately knowing if they play before or after player 17 or not without going to debug ( a thing I should already had done )
11. Player 4 skipped -- HUMAN SKIPPED NO . Human plays in 1716, after players 6-17
12. Players 6-17 play
13. Here or after the next point, the human ( player #4 ) and player #5 ( Russia ) play
14. Player 18 plays -- BARB TURN
15. 1718 starts
16. Players 0-3 play Also players 6-17
17. Player 4 plays -- HUMAN TURN

The bold parts are the ones in question, and the second one (player skipped) is the problem. Is this timeline accurate?

Red text is my comments/corrections

 

[EmperorFool]

# 1718 starts
# Players 0-3 play Also players 6-17
# Player 4 plays -- HUMAN TURN

So it seems that the human player #4 gets permanently moved to the end of the queue. I really don't see how this could be given the game mechanics of looping over players, but I'm not very familiar with this area of the code.