HOF Autolog timeshifts

da_Vinci

Gypsy Prince
Joined
Jun 13, 2004
Messages
4,182
Location
Maryland, USA
Hi all,

I am wondering when dates of events are posted in spoilers, if we are all talking the same time given the way the autolog shifts time, compared to what a manual record would have: (here is a test I ran)

The autolog has:

<b><u>Turn 0 (4000 BC)</b></u><br> (note that this line is html format, as I changed to forum style after the game started)
Berlin founded
Berlin begins: Worker
Research begun: Animal Husbandry

Turn 1 (3960 BC)
Tribal village results: a little gold (right date, wrong turn number vs. manual {turn 2})

Turn 2 (3920 BC)

Turn 3 (3880 BC)
Tribal village results: lots of gold (ditto)

Turn 4 (3840 BC)
Berlin's borders expand (this is more interesting. Manual is turn 6, 3800 that I can actually USE the fat cross. Berlin expanded at the END of date 3840, but isn't what matters when I can use it?)

Turn 5 (3800 BC)

Turn 6 (3760 BC)
Contact made: Russian Empire (manual gives me turn 7-8 transition, 3760 to 3720 transition for this. Again, met at end of 3760, but cannot USE that until 3720)

Turn 7 (3720 BC)

Turn 8 (3680 BC)

Turn 9 (3640 BC)
Tribal village results: a little gold (right date, wrong turn number vs. manual)
Hinduism founded in a distant land (manual is turn 11, 3600 that it is announced)

Turn 10 (3600 BC)

Turn 11 (3560 BC)
Tech learned: Animal Husbandry (manual is turn 13, 3520 that is is announced and usable)
Contact made: Persian Empire (manual is turn 13, 3520 that is is announced and usable. An INTER turn event)

Turn 12 (3520 BC)
Research begun: The Wheel (manual is turn 13, 3520)

Turn 13 (3480 BC)
Contact made: Roman Empire (manual was turn 14, 3480. An INTRA turn event)

Turn 14 (3440 BC)
Berlin finishes: Worker (manual has turn 16, 3400)
Buddhism founded in a distant land (manual has turn 16, 3400 as time announced)

Conclusions: I would like to see 4000 called turn 1 (not turn 0) so turn numbers match the years as displayed in the game.

Some events happen between turns: builds, research discoveries, meet neighbors on their moves. Autolog pins this to the previous turn date, even though it only affects us in the subsequent turn.

Some events happen during our turn: meet a neighbor on our move, and combat results (I think). Autolog puts them in the proper year, but of course the turn number is off by 1 vs a manual log.

I would prefer to see the autolog pin dates on the events that match when we can use them: for example, date of archery discovery is the date that we can start making archers. But maybe that is not a feasible change?

dV
 
You can choose (in the HoF mod) whether 4000 BC is displayed as turn zero or turn 1, so the fact that it displays that way in your game is your choice. It's a setting on the HOF2 Options tab. I don't think that setting affects the autolog output because the two mods were designed by two different people. However, as long as you choose to define 4000 BC as turn 1 in your manual records you'll be fighting against the software.

Internally, Civ4 calls 4000 BC turn zero, because it was designed by programmers, who count from zero. In the past it has been much more straightforward to use that convention. Particularly in Civ3 SGs (succession games) the round numbered turn numbers only line up with round numbered years if you start counting at zero.

Lots of stuff happens between turns. You learn archery in between turns, not during one turn. It happens after the end of one turn and before the start of another, when you can start to use it. You'll never fully reconcile turn, date and event.
 
AlanH said:
Lots of stuff happens between turns. You learn archery in between turns, not during one turn. It happens after the end of one turn and before the start of another, when you can start to use it. You'll never fully reconcile turn, date and event.
The issue is whether it makes more sense to assign the prior turn or the subsequent turn to events that happen on the fence. For example, if I build my last spaceship part in between 2010 and 2011, it is a 2011 space race victory. The autolog would say that I built the part in 2010. Just seems a bit inconsistent to me.

So I guess when we see dates in the spoilers, we don't know for sure if we are seeing the dates before or after boundary events, unless the poster specifies this.

dV
 
I guess we've managed to live with that level if uncertainty about players' reports for many years now. If it's really critical to you then you can now download a player's replay file and load it up to read the in-game log yourself.
 
da_Vinci said:
The issue is whether it makes more sense to assign the prior turn or the subsequent turn to events that happen on the fence. For example, if I build my last spaceship part in between 2010 and 2011, it is a 2011 space race victory. The autolog would say that I built the part in 2010. Just seems a bit inconsistent to me.

So I guess when we see dates in the spoilers, we don't know for sure if we are seeing the dates before or after boundary events, unless the poster specifies this.

dV
All builds and research happens at the end of a players turn in civ4 (in civ3 it happened after all players had ended the turn). This is why if you have a defender completing in 1 turn in a city, it will complete and defend before another civ can attack you (unless playing simultaneous turns)

Victory conditions are checked at the beginning of the players turn.
 
awww, I was gonna say that lol

anyway, yeah...anything that happens 'in between turns' for your civilization happens when you click 'end turn.'

The closest you can get it lined up the way you want is to click the HOF option that makes 4000BC turn 1. I'm a programmer myself, so I generally count things from turn 0, but in Civ, I'm playing to get away from all that programming homework(programming is a lot funner when it isn't being graded by a teacher) so I start it at turn 1 like I do anything else non-programming related.

*shrug* I understand you not liking that the turn you learn archery according to the autolog is not the turn you can build archers, but its the same for everyone, so I don't think it makes much of a difference overall.
 
AlanH said:
You can choose (in the HoF mod) whether 4000 BC is displayed as turn zero or turn 1, so the fact that it displays that way in your game is your choice. It's a setting on the HOF2 Options tab.

Lots of stuff happens between turns. You learn archery in between turns, not during one turn. It happens after the end of one turn and before the start of another, when you can start to use it. You'll never fully reconcile turn, date and event.

Is there any reason the mod can't have the same option for events that it does for zero vs. one and let the player choose whether it records a tech discovery the turn before or the turn after?
 
Alraun said:
Is there any reason the mod can't have the same option for events that it does for zero vs. one and let the player choose whether it records a tech discovery the turn before or the turn after?
It would be and extremely simple change to use the same options for both. BUT, it could get confusing when trying to compare games. Everyone would have to indicate not just their version of the save but clock setting. (i.e. "Contender, 4000BC=0"; "Challenger, 4000BC=1", etc.)

Autolog uses the "onEndGameTurn" event to write the Turn header line for the next turn. It uses the turn number + 1 to get the date for the next turn.
Code:
def onEndGameTurn(self, argsList):
	...
	self.turn = gc.getGame().getGameTurn() + 1
	self.year = gc.getGame().getTurnYear(self.turn)
I haven't checked, but I think that the player log and autolog probably agree on the turn number/year of the events. If they didn't than that would be an arguement for changing.
 
Gyathaar said:
All builds and research happens at the end of a players turn in civ4 (in civ3 it happened after all players had ended the turn). This is why if you have a defender completing in 1 turn in a city, it will complete and defend before another civ can attack you (unless playing simultaneous turns)

Victory conditions are checked at the beginning of the players turn.
@Gyathaar: I did not know about builds and research happening at end of my turn but before start of the same turn for some other players. In that case, assigning the event to the turn of completion (rather than the turn a tech can be used or a unit moved) makes perfect sense. I withdraw the complaint about the "on the fence" issues. And I can fix the turn 0 stuff if I wish.

It does raise the question of whether a single human player always moves first compared to the AI, or might some AI move before, and some AI after?

dV
 
da_Vinci said:
It does raise the question of whether a single human player always moves first compared to the AI, or might some AI move before, and some AI after?
In standard single player games the human is player number 0 and goes first..
However, in scenarios where you select what civ to play, the civs go in the order the selection list is shown
 
Back
Top Bottom