Clarifying OOS solutions

[Calbrenar]

My friend and I installed just the base mod (from gameflood) not sure what version of the patch that is.

We got continual OOS errors from 1st turn on until we installed the unofficial 3.13 patch.

Played 290 turns so far and no OOS errors.

Not using the media pack or any extra FFH patches other the nthe base install.

 

[Kentarion]

---- Delete ----

 

[Kentarion]

with 'NO EVENTS' enabled:

OOS comes when armageddon counter hits 40. (every round)

now going back to patch 'f' - hope it's playable with that.

 

[cephalo]

I'd like to point out something that I haven't seen in this thread yet. From the very beginning of map generation til the end of the game, in every case the random numbers generated have to also be generated at the same time on all of the clients no matter what. The numbers are generated in a pre-determined sequence, so if one player generates a number while the others do not, even if the result has nothing to do with them, the fact that the individual sequences are not in line will break the sync.

So, for the sequence:
1,2,3,4,5,6,7

player one and player two using the sequence for combat and events:
player one: result 1 result 2 norandevnt result 3 result 4 result 5 result 6
player two: result 1 result 2 randevent 3 result 4 result 5 result 6 result 7

Even though player one has no random event, to remain in sync he must also generate a 'dummy' random number that is thrown away. This can and should be handled automatically by the game, but from what I've seen in the dll code, no attempt is made to syncronize the raw random numbers.

Hope that clarifys how random numbers can effect things.

 

[Kentarion]

Played a game with PATCH 'G' to turn 240.

however since it was to boring because of early wins - we stopped it (no armageddon counter up to 40 etc...)

 

[Kael]

I redid the way the armageddon events occur in patch "i" to try to avoid OOS's. I am plannign on getting some OOS debug out this weekend, but it will be on the "i" codebase. So anything you guys can tell me about OOS activity on "i" would be greatly appreciated.

 

[Kentarion]

I redid the way the armageddon events occur in patch "i" to try to avoid OOS's. I am plannign on getting some OOS debug out this weekend, but it will be on the "i" codebase. So anything you guys can tell me about OOS activity on "i" would be greatly appreciated.

Patch "i" (Me Win XP, Friend Vista):

Played until Round 200 with no OOS happend. Can't say much about a-counter issues since it's just at about +7. We will play it further un sunday - update will come. Good work for now I LOVE EVENTS

 

[shrike2005]

We played for about 200 turns today, on patch "i" with no oos errors or any other errors. However we had the armageddon counter disabled.

Also i really like the new events. So i would like to say that this mod is the main reason i play civ4 and so a big thank you to everybody involved. This is truly an impressive thing you have achieved.

 

[onedreamer]

You should not take into account internet games if you want to find a real issue with OOS, you should test it in LAN because in internet OOS can easily happen due to ... internet.

 

[ShadowDrgn]

TCP ensures packet delivery and order so you can't blame the Internet for your games going out of sync. It's no worse than playing on a LAN since ping is irrelevant in a turn based game.

 

[ushram]

/bump

good one ShadowDrgn

 

[Lt_Kain]

Hello,

We just discovered this mod (we were waiting a warhammer mod ... and yours seems to be the most stable/advanced).

We are playing with the G patch, with the i patch, we just had a permanent OOS.

We are 4 players (3 in europe, one in Japan ) 3 XP and one vista and all his working fine.. an OOS happend sometime but the bad one just deco/reco and its good.

A good debug to the oos problem is to activate the logs and see the different actions made before the oos (barbarian attacked unit 1, etc). It is very easy to see what caused the OOS most of the time.

 

[Kentarion]

Patch 'j':

OOS is back.

Played to Round 180 - then every round OOS appears.

great trader turned after loading to great engineer - so thats where this oos comes i think.

next oos appeared with hunter event. i took a tiger after loading i had no animal.

next oos: council of esus support: i choose resistance to magic for recontype (didn't receive after loading...)

stopped the game.

started new game - ROUND 4 OOS. STOPPED playing. - waiting for patch 'K'.

 

[Seven05]

Odd that the previous poster complains of OOS with patch J and I was just positng to praise it as being OOS free

2 Player LAN game, large 'Creation' map, 9 civs patch J. Played for just shy of 500 turns without a single OOS or crash.

AC topped out at 29 I think so I can't say much for AC related issues.

For races in the game we had Hippus (me), Grigori (other player) and then AI controlling the Sidar, Bannor, Lanun (sp?), both elven civs, the Illian or whatever they're called and the Amurites. Sorry, I'm not so good with all of the civ & leader names. The barbarians wiped out three of them relatively early and I took out the Bannor about 300 turns in while my friend destroyed the Amurites about 50-75 turns later. Sidar created the gate and brought the Mercurians (or whatever they're called) into the game but the infernals never made an appearance. As far as I could tell their weren't any members of either council, I'm pretty sure all of the AI were still in 'no membership' when we quit and both of us were running no membership. Once the Bannor were eliminated all of the remaining civs were neutral, hense the rather stagnant AC which was really acting more as a counter of the number of cities razed by the barbarians.

We had a few religions widely spread, FoL and RoK being the most prominent with OO for the Sidar and Lunan, all of them were founded but not exactly active.

The majority of the combat was with the barbarians excluding the two wars brought on by the two of us. Most civs were at least cautious or better with everyone else.

We had events enabled, using the living world, raging barbarians and wil lands options. Most of the events we saw in our game were related to the guilds with a few for free forges, calibam refugees (they weren't in the game though), burning witches, rich guys dying and leaving us money and dedicating the obelisks to various gods. We also had the Dragon pop into the game early on, Orthos as well and then some giant 22 strength monkey. I was able to capture multiple animals (including the big monkey) without issue.

I tried just about every spell I could manage, which ended up being most of them since I had almost every type of mana. Both of us used our world spells but I don't recall ever seeing any of the AI players use theirs.

So, that's everything we did for the most part and none of those caused any OOS issues

 

[Kentarion]

ok, that's interesting, do you have different systems?

my friend has vista, me xp - (that's where most issues came)

we played a small map.
more events and animals enabled.
i played svartalfar friend ljosalfar. (difficulty prince)
my religion was CoE his was FoL.
other races were: amurite (killed them about turn 50), kazhad, luichurp, clan of embers.
however i don't think this all will have to do anything about it.
as we started a new game we got OOS @round 4 - no event occured then (not with me & him - but maybe the AI had one.) however i remeber some strange event with Faeryl Viconia wich happens exactly always turn 4. I didn't have that event...

we'll try patch 'k' tonight.

we (me & friend) feeling like testers for multiplayer here that's ok, i'd like to help getting it functional for all

 

[Goodgimp]

we (me & friend) feeling like testers for multiplayer here that's ok, i'd like to help getting it functional for all

Considering you're playing a use-created mod versioned at 0.30, I think being a tester should go without saying.

 

[Seven05]

ok, that's interesting, do you have different systems?

my friend has vista, me xp - (that's where most issues came)

...

we (me & friend) feeling like testers for multiplayer here that's ok, i'd like to help getting it functional for all

Yes, I'm on Vista64 and he was using my wife's PC which is using XP. Prior to patch J the only thing that worked for us was reverting to patch E. If you're still having problems and we didn't I'd suspect it is related to an event that didn't trigger for us or a civ that we didn't have in our game. If we do it again next weekend I'll try it with all 19 civs in and maybe go through the events xml and force them all to occur in our game.

 

[SwordofStriker]

I just played roughly 1040 turns into a LAN game before having an OOS problem, but what a problem it became.

Both patched to J
19 civs
Creation mapscript
Standard size
2 human players
Living World
Marathon speed
Both XP machines

At about turn 1040 we had the first OOS. Every few turns thereafter we had another. Sometimes one player could relog and the problem would get fixed, other times the game had to close, and then a save was loaded to fix the problem. Finally, we starting reaching this one turn in particular that the game went OOS on every single time, and the game became unplayable. Victory was in my grasp and now, disappointment.

 

[onedreamer]

TCP ensures packet delivery and order so you can't blame the Internet for your games going out of sync. It's no worse than playing on a LAN since ping is irrelevant in a turn based game.

Yes I can
The game doesn't work at the Transport layer, it works at Application layer. Do you know how does it pass data to Transport layer and how much does it wait in case of latecy (what you call "ping") ? Latency is always important in the internet including turn based games, besides Civ's multiplayer on the internet is almost always simultaneous.

 

[Roghar]

Played patch K for a few hours last night, no OOS. Looks like some of the changes in J and K have gotten rid of problems from F - I. Hope it wasn't just good luck last night, its so much more enjoyable when it behaves!