Taking a break...

[treespider]

Any work arounds for serious syncing issues...I am about to decide whether to stick with WitP or computer games in general. Jim Burns and myself have once again encountered serious syncing issues. I run our turns after a fresh start of WitP and we have started having syncing issues after all of 5 turns. I do have two other PBeM games going under this same install but I always start each PBeM game's turn with a fresh start of WitP.

All of the games are being run under v1.602 but IRC the syncing issue was present before 602.

This game takes way too much of a time investment to find out four turns into a PBeM that it is busted. I have pretty much come to the conclusion that computer wargaming sucks. There is no going back and having a do over as there is with good old cardboard and paper. I have yet to find a computer game without game-ending bugs. What's the sense in running a PBeM game when the two plauyers aren't even playing the same game...

[ChezDaJez]

Don't know what to say. I have never experienced this problem in my PBEMs.

I assume you guys are using the same version, database and everything.

Chez

[treespider]

ORIGINAL: ChezDaJez

Don't know what to say. I have never experienced this problem in my PBEMs.

I assume you guys are using the same version, database and everything.

Chez

Yep...

V1.602
CHS 1.06


Last turn as an example ...the combat replay shows a strike by 27 Nells on the PoW and Repulse scoring a torpedo hit on the Repulse that never happened, a torpedo hit on CL Boise that never happened, the sinking of a japanese submarine that never happened and the capture of Manado that never happened.

I might add that i never witnessed any of these as the Japanese. The only reason I picked up on it was that Jim noticed it the difference between his combat replay anbd the combat report.txt file I sent. So when i reinstalled the combat replay save and let it run it provided a different result.

Maybe something is screwy with my machine....I'd hate having to shut down my virus software and firewalls and spyware and whatever else just to play a game.

[jwilkerson]

ORIGINAL: treespider

Any work arounds for serious syncing issues...I am about to decide whether to stick with WitP or computer games in general. Jim Burns and myself have once again encountered serious syncing issues. I run our turns after a fresh start of WitP and we have started having syncing issues after all of 5 turns. I do have two other PBeM games going under this same install but I always start each PBeM game's turn with a fresh start of WitP.

All of the games are being run under v1.602 but IRC the syncing issue was present before 602.

This game takes way too much of a time investment to find out four turns into a PBeM that it is busted. I have pretty much come to the conclusion that computer wargaming sucks. There is no going back and having a do over as there is with good old cardboard and paper. I have yet to find a computer game without game-ending bugs. What's the sense in running a PBeM game when the two plauyers aren't even playing the same game...

There is a "fix" it doesn't work forever - but it works for a while - there are many theories about what causes it to break - but it has been around from the beginning and is still present.


The fix is first to ensure the Allied player has Combat Reports turned on.

Then assuming the Japanese player has Combat reports turned on ... the Japanese player turns off Combat reports .. processes the turn ... sends to Allied player ( no combat reports this turn ) ... the Allied player processes and sends back .. then the Japanese player turns CR back on and processes the turn ...

This should "reset" the synch issue ... after that .. it may recur ... but can be dealt with by repeating this process - or sometimes it clears itself up.

According to many of the theories - it is best to play one game per instance - or at least not to play Allies and Japanese in the same instance ( depending on the theory ).

[treespider]

According to many of the theories - it is best to play one game per instance - or at least not to play Allies and Japanese in the same instance ( depending on the theory ).

Do you mean one game per install?

[treespider]

I think i'm going to have to send my wp.prfs file with each game. I noticed that it updates itself each time I run a turn.

[Captain Cruft]

I get the out of sync problem regularly. My thoughts:

Fact: The syncing issue only ever effects the Allied player - the Japanese player always sees the reality since the execution phase is "generated" on his PC.

Postulation I: Apart from patch installations what causes this to occur is ships getting badly damaged (by any mechanism). This may then cause the "DIVIDE CRIPPLED TFS" phase to mess up and take a lot longer than usual. I notice this because I have a relatively slow PC and it goes from c.10 seconds to c.3 minutes ...

The "DIVIDE CRIPPLED TFS" phase is where the computer is forming up new "Escort TFs" out of the badly damaged ships and any spare DDs that are to hand etc.

Postulation II: It may be something to do with ship or location (TF) slots that is the root cause. I have seen it happen more often in CHS than in vanilla.

Observation: The good news is that it always seems to self-correct 2 or 3 turns later. When the problem does occur I just settle for reading the combatreport.txt (sent by my opponent) for a couple of turns instead of watching the movie.

[Andrew Brown]

ORIGINAL: treespider
Do you mean one game per install?

This was a "last ditch" fix I was about to suggest, although it may not prevent it either, I guess...

[Captain Cruft]

If you are going to try manual wp.prfs synchronisation it needs to be the Japanese player's version that is the "master". The game runs on the Japanese PC, the Allied PC is just an observer apart from the order entry.

[treespider]

ORIGINAL: Captain Cruft

If you are going to try manual wp.prfs synchronisation it needs to be the Japanese player's version that is the "master". The game runs on the Japanese PC, the Allied PC is just an observer apart from the order entry.

Yes I know....as i've sent this message to my PBeM opponents.

The synch problem is:

The witp001.pws file that I send you when replayed does not correspond to the combat that the japanese player sees. Nor will it mesh with the combatreport.txt file.

I believe the work around will be to send my wp.prfs file with each zip, for each game, every time. I noticed that my wp.prfs file seems to update each time I run a turn from a different game.

I am the Japanese player. The reason I think it is tied to the wp.prfs file is when I ran the combat replay from mine and jim's game the prfs file had already been updated and the combat replay was different from the one that I had witnessed last night. In addition, when we had a game running a couple of months ago this exchange seemed to correct the problem and IRC i only had one game running at the time.

[jwilkerson]

ORIGINAL: treespider

According to many of the theories - it is best to play one game per instance - or at least not to play Allies and Japanese in the same instance ( depending on the theory ).

Do you mean one game per install?

Yup ...

Currently, I'm not following this religiously .. but I am following the only play one side or the other per install ...

I find that some instances of the problem "self-correct" after a few turns ... other do not ... I've seen the problem start due to upgrades ( self correcting ) ... and firing up another game as head to head ( not self correcting ) ... flipping the combat reports switch also supposedly matters .. but as you can see others even have other theories ! There is also the "PT Boats just fought a battle" theory ...

But if it doesn't self-correct then use the "POMPAC" method I described above .. that has worked many times for me and others ... it doesn't does work forever ... eventually no matter what apparently .. the problem will re-occur ...

[Captain Cruft]

Yep, I don't this problem will ever go away. I could rant on about the totally bizarre design but I think I did that already some time ago ...

Someone should try reverse-engineering the wp.prfs file. It's about the only mystery left Unfortunately I think that it is at least partially duplicated in the game save file, which then over-writes wp.prfs when it is loaded.

[Jim D Burns]

Treespider sent a PM to his PBEM opponents and a thought occurred to me. Here's a copy of his PM and my reply. Would some energetic Japanese players with multiple games running test this with their opponents and see if it is in fact the wp.prfs file that’s getting overwritten that causes the problems?

Treespiders post:

Andrew,

The synch problem is:

The witp001.pws file that I send you when replayed does not correspond to the combat that the japanese player sees. Nor will it mesh with the combatreport.txt file.

I believe the work around will be to send my wp.prfs file with each zip, for each game, every time. I noticed that my wp.prfs file seems to update each time I run a turn from a different game.

Forest

My reply:

I suspect this is the culprit then, each time you run a different saved game, a new set of die rolls gets generated and that's why we are seeing different results on the allied side, we are using die rolls for a different games turn during our replays. The in game results don't actually change because the game has its own save slot, but the replay does because all your email games share this same file.

So try not running more than one game between emails. Do your turn and pack up the turn and send it off. That may fix the problem without having to do multiple installs of the game. Whatever you do don't let the wp.prfs file get overwritten with a new save games rolls till you send off that players turn.

Jim

[ChezDaJez]

If this is the culprit, I would think that it would be an easy fix. Simply quit the game after you've sent a each turn. Then restart it for your next opponent.

Anyone esle think that would take care of it?

Chez

[Captain Cruft]

That doesn't work. Or at least it's not the whole story. At one time I only had one game going and still had the same problem. Never touched the game other than to run the PBEM turns from my single opponent.

[jwilkerson]

BTW the host ( the Japanese player ) is not corrupted or broken or damaged as far as I know or as far as I've ever heard ... only the ability of the Allied player to see what is happening. It is fairly standard to send the combatreport.txt file from the Japanese player ... along with the turn file and the 001 file .. this way the Allied player can determine if there is a synch problem ... and can see what really happened regardless.

And this is a ( fairly ) well known issue which has had a number of threads about it over the past year ... which is how I learned about the "POMPAC" method !

[Jim D Burns]

ORIGINAL: jwilkerson
BTW the host ( the Japanese player ) is not corrupted or broken or damaged as far as I know or as far as I've ever heard ... only the ability of the Allied player to see what is happening. It is fairly standard to send the combatreport.txt file from the Japanese player ... along with the turn file and the 001 file .. this way the Allied player can determine if there is a synch problem ... and can see what really happened regardless.

Problem is Treespider sends those files every turn and we still have the sync issue. In fact he loaded the same files from the zip he sent me back into his game and he got a false replay too, so it's not those files, it's something else.

Jim

[Jim D Burns]

ORIGINAL: Captain Cruft

That doesn't work. Or at least it's not the whole story. At one time I only had one game going and still had the same problem. Never touched the game other than to run the PBEM turns from my single opponent.

ARGH! couldn't you have let us have some hope for at least 30 minutes! [:@]

[;)]

Jim

[jwilkerson]

ORIGINAL: Jim D Burns

ORIGINAL: jwilkerson
BTW the host ( the Japanese player ) is not corrupted or broken or damaged as far as I know or as far as I've ever heard ... only the ability of the Allied player to see what is happening. It is fairly standard to send the combatreport.txt file from the Japanese player ... along with the turn file and the 001 file .. this way the Allied player can determine if there is a synch problem ... and can see what really happened regardless.

Problem is Treespider sends those files every turn and we still have the sync issue. In fact he loaded the same files from the zip he sent me back into his game and he got a false replay too, so it's not those files, it's something else.

Jim

Sorry - I wasn't trying to say sending the combatreport.txt would fix the problem just allow it to be detected ... the only "fix" I've ever seen is the combat report switch toggle 2 turn process described above ...

[Captain Cruft]

There is no hope. Not with this issue anyway.

The fundamental cause is that PBEM was not part of the original design (WiR/PacWar). Or, to state it more precisely, the combat resolution engine is inextricably tied up with the user interface. Bad design but there you go

[EDIT]: Whilst my above unnecessarily snidy comment is probably true I think I may have worked out what is going on here. It's a "buffer overflow" issue, where the data involved with random numbers gets over-written by "slot arrays" becoming too full. The kludgy solution to which is to read the random numbers/seeds/whatever from file every time they are used, rather than just once at start of turn. I will try and contact the devs about it tomorrow.