Patch for Synchronicity Bug? (Full Version)

All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition



Message

Rusty1961 -> Patch for Synchronicity Bug? (10/4/2017 11:24:20 PM)

Very frustrating. I watched the replay of the turn- a rather slow, uneventful turn-then while taking my turn I notice damage to ship that shouldn'thave any so I looked at the combat report. Nothing like the combat report I watched.

Nothing.

So I don't know what is the truth here. Did the Japanese player see the actual report as it was in totality and I see nothing near what happened? Is my combat report I read when I open my turn the truth, or is there fog of war?

All I know was my replay didn't have 10% of what appeared to have happened during the combat phase.

Any chance this bug gets fixed?



IdahoNYer -> RE: Patch for Synchronicity Bug? (10/4/2017 11:38:56 PM)

No patch that I'm aware of. Been dealing with it since L_S_T and I started our PBEM a few years ago. Happens every so often.

Best solution we've come up with is that the Jpn player sends not only the two save files, but also both combat report files. After you go through the replay, replace the combat report files and you have "the truth". Also helps if you're using tracker or combat reporter - replace those files before you run those utility programs.



BillBrown -> RE: Patch for Synchronicity Bug? (10/4/2017 11:47:40 PM)

I think it helps if the Japanese player does not use the ESC key to speed up combat, but uses the Z key instead.



Rusty1961 -> RE: Patch for Synchronicity Bug? (10/5/2017 12:55:40 AM)

what is the difference in Z key vs. ESC?



Chris21wen -> RE: Patch for Synchronicity Bug? (10/5/2017 5:31:45 AM)

If it's just two ships in the same TF then it's possibly ship collision.



BBfanboy -> RE: Patch for Synchronicity Bug? (10/5/2017 6:07:55 AM)


quote:

ORIGINAL: Chris H

If it's just two ships in the same TF then it's possibly ship collision.

+1.
Look in the ops report for a line about collision, grounding, attack by search aircraft, etc.



crsutton -> RE: Patch for Synchronicity Bug? (10/5/2017 4:08:32 PM)

Happens once in a while. Usually when a big fight takes place. Annoying for the Allied player but rare enough to now be a game breaker. Don' plan on seeing a fix. If it could have been fixed it would have by now.



HansBolter -> RE: Patch for Synchronicity Bug? (10/5/2017 4:43:59 PM)

Allied players need to stop caving on this and DEMAND the results they see.

What you all keep referring to as a 'replay' is nothing of the kind.

It's supposed to be a RECORDING of the actual events as they transpired which means that it is NOT the recording that is wrong, but rather the saved game file the Japanese players foists on the Allied player while demanding he accept a corrupted save file and discard the recording of actual events.



Encircled -> RE: Patch for Synchronicity Bug? (10/5/2017 4:44:35 PM)

It can be annoying, but as long as you have the combat report and the next turn you can check what actually happens.

Finding out I'd lost 4 carrier in exchange for one bomb hit after the replay came across as pretty even was a knock, but it shouldn't be a game ender if we are being perfectly honest.



Lokasenna -> RE: Patch for Synchronicity Bug? (10/5/2017 4:54:44 PM)

quote:

ORIGINAL: BillBrown

I think it helps if the Japanese player does not use the ESC key to speed up combat, but uses the Z key instead.


quote:

ORIGINAL: Rusty1961

what is the difference in Z key vs. ESC?


There isn't one, practically speaking.

quote:

ORIGINAL: crsutton

Happens once in a while. Usually when a big fight takes place. Annoying for the Allied player but rare enough to now be a game breaker. Don' plan on seeing a fix. If it could have been fixed it would have by now.


My personal theory is that it's related to (carrier) CAP. I've only had sync bugs occur beginning with either:

1) immediately in the turn (which will be a ship/TF because that's the first thing that can happen), potentially related to the "air superiority" calculation that is performed as the replay is generated

2) an air attack on a TF


quote:

ORIGINAL: HansBolter

Allied players need to stop caving on this and DEMAND the results they see.

What you all keep referring to as a 'replay' is nothing of the kind.

It's supposed to be a RECORDING of the actual events as they transpired which means that it is NOT the recording that is wrong, but rather the saved game file the Japanese players foists on the Allied player while demanding he accept a corrupted save file and discard the recording of actual events.


This is 100% backwards.

The WRONG file is the "results" that the Allied player sees in the incorrect replay. It is the replay that is incorrect and the turn file that is correct; this is why there was a change some time ago to put the 100% Real Actually Happened combatreport.txt into the orders turn file itself.

The only exception to this is if the Japanese player does not perform the beta patch update in the correct order (update first and then generating the replay with the new version, instead of with the old version as is proper).



HansBolter -> RE: Patch for Synchronicity Bug? (10/5/2017 6:25:17 PM)

You didn't grasp the point I was trying to make.

The saved game isn't created until after the turn execution is complete.

If the Japanese player observes a turn execution that matches the saved game and combat report generated at the end of the turn execution and then a so called 'replay' is created that doesn't match any of the first three events than the 'replay' is NOT the recording it is supposed to be.

Therein lies the root of the problem.

If the recording of the turn execution was exactly that instead of the 'replay' that it appears to be instead none of this would be occurring.

Why does the game generate a false replay instead of an actual recording?

This is one of the main reasons that I don't play PBEM.

I don't want to watch a replay, I want to watch the turn execute and only Japanese players get that honor in PBEM games.

Allied players are second class citizens in PBEM.



witpqs -> RE: Patch for Synchronicity Bug? (10/5/2017 6:34:47 PM)


quote:

ORIGINAL: HansBolter

Allied players need to stop caving on this and DEMAND the results they see.

What you all keep referring to as a 'replay' is nothing of the kind.

It's supposed to be a RECORDING of the actual events as they transpired which means that it is NOT the recording that is wrong, but rather the saved game file the Japanese players foists on the Allied player while demanding he accept a corrupted save file and discard the recording of actual events.

That's not the way it works. A recording would be huge. Massive. Keep in mind all the technology compromises that had to be made in making the game. In fact the replay mechanism was in the first version - actually in all the prior work for more than a decade that led up to the first WITP version published in 2004.

Because of the size of a recording (how fast was your internet access in 2004?) and how impractical that would be to play by email they implemented a replay. The game starts with a seed for a random (or pseudo-random) number and all chance events through the game are derived from that, in order one after the other. The turn file contains the then current state of that chain of random numbers. The Japan player sees the turn resolution, which starts with the (then) current random number.

The replay which the Allied player sees starts with the very same random number that the turn resolution started with and - because of that - all the random events are supposed to come out the same (they follow the same chain of random numbers). If the code were perfect that would work every time. But code is rarely perfect, and this code has to contend with things like outside interruptions in addition to the likelihood that the code contains its own bugs.

There used to be a lot of sync bugs but they did find and did eliminate almost all of them. We are left with some. That is not desirable but over the course of all the turns PBM players flip to each other they are rare. When it happens to me I don't like it any more than you do, but before Michael had to bow out from maintenance he spent significant labor looking for the remaining sync bug(s) and could not find the cause(s).

BTW, if you go through the explanation I gave above you will notice one saving grace in the situation. The (then current) random number that turn resolution starts from is the same random number that the replay starts from. That means that each replay starts in sync with the turn resolution - each sync bug occurrence is inherently a one-turn event. There are times when a sync bugs happens to a player twice in a row and they might even be caused by the same (as yet unidentified) bug, but they are separate events.



HansBolter -> RE: Patch for Synchronicity Bug? (10/5/2017 6:41:21 PM)

Thanks for the succinct explanation of how the Allied player got screwed from the very beginning.

I want to watch the turn execution.

That's why I won't play PBEM as Allies.

Furthermore, if the game is executing two different turns for two different players why shouldn't the Allied player have just as much right to demand that the turn he witnessed be the one the game moves forward with?

Why do the Japanese players get preferential treatment?



witpqs -> RE: Patch for Synchronicity Bug? (10/5/2017 7:09:09 PM)

Go spend some time with Jojo. Hugs are recommended, maybe your hand on his side while he sleeps. [:)]

Re-read my explanation again tomorrow. They did the best they could to show the Allied player the exact same thing the Japan player sees. It works about or greater than 99% of the time. But they are not magicians and that last fraction of a percent has proven elusive. Still the best game in town. [8D]



HansBolter -> RE: Patch for Synchronicity Bug? (10/5/2017 7:24:35 PM)

Sound advice.

I'll never dispute your last sentence.

And BTW Jojo is a female.



Lokasenna -> RE: Patch for Synchronicity Bug? (10/5/2017 7:27:20 PM)


quote:

ORIGINAL: HansBolter

You didn't grasp the point I was trying to make.

The saved game isn't created until after the turn execution is complete.

If the Japanese player observes a turn execution that matches the saved game and combat report generated at the end of the turn execution and then a so called 'replay' is created that doesn't match any of the first three events than the 'replay' is NOT the recording it is supposed to be.

Therein lies the root of the problem.

If the recording of the turn execution was exactly that instead of the 'replay' that it appears to be instead none of this would be occurring.

Why does the game generate a false replay instead of an actual recording?

This is one of the main reasons that I don't play PBEM.

I don't want to watch a replay, I want to watch the turn execute and only Japanese players get that honor in PBEM games.

Allied players are second class citizens in PBEM.


No, I grasped the point you were making - but you aren't grasping how PBEM turns actually work.

The Allied replay shows different things to the Allied player than the Japanese replay shows to the Japanese player. The replay IS watching the turn execute. Sometimes it bugs out and you have to look at the combatreport.txt that is within the orders file. The orders file is always, always, always 100% correct.

I've played 4000+ turns in PBEM and had less than 10 sync bugs.

witpqs is correct as well - it's a problem somewhere with the random number seed being used in the turn (random numbers for computers are drawn from a giant table of randomized numbers; the seed is the number it starts on, e.g. a seed of 10 is the 10th number in the table).



Lokasenna -> RE: Patch for Synchronicity Bug? (10/5/2017 7:28:04 PM)


quote:

ORIGINAL: HansBolter

Thanks for the succinct explanation of how the Allied player got screwed from the very beginning.

I want to watch the turn execution.

That's why I won't play PBEM as Allies.

Furthermore, if the game is executing two different turns for two different players why shouldn't the Allied player have just as much right to demand that the turn he witnessed be the one the game moves forward with?

Why do the Japanese players get preferential treatment?


It's one turn with two replays, because there are two players playing opposing sides that can't see everything the other side is doing.



witpqs -> RE: Patch for Synchronicity Bug? (10/5/2017 7:36:50 PM)


quote:

ORIGINAL: HansBolter

Sound advice.

I'll never dispute your last sentence.

And BTW Jojo is a female.

Oh, she would never let me hear the end of that!



modrow -> RE: Patch for Synchronicity Bug? (10/5/2017 7:42:58 PM)


quote:

ORIGINAL: HansBolter

Why do the Japanese players get preferential treatment?


Based on my understanding of the effect (which may not be 100% correct), my guess would be because that is the conclusion you want to reach.

I want to stress that really no offense is intended with that remark.

You are correct in that the Allied player loses information from the combat animation files, which can be essential especially in naval combat and is at least nice to have in a2a combat. But the in-game combat report file you obtain from the game file is correct according to my experience, so on the combat report level you get the right information.

On the other hand, however, it's not just the results that change but entirely different battles can occur. So you gain additional intel on local force distribution which the IJ player never knows you have gained. Also, especially for land combat it can be very instructive to see the bandwidth of results of two executions of the same ground battle for making calls how to continue that battle.

So Allied player loses naval and a2a info, but gains intel on asset allocation/distribution and info on ground combat.

Just my 2cts.

Hartwig



HansBolter -> RE: Patch for Synchronicity Bug? (10/5/2017 7:54:41 PM)

So you are saying the Allies don't get as much of the short end of the stick in this deal as I have portrayed?

I can accept that.

In case you haven't been around long enough or regularly enough to notice, I often take extreme positions for the sake of stimulating debate.



Lokasenna -> RE: Patch for Synchronicity Bug? (10/5/2017 8:50:12 PM)

quote:

ORIGINAL: hartwig.modrow

On the other hand, however, it's not just the results that change but entirely different battles can occur. So you gain additional intel on local force distribution which the IJ player never knows you have gained. Also, especially for land combat it can be very instructive to see the bandwidth of results of two executions of the same ground battle for making calls how to continue that battle.

So Allied player loses naval and a2a info, but gains intel on asset allocation/distribution and info on ground combat.



Yep, this too.

A sync bug might cause an air strike to occur on a TF that had low DL or perhaps no DL at all in the actual sequence of events, and then the Allied player knows that Yamato and Musashi are 2 hexes away from their CV fleet on their first major amphibious operation of the war despite not actually having them spotted in the for-real turn.



modrow -> RE: Patch for Synchronicity Bug? (10/5/2017 8:59:24 PM)


quote:

ORIGINAL: HansBolter

So you are saying the Allies don't get as much of the short end of the stick in this deal as I have portrayed?

I can accept that.

In case you haven't been around long enough or regularly enough to notice, I often take extreme positions for the sake of stimulating debate.


I actually appreciate that approach to stimulating debate, but work in a field where you have to be able to argue both sides of an issue, so there's always a chance that a pointed statement triggers one of the rare events of my leaving the lurker position and posting.

At first, when being hit by the synch bug, my sentiments were quite similar to yours - but sentiment tend to distort the picture. At that point of time I had not yet realized that you get a correct combat report with the game file, which means Allied player can retrace what happened. I still prefer if no sync bug occurs, but believe that also the IJ player is better off in some aspects if it does not happen.

So all is well - especially as many triggers of the sync bug have been successfully removed [&o] - and you can start a PBEM. It's great fun !

Hartwig



Zorch -> RE: Patch for Synchronicity Bug? (10/5/2017 9:03:29 PM)


quote:

ORIGINAL: Rusty1961

Very frustrating. I watched the replay of the turn- a rather slow, uneventful turn-then while taking my turn I notice damage to ship that shouldn'thave any so I looked at the combat report. Nothing like the combat report I watched.

Nothing.

So I don't know what is the truth here. Did the Japanese player see the actual report as it was in totality and I see nothing near what happened? Is my combat report I read when I open my turn the truth, or is there fog of war?

All I know was my replay didn't have 10% of what appeared to have happened during the combat phase.

Any chance this bug gets fixed?

Listen to the Police!


[image]local://upfiles/34241/CF058B65D59D41D7A548FB237CE14270.jpg[/image]



spence -> RE: Patch for Synchronicity Bug? (10/6/2017 2:39:03 AM)

To combat the Replay Bug you can have the IJ Player send his written Combat Summary along with the Replay and Turn. Before you run the replay file just read over the Combat Summary file to see if what you afterwards see matches what you read. If not then at least you know the Replay Bug bit you.



wdolson -> RE: Patch for Synchronicity Bug? (10/6/2017 3:25:33 AM)

The "sync bug" is actually a slew of bugs that all behave the same way. Michael fixed a ton of them, but I don't think he's actively working on the code at the moment. That said, if someone has a sync bug that is reproducible and you can send a save to Michael, he can fix it when he gets back to the code. Fixing bugs that are consistently reproducible is not that difficult. It's the intermittent ones that make programmers pull their hair out, drink, or whatever they do under stress.

Bill



Lokasenna -> RE: Patch for Synchronicity Bug? (10/6/2017 5:15:58 AM)


quote:

ORIGINAL: wdolson

The "sync bug" is actually a slew of bugs that all behave the same way. Michael fixed a ton of them, but I don't think he's actively working on the code at the moment. That said, if someone has a sync bug that is reproducible and you can send a save to Michael, he can fix it when he gets back to the code. Fixing bugs that are consistently reproducible is not that difficult. It's the intermittent ones that make programmers pull their hair out, drink, or whatever they do under stress.

Bill


Throw things.

Make up new swear words.

Constantly mutter to self and/or at the computer screen, holding a conversation with it.



witpqs -> RE: Patch for Synchronicity Bug? (10/6/2017 6:12:13 AM)

I'm pretty sure the ones that pretty ran into which he could not find did not reproduce on his machine. I know he spent a ton of time looking for possible causes anyway. [&o]



Bullwinkle58 -> RE: Patch for Synchronicity Bug? (10/6/2017 11:19:57 AM)

quote:

ORIGINAL: HansBolter

Thanks for the succinct explanation of how the Allied player got screwed from the very beginning.

I want to watch the turn execution.

That's why I won't play PBEM as Allies.

Furthermore, if the game is executing two different turns for two different players why shouldn't the Allied player have just as much right to demand that the turn he witnessed be the one the game moves forward with?

Why do the Japanese players get preferential treatment?


Hans, you can't "watch the turn resolution." If you could, you would see EXACTLY what the Japan player sees, including the location of all of his forces. As part of the "reconstruction of the replay" the game uses and the posters here have described, the code edits out the visual and text message parts that are opsec to each player.

I agree with witpqs. Synch bugs, if both players are exactly patched in congruence, are exceedingly rare now. In at least a third of the cases I've thought there was one it turned out there was a collision in the combat events file that I had missed. IF synch bugs are the reason you don't play PBEM, play PBEM.

Edit: yet another case where I should have read the whole thread first.



LargeSlowTarget -> RE: Patch for Synchronicity Bug? (10/6/2017 12:55:20 PM)


quote:

ORIGINAL: Lokasenna


quote:

ORIGINAL: wdolson

The "sync bug" is actually a slew of bugs that all behave the same way. Michael fixed a ton of them, but I don't think he's actively working on the code at the moment. That said, if someone has a sync bug that is reproducible and you can send a save to Michael, he can fix it when he gets back to the code. Fixing bugs that are consistently reproducible is not that difficult. It's the intermittent ones that make programmers pull their hair out, drink, or whatever they do under stress.

Bill


Throw things.

Make up new swear words.

Constantly mutter to self and/or at the computer screen, holding a conversation with it.


The language most programmers know best is CURSING [:D]



rustysi -> RE: Patch for Synchronicity Bug? (10/12/2017 7:30:30 PM)


quote:

ORIGINAL: wdolson

The "sync bug" is actually a slew of bugs that all behave the same way. Michael fixed a ton of them, but I don't think he's actively working on the code at the moment. That said, if someone has a sync bug that is reproducible and you can send a save to Michael, he can fix it when he gets back to the code. Fixing bugs that are consistently reproducible is not that difficult. It's the intermittent ones that make programmers pull their hair out, drink, or whatever they do under stress.

Bill


This is true not only for programmers, but techs as well. The worst intermittent problem I remember from my career took a year and a half before someone finally resolved it. The man hours had to run in the hundreds if not a thousand or more.



Page: [1] 2   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.796875