Jap Player running turn multiple times

[rtrapasso]

ORIGINAL: Yamato hugger

I have tested this using the same PBEM turn on 2 different machines (me as Japan). I tried shutting off the machine mid run, I tried cutting off the replays on 1 and not the other ect, I tried putting the settings at different settings. Results: identical results.

Now doing all that (as far as changing settings and cutting off the replays) will likely change the result the allied player sees, but it will NOT change the actual result regardless of the machine it is run on. Or at least my tests failed to produce any result.

As far as AE not correcting the replay bug I'd like to ask why not? Good God, re-write that routine. Its broken, and you KNOW it.

FIX IT

Good luck on that - from what i understand, it would involve an extensive rewrite of code... Maybe a couple of man-years of work - you want to volunteer?

It is a simple matter to produce replay errors, but i am NOT going to tell people how to do this.

[Yamato hugger]

Extensive re-write? Why?

All the data you need to produce an accurate allied replay is produced by the Jap running of the turn. This data is already saved and written to a file (actually files).

A MINOR routine to read these files and display it as it happened is all that is really needed. (And yes Bob, I could actually write it in basic but my understanding is the code is written in some form of C. Not sure if todays C would read my basic *although I could probably write an emulator, I used to do this to translate between 80 column cards written in RPG II to Basic plus for a PDP 11/70 back in the early 80s*. I ended my programming career before C even came out).

It is not as hard as they would have you believe:
Step 1 retrieve random seed that was sent by the allied save.
Step 2 begin resolving the replay for Jap.
Step 3 every time something is to show on the screen, write this random number to a replay file.
Step 4 repeat steps 2 and 3 until replay is finished
Step 5 on allied replay, instead of doing the Jap replay sequence (that can produce the replay bug), read the replay file created instead.
Step 6 show what happened in step 2 above.
Step 7 repeat step 6 as often as needed.
Step 8 flash up on screen the following: *** REPLAY BUG SQUASHED BY YAMATO HUGGER *** (only once, I am modest after all)

FIX IT!

[rtrapasso]

ORIGINAL: Yamato hugger

Extensive re-write? Why?

All the data you need to produce an accurate allied replay is produced by the Jap running of the turn. This data is already saved and written to a file (actually files).

A MINOR routine to read these files and display it as it happened is all that is really needed. (And yes Bob, I could actually write it in basic but my understanding is the code is written in some form of C. Not sure if todays C would read my basic *although I could probably write an emulator, I used to do this to translate between 80 column cards written in RPG II to Basic plus for a PDP 11/70 back in the early 80s*. I ended my programming career before C even came out).

It is not as hard as they would have you believe:
Step 1 retrieve random seed that was sent by the allied save.
Step 2 begin resolving the replay for Jap.
Step 3 every time something is to show on the screen, write this random number to a replay file.
Step 4 repeat steps 2 and 3 until replay is finished
Step 5 on allied replay, instead of doing the Jap replay sequence (that can produce the replay bug), read the replay file created instead.
Step 6 show what happened in step 2 above.
Step 7 repeat step 6 as often as needed.
Step 8 flash up on screen the following: *** REPLAY BUG SQUASHED BY YAMATO HUGGER *** (only once, I am modest after all)

FIX IT!

First of all - that WOULD (potentially) require a rewrite of the code - you possibly aborting a large chunk of the normal gameplay, and creating a whole new set of programming... and perhaps potential new problems.

And it still wouldn't address one of the most common causes of replay problems, ie. mismatched EXE files and mismatched map files... i would think this would be more easily addressed by having the game check that they matched, though.


EDIT: Thinking on it more, this actually may be how the game is already written (i don't know - i've never seen the programming per se) ... however, if there are memory management problems (which there are) - things might not behave the same when you get a result number and you apply them to values in memory that are mucked up... i think you still will have to address the memory management issue.

[Grotius]

Thanks for your reply, rtapasso. It sounds like stomping the sync bug would be a non-trivial fix. And I appreciate that there may well be higher priorities -- a reasonable AI is certainly higher on my list, as I play the AI more than PBEM. Still, would you mind if I posted the same question in the AE forum, to see if Joe etc have anything to add?

[rtrapasso]

ORIGINAL: Grotius

Thanks for your reply, rtapasso. It sounds like stomping the sync bug would be a non-trivial fix. And I appreciate that there may well be higher priorities -- a reasonable AI is certainly higher on my list, as I play the AI more than PBEM. Still, would you mind if I posted the same question in the AE forum, to see if Joe etc have anything to add?

Nope, don't mind at all... as i said above, i am guessing it won't be much difference as it is the same game engine, but i don't KNOW.

[Yamato hugger]

ORIGINAL: rtrapasso

First of all - that WOULD (potentially) require a rewrite of the code - you possibly aborting a large chunk of the normal gameplay, and creating a whole new set of programming... and perhaps potential new problems.

And it still wouldn't address one of the most common causes of replay problems, ie. mismatched EXE files and mismatched map files... i would think this would be more easily addressed by having the game check that they matched, though.


EDIT: Thinking on it more, this actually may be how the game is already written (i don't know - i've never seen the programming per se) ... however, if there are memory management problems (which there are) - things might not behave the same when you get a result number and you apply them to values in memory that are mucked up... i think you still will have to address the memory management issue.

A re-write? Nope, an add-on to existing code. And certainly not "an extensive rewrite".

I agree the game itself should make sure the versions and map files are the same. This is 1 (count 'em ONE) if/then statement [if player_jap_map<>player_allied_map or player_jap_gamever<>player_allied_gamever then print "incompatable map or version"]. If you wanted to get more detailed you do this:
[if player_jap_map<>player_allied_map then print "incompatable map"]
if player_jap_gamever<>player_allied_gamever then print "incompatable version"]

But thats 2 lines of code [X(] (4 if you count the line to add it to the .001 file and the one to retrieve it from the .001 file)

As far as memory issues, when you are reading off a sequential file from the disk, there is NO memory issues other than the variables already in use.

FIX IT!

[rtrapasso]

ORIGINAL: Yamato hugger

ORIGINAL: rtrapasso

First of all - that WOULD (potentially) require a rewrite of the code - you possibly aborting a large chunk of the normal gameplay, and creating a whole new set of programming... and perhaps potential new problems.

And it still wouldn't address one of the most common causes of replay problems, ie. mismatched EXE files and mismatched map files... i would think this would be more easily addressed by having the game check that they matched, though.


EDIT: Thinking on it more, this actually may be how the game is already written (i don't know - i've never seen the programming per se) ... however, if there are memory management problems (which there are) - things might not behave the same when you get a result number and you apply them to values in memory that are mucked up... i think you still will have to address the memory management issue.

A re-write? Nope, an add-on to existing code. And certainly not "an extensive rewrite".

I agree the game itself should make sure the versions and map files are the same. This is 1 (count 'em ONE) if/then statement [if player_jap_map<>player_allied_map or player_jap_gamever<>player_allied_gamever then print "incompatable map or version"]. If you wanted to get more detailed you do this:
[if player_jap_map<>player_allied_map then print "incompatable map"]
if player_jap_gamever<>player_allied_gamever then print "incompatable version"]

But thats 2 lines of code [X(] (4 if you count the line to add it to the .001 file and the one to retrieve it from the .001 file)

As far as memory issues, when you are reading off a sequential file from the disk, there is NO memory issues other than the variables already in use.

FIX IT!

i don't believe you are correct in this... but i am not a programmer by profession (although i have done programming for "fun" in the past, but not in C++), so i am not going to argue the matter... you can volunteer to do programming if you like...

[Yamato hugger]

Bob, the point is this:

The replay bug is the single biggest complaint with the game (note, I did not say the most important complaint) and it has been there since UV. Its time to end the BS and FIX IT!

Leader bug probably the 2nd biggest complaint (and in my opinion the most important considering how many die rolls in this game are influenced by the leader stats).

[rtrapasso]

ORIGINAL: Yamato hugger

Bob, the point is this:

The replay bug is the single biggest complaint with the game (note, I did not say the most important complaint) and it has been there since UV. Its time to end the BS and FIX IT!

Leader bug probably the 2nd biggest complaint (and in my opinion the most important considering how many die rolls in this game are influenced by the leader stats).

The replay bug is not one bug: it is several... and despite what you might think, it is not easy to fix from what the programmers have told me... they've been knocking down some of the problems, but certainly not all of them... if it was as easy to fix as you suggest, it would have been done a long time ago.

It is the leading problem now since other huge problems have been alleviated.

But, you are welcome jump up and down and yell "FIX IT" if you think it will help.

[Yamato hugger]

Well, my opinion is and always has been to trash the code, start over and dont bother with an AI (the justification for this is that AE isnt a stand alone game as you have to have WitP to play it) therefore AE would be just another WitP mod with no AI option [8D]

[Yamato hugger]

ORIGINAL: rtrapasso


But, you are welcome jump up and down and yell "FIX IT" if you think it will help.

I think anything short of it being fixed is an insult to everyone that plays the game. Its broken, always has been, yet nothing is done about it. "we would have do actually work on it" is no excuse. Want us to PAY for this AE? Better work then, huh? I mean its simple really. Would you buy a car that came with 3 tires?

Salesman: oh, to get the 4th tire we would have to cut down another rubber tree, and that would take a total rework of the plantation.

Bob: OK, I guess I will take it as it is.

[&:]

Fix it.

[rtrapasso]

ORIGINAL: Yamato hugger

ORIGINAL: rtrapasso


But, you are welcome jump up and down and yell "FIX IT" if you think it will help.

I think anything short of it being fixed is an insult to everyone that plays the game. Its broken, always has been, yet nothing is done about it. "we would have do actually work on it" is no excuse. Want us to PAY for this AE? Better work then, huh? I mean its simple really. Would you buy a car that came with 3 tires?

Salesman: oh, to get the 4th tire we would have to cut down another rubber tree, and that would take a total rework of the plantation.

Bob: OK, I guess I will take it as it is.

[&:]

Fix it.

[treespider]

ORIGINAL: Yamato hugger

ORIGINAL: rtrapasso


But, you are welcome jump up and down and yell "FIX IT" if you think it will help.

I think anything short of it being fixed is an insult to everyone that plays the game. Its broken, always has been, yet nothing is done about it. "we would have do actually work on it" is no excuse. Want us to PAY for this AE? Better work then, huh? I mean its simple really. Would you buy a car that came with 3 tires?

Salesman: oh, to get the 4th tire we would have to cut down another rubber tree, and that would take a total rework of the plantation.

Bob: OK, I guess I will take it as it is.

[&:]

Fix it.

Take it up with Wilkerson as he knows the code and has explained the sync bug in the past.. It is not one bug but could be several thousand bugs that all have the same result - Out of Sync.

[Yamato hugger]

From what I have heard the air routines were completely redone bottom to top. Same with naval. Same with land combat. The map was completely redone and the scale was changed. This means the whole data base was changed (air ranges, ship range and fuel). They added 2 (?) waypoints into naval movement. I dont remember what else they touted in their press release.

But the replay routine cant be redone so it works? Bull.

I havent seen the code, but I was told by an informed source that when GG first started writing games, he was a game designer not a programmer. He went out and taught himself how to program, but had no real world experience to draw on. He didnt know about subroutines and system calls and other little tricks of the trade so I would imagine that every die roll in the game is a seperate line of code meaning if you change one thing, you have to find every single place in the program that does the same thing and make that change, miss one and you really screw things up, and that being the case its no wonder it is so hard to track and squash bugs in it.

And for that very reason the old code should be flushed and replaced by maintainable code. I worked for a software house back in the early 80s that had 6 year old code that had been changed and modified by probably a dozen different people, and I know first hand how difficult it is to try to maintain. There is a time when you have to say "screw it, this is going to be replaced". I see no time like the present for this engine. The engine they are using is broken, always has been. They know it. If they choose to use it regardless, I cant see where that merits your defending it.

I dont know how much time you put into 1.807, but I put a lot of man hours into it. Was due out early last Sept and was a week away from release and JW started working on AE and never heard from 1.807 again. No explaination, no release, nothing. It was a week away and was never released. Why? That left a sour taste in my mouth. Seeing some of the early develpomental discussions with AE convinced me that I didnt want to be a part of it. I wont elaborate.

I didnt even consider buying WPO. Chief reason is it used this same engine and has the same bugs in it. I didnt see a point in wasting the $$$ on it. Frankly I have my doubts about AE because of the same thing. The replay and leader bugs are going to remain as a part of the game, I see no need to fork out more money for the same broken game.

Leaders play such a huge part of this game, and probably more so with the new routines and yet the leader bug will persist. Replay is the game. Its the summary. If it isnt accurate why even have it? Just load up the turn and look around the map and try to figure out what happened. If they are going to use the same basic engine, they will have the same basic problems. So all you are buying is the same game with a slightly different look (al la WPO).

Now if you think my concerns arent worthy of your time to read, then take out your popcorn and sit in the corner. My request to FIX IT clearly isnt aimed at you since you havent the power to fix it. But you seem to defend it like you do. If Matrix chooses not to listen at least to concerns from a potential buyer, then that is their business decision.

[pompack]

[>:]

[donkey_roxor]

ORIGINAL: Yamato hugger

Well, my opinion is and always has been to trash the code, start over and dont bother with an AI (the justification for this is that AE isnt a stand alone game as you have to have WitP to play it) therefore AE would be just another WitP mod with no AI option [8D]

And tough luck for those of us who only play the AI, huh?

Personally, I don't care about the replay bug or the leader bug. I get a lot of enjoyment playing WiTP as it is, and I'm looking forward to AE, especially with any AI "improvements". I fully support the direction the dev team is going in.

[Yamato hugger]

ORIGINAL: donkey_roxor

ORIGINAL: Yamato hugger

Well, my opinion is and always has been to trash the code, start over and dont bother with an AI (the justification for this is that AE isnt a stand alone game as you have to have WitP to play it) therefore AE would be just another WitP mod with no AI option [8D]

And tough luck for those of us who only play the AI, huh?

Personally, I don't care about the replay bug or the leader bug. I get a lot of enjoyment playing WiTP as it is, and I'm looking forward to AE, especially with any AI "improvements". I fully support the direction the dev team is going in.

Is the AI a challenge to you? Seriously? If the same engine is used in AE, you can bet the AI will also be the same. I mean if the code is so screwed up that they wont even re-write it to fix the sync bug, how hosed up do you think the AI stuff is? AI is good for getting the fundementals down, but I cant say I have ever played a game where the AI was a serious challenge.

[donkey_roxor]

ORIGINAL: Yamato hugger

ORIGINAL: donkey_roxor

ORIGINAL: Yamato hugger

Well, my opinion is and always has been to trash the code, start over and dont bother with an AI (the justification for this is that AE isnt a stand alone game as you have to have WitP to play it) therefore AE would be just another WitP mod with no AI option [8D]

And tough luck for those of us who only play the AI, huh?

Personally, I don't care about the replay bug or the leader bug. I get a lot of enjoyment playing WiTP as it is, and I'm looking forward to AE, especially with any AI "improvements". I fully support the direction the dev team is going in.

Is the AI a challenge to you? Seriously? If the same engine is used in AE, you can bet the AI will also be the same. I mean if the code is so screwed up that they wont even re-write it to fix the sync bug, how hosed up do you think the AI stuff is? AI is good for getting the fundementals down, but I cant say I have ever played a game where the AI was a serious challenge.

I never said the AI was a challenge, although I have been surprised by some of its moves before. Nevertheless, I do enjoy playing the game as it is. And it is much more convenient for me to play against a rather anemic AI whenever and wherever I want than to start a more challenging PBEM restricted by the real-world obligations of both sides. I understand that the replay bug can be particularly galling for the PBEM contingent, but as an AI player (which I understand may be the majority of WiTP owners, as I seem to recall seeing in a poll here), I'm happy to hear that the dev team is attempting to improve it, even if the results aren't everything I'd hope for. I am hesitant to say that this will be an absolute purchase for me (final details and price will determine that), but I generally approve of the direction the development team has taken.

[Yamato hugger]

Well as I said before you have to have WitP to play AE (by my understanding anyways), the AI is in WitP. AE is just a scenario in this case that you have to pay for. Not being able to play AE without an AI is just not being able to play 1 scenario without an AI and this in my humble opinion is noting to complain about when you consider that you have to PAY for AE. If you know AE doesnt have an AI going to begin with, then the people that only play the AI wont waste their hard earned cash on it. For those of you that enjoy the AI, you will still have it, you just wouldnt have it in that ONE scenario.

And who knows? With fresh maintainable code in AE, maybe they would put an AI into it and make it a stand alone game and not require you to have WitP to play it. I am not affiliated with Matrix in any way, but if I had to speculate I would guess that requiring you to purchase BOTH WitP and AE in order to play AE is to appease GG in some way. I mean it makes no sense to me if AE is going to use the same engine as WitP that you should actually have to have WitP for it in any other event.

[Charbroiled]

Yamato Hugger

I respect your opinion, but I believe that you are way off base. From what I have read:

per Eric:

Seems like we had some confusion here. Let me reiterate what was in the press release to clear things up.

1. WITP AE will require WITP to be installed in order to install and play, but it does not require that the installation of WITP be updated, you could be running stock WITP v1.00 and that would be fine.
2. WITP AE will install in a separate location, so that you can still play the original WITP as well as the new WITP AE, once installed they will effectively function as separate games.

Regards,

- Erik

From my understanding, AE will be a completely separate game platform. The team has been working on the AI a lot in order to improve it. If I understand correctly, because of this focus on improving the AI, the release of AE has been postponed/pushedback. I would hazard to guess that they felt that an improved AI was much more important in the long run then trying to chase down the sync bug. I would agree with this prioritization 100%. While the sync bug is a nuisance, it is not major problem IMHO.

I am VERY satisfied with WITP, even with some of the bugs that are still left. The support for this game that has been provided by Matrix and volunteers has been unprecedented. Because of this support and because of the enjoyment I have received from playing WITP, I will be buying AE the day that it is released without waiting for a review or comments on how good the game is. This is also unprecedented by me.