Our ReVolt Pub

Download this replay. Play it in Arcade Mode in b11.0208 and in a13.0820 (obviously, you will need to use -dev for that). I want Huki or Jigebren to explain all the differences between the two results.

I will not comment anything so far as it is pretty self-explanatory.


[HIDE=Offtopic]I have also just noticed that 7-zip is able to open the .rpl file and it seems to contain a single HUGE file lol[/HIDE]

You want to
Maybe you rather meant you'd like to...

I meant both. I'd like to and I want to. As I want to implies I'd like to, I chose I want to.

We want...

nevermind @ 8 Mar 2015, 05:27 PM wrote:I meant both.

Yep, you did. And you shouldn't.

We want...

Then keep wanting.

what is there to explain?
replays are not compatible across versions.

1.2b and 1.2a are drastically different... for some unknown and unneeded reason

[hide=Offtopic]

jigebren @ 8 Mar 2015, 07:18 PM wrote:

nevermind @ 8 Mar 2015, 05:27 PM wrote:I meant both.

Yep, you did. And you shouldn't.

Not that I see this discussion going anywhere but thats really just nitpicking at this point.

If he intended an aggressive wording he could have very well used something like "demand" or "request" which he didn't, besides he even said "please" in the title.[/hide]
As for the actual topic, I guess the differences between the replays are due to changed behavior in AI and/or physics.
The replay file only stores "events" from the race and lets the game do the rest, of course stuff gets corrected along the way which are the parts where cars get teleported or generally behave weird.

First, when you go to the butcher, you don't say "I want a steack". Even if you have a sign on your forehead that says "Serve me, Please"...
Same rule here. We're not at your goddamn service.

Second, what's the point of this topic actually?
- You're really curious about the Re-Volt Replay internal? That's not the impression your message gives. Then why using an old version like a13.0820 and not just the last alpha? Either there's a point in using it instead of a15.0131, in that case let us know why. Or there's no real reason then the report (if that's a report) sounds almost irrelevant (not to mention Replay saving for later use was not even officially added by the time of both those releases).
- Or maybe you're just trying to prove that a13.0820 is not like the good old 11.0208 sanctified by God himself (if it were the case, I can assure you the Re-Volt Replay design is broken enough by itself that it would actually demonstrate something like nothing at all).

Third. Even if the issue is totally perfectly 100% self-explanatory, we may:
1. already know what it's about (if you had let us know before)
2. don't give a damn about that
3. have fixed it already
4. found out already that it's not fixable
5. whatever that doesn't make it worth spending any time on
In all those cases you've just make us lose our time for nothing. So at least make a kind of summary of what are the issues before asking instructing us by name to justify anything.

Fourth. "nitpicking..." yeah, so far that's what I have the feeling this topic is all about, indeed. If I misread it, then reformulate your request in a straightforward way that leave no doubt for no one about what the point really is. Thank you.

[hide=Offtopic]@1st: I don't think your analogy really fits here, after all the butcher is dependent on his customers and can't afford to flip them off just because they weren't nice enough in his opinion.
Because otherwise they would just buy their meat elsewhere and he would go out of buisness very soon with that attitude (which doesn't apply here).

@4th: The same thing goes the other way around, there is no point in saying "Pretty please" and "Thank you very much" every time there is something to point out or to discuss because then about half of the text written in this forum would be spent on "pleasant" formalities.

Not to mention that "forcing" these formalities on the users makes them kind of meaningless in the end.
And yes, by not giving an informative response to a user just because they didn't use the correct wording pretty much boils down to "forcing". (note that I'm just speaking about your initial two responses here, not point 2) and 3) from your last post which are totally fine and would have been a great initial response instead)

Of course I'm not saying that being rude and tersely is totally fine but I think both sides should have their tolerance borders.

Last but not least, this post here is just meant to be viewed as a personal opinion on things and a more detailed explanation about what I wrote in my last post and why I wrote it.
Therefore I'm totally fine if the discussion about this side topic ends here and no answer to this post is given (but if you want to continue, feel free to do so).[/hide]

Ok, I have just learnt that the Universe has been configured in a way that, if anyone mentions an undocumented difference between last beta and a13.0820 in this forum, all answers will turn weird. Of course, my post was far from the best to start a topic anyway.


The point is, a replay from a b11.0208 does not work the same when run in a13.0820 as it does in b11.0208: when a car collides or makes contact to the ground, its movement as estimated from a13.0820 does not always match its actual behaviour calculated in b11.0208, and that is shown as a weird replay of their path.

It is probably not possible to explain that as flaws in the (same) replay system for both versions, as the differences between the actual race and the replay run in its matching version are orders of magnitude lower than the differences between playing the replay in different RV versions. I have not been able to find anything in the changelog that might explain this. Then, I guess that something has been changed and that has not been added to the changelog. I do not mean that something has gone wrong. I mean that something is different.

That being said, would you mind telling me whether I am wrong about this?

[HIDE=Offtopic]@Kenny
@1st:
Read yourself again, and notice how you're demonstrating the exact opposite of what you expected. The butcher can't afford to flip his customer, so he'll accept being talked rude. That is your point. Still his customers don't even dare to be that commanding, and THAT was my point. Your comment only makes it even more valid.

>> not point 2) and 3) from your last post which are totally fine and would have been a great initial response instead
My first post was an attempt not to lose a quarter of an hour just to explain the evidence.

@4th:
And you've been the first one talking about nitpicking. Seriously...
Not to mention my 4th point was about the request accuracy or goal, not a matter of courtesy anymore, so there's no "thing going the other way around", this remark is just off the mark.

>> by not giving an informative response to a user just because they didn't use the correct wording pretty much boils down to "forcing"
What informative response? The request itself was not even informative. It was roughly: "you two X & Y the developers, try the protocol A and B I'm giving to you, I won't tell what happens, you'll have to find out by yourself, then I want you to explain to me every differences between A and B". It was not only rude, it was also ridiculously arrogant.

>> Therefore I'm totally fine if the discussion about this side topic ends here and no answer to this post is given (but if you want to continue, feel free to do so).
That kind of sentence is just as formal, stereotyped and useless than my "Thank you".[/HIDE]

[hide=Offtopic]

jigebren @ 9 Mar 2015, 03:20 AM wrote:Your comment only makes it even more valid.

Err...I can't really follow your reasoning of why exceptionally all of the butcher's customers are nice (enough) to him but I guess we just have a different point of view regarding butchers and their customers

jigebren wrote:My first post was an attempt not to lose a quarter of an hour just to explain the evidence.
...
And you've been the first one talking about nitpicking. Seriously...
Not to mention my 4th point was about the request accuracy or goal, not a matter of courtesy anymore, so there's no "thing going the other way around", this remark is just off the mark.
...
What informative response? The request itself was not even informative.

My point was that your response was just as rude and uninformative as his question to begin with. And by uninformative I mean that it contained no information about what things he missed in his question that would be required for a proper answer.
Instead you just insisted on a nicer wording and derailed the thread into a discussion about that (only talking about the first 6 posts here).

You could have for example just said "Could you explain what you mean with x and y in greater detail? And a nicer wording would be appreciated."
Or "We don't have the time and patience to look into every matter ourselves so it would be appreciated if you could explain your problem in greater detail instead of being mysterious about it."
There, just as short and time effective as your initial response.

And now that the thread opener rephrased his question to be more appropriate like you requested you still completely ignore it, I wonder why that is the case?

About me being nitpicky - feel free to think about me whatever you like, I merely wanted to point out a few things here and its up to you to think about them or just completely ignore them.
Anyway I have a feeling that this discussion will be going nowhere (thats actually why I wrote that last sentence in my previous post and not as a formality but whatever), so I'll just keep idly watching this thread whatever the response to this post may be.[/hide]

When reporting any bug please be sure to explain your problem in as much detail as possible so as to leave nothing for guess work. Providing clear information (and steps to reproduce if applicable) would greatly increase the chances of getting a response. Others like this one may simply be ignored in the future.

The replay format is already broken and it's very easy to get a different result each time the replay is played. And as jigebren pointed out, this feature was only added in latest alpha and not considered "supported" in previous alphas or beta. So even if you found weirdness by playing replay saved by one version on another, you just have to have fun with it instead of demanding poor developers to study your entire data and submit report (hey it's Monday and I thought I already missed the deadline, don't fire me )

[hide=Offtopic]@Kenny
Oh oh, that's all too much psycho-analysis for me to follow. we were simply baffled at the kind of frustrating demand in this topic: imagine we already know about whatever bug was being reported, but we'll be watching and watching the replay with no clue what we're supposed to look for! So jig had simply used a convenient analogy to try to explain why users should be more responsible with their queries, that's all. There's no need to keep drawing that thread longer and longer (thread for thread, sounds like a better analogy to me ).
[/hide]

[HIDE=Offtopic]End of the story...

Kenny wrote:Err...I can't really follow your reasoning

Really? Here's another chance. The customers don't come and say "I want a steak" because:
A. They are polite and they show a minimum of respect for the butcher.
B. They're just afraid the butcher will cut his steak out of them otherwise.
You've illustrated yourself that the butcher is not going to do anything against his customers, making answer B invalid. A and B are on a boat, B falls off. Who's left?

My own point was already valid in both cases, A and B, as no matter the reason I just meant that you don't speak rude to someone when you have something to ask. You've made it even clearer by proving that it's not for reason B. Now we know that people don't speak rude just because they are polite. If you are not polite, either you have a good reason - because you're my boss or you're mad at me since I may have just spat on your steak -, or you're just showing how disrespectful you are. And "the thread opener" as you call it was neither my boss nor his piece of meat has been snatched or wasted by me as far as I know...

Kenny wrote:You could have for example just said

Of course, I'd have better... Wait, if you want to apply for the educator job, the place is free.

Kenny wrote:And now that the thread opener rephrased his question to be more appropriate like you requested you still completely ignore it, I wonder why that is the case?

Lack of time maybe, you now what that means?

The claiming tone (that you seem not to want to notice) of requests like this one also strongly affect my will to make the slightest effort. The rewording would have been an acceptable initial question, but that's just not how it happened.

And I've answered already in my post above it you had better concentrated on the interesting parts instead of nitpicking for the last two words. To my question "Second, what's the point of this topic actually?", nevermind's reply lets me guess it's the second suggested answer. In that case I've already said what I think about using Replay for that kind of purpose.
He's not convinced with that, that's another matter. But if a more detailed answer is needed then 1) it actually belongs to the bug report section, 2) you really think I can give the answer off the top of my head?
Not to mention I still don't get the point to test a13.0820 instead of the latest one, as I mentioned and hasn't been answered yet. But weirdly, you haven't reacted to that. Talk about skewing...
[/HIDE]

ThugsRook @ 8 Mar 2015, 04:36 PM wrote: what is there to explain?
replays are not compatible across versions.

1.2b and 1.2a are drastically different... for some unknown and unneeded reason

Actually replays are compatible (they've been the same since 1.0) what changes is how the thing is handled by the engine.

Abc @ 9 Mar 2015, 11:54 AM wrote: Actually replays are compatible (they've been the same since 1.0) what changes is how the thing is handled by the engine.

good point.

what we'd like to know, is why?
what mod is so important that this has been changed and is broken?!?
and undocumented of course.

this is the sort of problems that need to be straightened out, or at least explained

This is not a bug report. I would have started the topic in Bug Reports and I would have provided more information if I thought this is a bug.

The replay format may be inaccurate. But the fact that the same replay run twice gives slightly different results is not enough to explain why the same replay run in two different versions shows way different results. And nothing in the changelog seems to have anything to do. That is my point.

I mean, it is as if you have two weighing scales from the same manufacturer. When you weight yourself on the first one you get different results, eg: 66, 69, 67, 70, 67, 68, 65. But, when you weight yourself on the second one you get: 91, 93, 92, 91, 91, 95, 92. Of course, both weighing scales are somewhat awful, but, as you can not explain the different ranges as random results within the same range, isn't it clear that one is different from the other?

Same with GameGauge, fps.txt's first line still reveals "<FPS> Re-Volt v1.00"

The replay format is not "inaccurate". It is broken...

> broken <

Why it reacts differently between b11.0208 and a13.0820, I don't know, and so far frankly, I don't care. And who cares? The discussion could have revealed interesting, if it had been about the Replay internal actually. But no, in fact it rather looks like just yet another attempt to prove that physics have been modified somewhere by the evil v1.2 developers, that - the bastards - don't even document each and every change for the cute little user that is so frustrated and really not happy at all.

Now, if you have specific question about Replay, I'll gladly answer them. Or if you precisely list what's wrong in this Replay (if at least we had known which version has been used to record this Replay), and what should be the expected behaviour, I'd gladly have a look. We may also know why you chose to use an old build like a13.0820 and not a15.0131. But I doubt this was the purpose of this topic anyway... The real point if I read again is that according to you we have to be answerable for any modification you may find between version X and Y, isn't it? Then sorry to destroy your world of happiness, we're not.

Until then, as far as I'm concerned I consider the matter closed.

physics have been modified somewhere by the evil v1.2 developers.

as to where, why, or if it was done on purpose, we dont know.

I don't really get what all the fuss is about regarding the replays behaving different. I mean up until the most recent alpha (15.0131) external replays were a dev-only feature and couldn't even be easily accessed/used otherwise.

And the whole physics topic is an entire different matter that has already been discussed to death, as far as I can tell replays won't be a big help in revealing more about that so what else is there to say?

I think it would actually be better to change replays back again to dev-only for the next version(s), at least until the system has been overhauled with some improvements (as well as file format changes if necessary, that would also make sure that there are no compatibility issues with non-dev replays).

Abc wrote:Same with GameGauge, fps.txt's first line still reveals "<FPS> Re-Volt v1.00"

I can only assume that they either forgot to update the version or just kept using the original number because they didn't change anything in gamegauge since then.
Either way thats probably the least problematic issue right now.

enabling something that is broken wasnt a smart decision, and very much a part of the overall problems.

ThugsRook @ 10 Mar 2015, 12:41 AM wrote:enabling something that is broken wasnt a smart decision, and very much a part of the overall problems.

You know when we discovered that it was broken? The first email of Huki reporting an issue with a replay file to me was dated 01 Feb 2015. The day when we discovered the reason was 08 Feb 2015. You know the the date of the release that enabled load / save replay? It's 15.0131. That is to say one day *before* Huki reported this issue.
Now Mr. Smart Guy, what do you have to say? Still not smart enough for you?

Part of the overall problem? Which overall problem? It's so easy to make such sentences empty of meaning . So let us know now, whit clear words, what is exactly this overall problem you were talking about, just for a laugh.

You haven't even noticed that none of the version reported in this "overall problem" has the Replay save / load feature enabled except in Dev mode.

And only that illustrates the relevance of your last comment. And the one before. And the one before...

^ your updates are not properly tested, and since noone really knows what youve done we dont know what to look for.

these guys are racing, not looking for bugs. only a few of us are actively looking for problems.

full disclosure of all modifications would help



and really, what did you do that could possibly make the replays incompatible across versions? you dont know do you? thats a problem.

look i dont want to fight, i just want to fix the game. you arent exactly cooperative ya know.

I'm a tester and I would like to work closer with the devs. people who just race are "noobs". (kinda, not in a mean way either)

the point is that all this was originally built back in 1.0!!! (I don't see why you guys are trying to break and fix something that doesnt need to be fixed)

ThugsRook @ 10 Mar 2015, 10:12 AM wrote:and really, what did you do that could possibly make the replays incompatible across versions? you dont know do you? thats a problem.

Abc wrote:the point is that all this was originally built back in 1.0!!! (I don't see why you guys are trying to break and fix something that doesnt need to be fixed)

I don't think you two get it..

As i understand it, the replay system just saves the race state (cars position, speed, powerups of each car, etc...) in certain intervals.
So it's logical that any subtle change in either the cars AI, the game physics, collision system, etc will make replays from different game versions "incompatible".

Incompatible: That is, replays from version A will load and play in version B, but the races will be different.
Of course, when the next interval is loaded, the cars will be repositioned in the place they were recorded in version A, which most likely will be different from what happens in version B, so expect cars teleporting randomly.

The replays are not buggy or broken. It's just the logical behaviour when you mix differernt versions.

hi&#045;ban @ 12 Mar 2015, 12:21 AM wrote: As i understand it, the replay system just saves the race state (cars position, speed, powerups of each car, etc...) in certain intervals.
[...]
The replays are not buggy or broken. It's just the logical behaviour when you mix differernt versions.

The replay system is buggy (and it has been that way since 1.0), but otherwise you're quite right. It'd be too long to explain what exactly is wrong, but even on the same version, when a replay is viewed several times you'll notice different behaviour. And as you said, it's very logical that the there will be more differences when mixing different versions.

So then, even when replaying it tin the same version, there are still some differences?

In that case, the only logical explanation is randomness. I mean, we have circuits with different paths, in which the cars AI might have some randomness when choosing one path or another. Besides that, there also might exist some randomness in the car's trajectory when going through the same path.

We could also take into account the possible randomness of some powerups, like trajectory of rockets, water balloons, etc... All that can cause the cars to get thrown into different places when hit by a rocket or balloon.

The possible solutions are:

A- Remove all random factors in the game (AI, projectile trajectory, etc...) and make everything act in a fixed way.

B- Increase the frequency at which the replay system stores the events/states in the replay. The more frequent, the better.

hi&#045;ban @ 11 Mar 2015, 08:38 PM wrote:So then, even when replaying it tin the same version, there are still some differences?

Exactly.

The possible solutions are:

Could be, but you know, this topic is more about beating a dead horse than trying to improve the Replay in any way...

then why did you enable it?


...even when using the same version, pickups are in incorrect locations.

Replays work like a networked game: they are sync'd every x seconds, When there's no answer from the peer, the car stays doing what it been doing last and because of the timed car position poll that's why car teleporting exists and desync.... Right?

Guys, stop beating the old horse, indeed. The replay system maybe buggy, but it _works_ well enough that if you replay it with the same game version and had no random elements (I mean weapons) in-game, then there’s practically no perceptible difference with what happened in-game.
Proof? All my videos are made from replays, and I have had maybe 2 or 3 occasions in years where a crash was too random to repeat correctly in replay.

Just keep the same game version and it works quite well (can’t do much about weapons, of course...)

P.S. I wouldn’t want to live without the replay system, how else would I make videos?

btw, did you know that if you shoot weapons with sadist cheat it messes the replay really bad?