Acrobatics Task

[Tarius]

without viewing .exe directly.

I don't know why we aren't using that straight way. Theoretically:
1) If I was not agreed the EULA (if I was not install the software or run it) or any other document, that forbid me to disassemble
2) If I didn't get any money from disassemble
3) If I will not use the code of Morrowind in any other software.

Which laws are violated?

Oh, and you're already reverse-engineer the code (making some tests), moreover, using the MWSE, external program, that modifies the Morrowind code

Actually, you dont talk about this sort of thing. We have alot of equations floating around but of course no way to absolutely verify where they came from. They could have been discovered from testing or they could have been copied from the game itself. No way to know. If you find an equation, best to either not say anything, or at least show you could have gotten it from testing.

[Epsilon]

without viewing .exe directly.

I don't know why we aren't using that straight way. Theoretically:
1) If I was not agreed the EULA (if I was not install the software or run it) or any other document, that forbid me to disassemble
2) If I didn't get any money from disassemble
3) If I will not use the code of Morrowind in any other software.

Which laws are violated?

Oh, and you're already reverse-engineer the code (making some tests), moreover, using the MWSE, external program, that modifies the Morrowind code

We've done some discussion on IRC about this.

It's not about "a certain law being broken" so much as "plausible deniability" and "being over-cautious."

I'm actually somewhat convinced (again based on the IRC conversations) that it wouldn't break any laws, but you're looking at two options:

A: We do black-box testing, which is explicitly legal, and there's no chance of legal complications down the road. It makes equation-finding go a bit more slowly, unfortunately.

B: We decompile the code, which we think is legal, but have the risk of legal complications happening later. Obviously, this would be quicker.

Barring a couple of (real) lawyers on the project that are willing to research the decompilation angle, I'll personally opt for A out of a sense of caution. The last thing I'd want is for the project to end up sunk just because we wanted to get things done more quickly.

Incidentally, last I checked, you agreed to the EULA when you opened the box. That's a legal mess in and of itself, but there you have it.

[Nandorianen]

Sadly, some things won't take "slightly longer", they'll be pretty much impossible to find out and you'll only have a very sketchy estimate of some formulae.
As for the legal thing. Well, you guys still have presumption of innocence in your countries, right? There is no way to prove that you obtained some formula one way or another. If you've not been caught physically by someone in the actual process of dissecting the .exe, that is. So it should not matter as long as you don't go and boast about your methods on every corner. And finally, there is a matter of obtaining information indirectly. If you go and steal some government secret, you are commiting a crime. If I already stole it and just show it to someone, it would be me who commits another crime, not the person I show it to. So when Hrnchamd provides another formula from half a day of debugging, you can't be held responsible for it's use. That, and I also don't see anyone antagonizing Hrnchamd, who dissects .exe for years on official forums. So relax, keep deniability and use methods that are actually effective.
This posts should be burned or hidden after this discussion ends, though, if you want to be extra cautios.)

[Tarius]

Sadly, some things won't take "slightly longer", they'll be pretty much impossible to find out and you'll only have a very sketchy estimate of some formulae.

I would say that finding jump height to within less than like .3 of a height unit difference compared to the original MW is a bit more than a sketchy estimate.

[Nandorianen]

As I already said many times, this formula needs a complete rework with all GMSTs taken into account, specifically ones for fatigue mult. But it's not the trickiest one, since it's pretty much static. I'm talking mostly about formulae involving random chance that aren't easily tested. Like haggling and knockdowns. AI can prove to be quite a challenge to replicate, too. Maybe.

[HiPhish]

Incidentally, last I checked, you agreed to the EULA when you opened the box. That's a legal mess in and of itself, but there you have it.

Actually, in most countries the EULA has no legal validity since you cannot know it before buying the product. General copyright laws still apply though, so you are still not free to do anything you want like redistributing the software on your own.

[Kirtai]

I believe what some people and corporations do is have two separate groups. One group does the reverse engineering and writes down a prose description of what they find (no code). The other group takes the description and uses it to write their own code from scratch. I think it's called clean room reverse engineering and believe is's legally defensible but it does require that the reverse engineers not contribute any code to the project.

I don't know if that would be feasible though.

[Tarius]

As I already said many times, this formula needs a complete rework with all GMSTs taken into account, specifically ones for fatigue mult. But it's not the trickiest one, since it's pretty much static. I'm talking mostly about formulae involving random chance that aren't easily tested. Like haggling and knockdowns. AI can prove to be quite a challenge to replicate, too. Maybe.

No, it does not need a complete rework, only some tweaking and some additions to it; provided of course that the other 2 GMSTs are doing what I think they are.
Convert those numbers to velocities since that is what we were messing with, instead of jump heights.

[Hrnchamd]

Acrobatics mechanics are posted at http://wiki.openmw.org/index.php?title= ... Acrobatics

Your empirical regressions and the transition point to an exponential regime were very close, it's very impressive.

[Greendogo]

Great job! I believe Acrobatics was giving us trouble because of the separation at Acrobatics = 50. Why they did that, I'll never understand.

I have a question, what is your method (briefly)? Are your findings exactly correct then? Can you find the Athletics formulas so exactly?