Sooo was it bad coding?

sooo was it bad coding?

Attached: but first.jpg (1280x720, 169K)

Do you really think the devs are gonna think "But what if someone tries to build speed for 12 hours?"

no but
>converting position to integer in a 3d game ever
no wonder the collisions are so fucked up

It does not ruin the game experience in any way, so there's really no problem.

The game's glitchy as hell if you play it with a high IQ

>still fluid as fuck
Seemed to work out okay

To be fair...

It was one of the first 3D platformers ever made, of course it's going to be full of janky code and weird quirks

hmm let's see, are there any people in the house who are experts in programming nintendo 64 games? or just a bunch of virgins who torrented RPG maker and made a trapezoid in blender once

Everyone who genuinely thinks Nintendo programmed in parallel universes is a fucking idiot who doesn't even have the attention span to listen to what someone's saying in a youtube video that they claim to have watched so many times. The parallel universes are just a way of visualising how the game converts a float to a short using the modulo operator.

no idiot, this is how every game works
when bits go too high, they overflow
thats all that "parallel universes" are
most games crash, but since you can lock the camera in mario, it doesnt

this.
Every game is made this way, and every game will fuck up, glitch, and act in semi unpredictable ways when it's played way beyond the limits it was intended to be played.
A game has to work within the limits of what the in-game common sense allows, plus some basics "what ifs" so that it doesn't break at every single attempt at doing something small the devs didn't think of.
If devs were to actually think about all and any possible combination of overflow values, no game would ever be finished, because the only way to do that is to no set limits to any of the values, which defeats the purpose of making a game, other than being impossible anyway

You can't even tell during normal gameplay. Maybe they realized Parallel Universes were theoretically possible, probably not though. In any case it's certainly something they didn't even think to worry about because they didn't know there were ways to obtain QPU speed.

yea because the original position numbers are floats

Why would it be? This kind of shit isn't something you're going to find out unless you're a pancake wizard.

If you have standalone code that holds up for 15 years, then it's pretty damn good.

no. the only reason people know about all these weird glitches is because the game has been analyzed front-to-back non-stop by a small army's worth of autists for literally more than 20 years. obscure glitches that are well outside the intended use case of a program and the bounds of reasonable user behavior are not and should not be a concern for your programmers unless security is of the utmost importance (which, in the case of vidya, it is not).

haha, i understand that reference. nice meme, fellow channer :)

To this day collision and the visual model that represents the character are still two separate things.

When development started of SM64 started the concept of a 3D platform only existed in people's minds, and by time any 3D platformer was actually shown to the public SM64 was heading to the end of its development. Are things ideal? No, but when you have fuck all experience in 3D and no no examples to follow shit like that happens.

Honestly, many games probably have game changing glitches like SM64, but no one knows about them because they haven't had an autistic legion of people spending decades of their lives trying to fuck the game.

>Game almost never crashed
>Little to no glitches during normal gameplay
>All these weird-ass exploits only surfaced like ten or fifteen years later, and only under extreme/obscure circumstances

Meanwhile, some modern games will crash if you look at them too hard

Testing floor collision with shorts is more of an optimization, you don't need that much precision, and they needed to get all the performance they could out of the N64.

i haven't read up on the details of N64 hardware but i would not be surprised if the system was incapable of handling floating point math at the speeds necessary for realtime movement while also doing everything else it needs to do to run the game. the performance gains from using integers were probably necessary for the game to be feasible.

I've played through SM64 at least 6 times and I can't recall ever seeing it crash, which is pretty impressive when you consider the stability issues a lot of its early 3D contemporaries had.

NEVER EVER

Attached: maxresdefault.jpg (1326x746, 51K)

You can make it crash but you have to go out of your way to do it, like cloning too many objects until you hit the object limit or killing an uninitialized Monty Mole.

Has anyone tried to show pannen's videos to Miyamoto? I'm curious about what he would say after watching the 0xA press videos or those videos where pannen explains how various parts of the game work.

On a similar note

>First pokemon games
>Literally the most ambitious games at the time. 151 unique creatures with a large number of moves, all with different sprites, animations, and stat values
>So much was in this game the hardware could barely hold it all.
>It got to the point where they were moving where data was physically stored on the cart to fit it all
>Had to juggle values and memory bits around just to keep the file size down
>Meanwhile, at release, X and Y had an entire city you couldn't save in or else it would wipe your save

Just because you can make a game do weird shit by faffing about around the safari zone in ways that a normal gamer would never do doesn't mean the game is badly programmed. They did fucking amazing with what they had access to at the time

Attached: 220px-Pokémon_box_art_-_Red_Version.png (220x221, 110K)

He doesn't know English so I don't think Pannen's explanations would mean anything to him.

Are there any other game engines with such fascinating quirks and possibilities like this? Nothing seems to compare. Any others like pannenkoek?

>Every game is made this way, and every game will fuck up, glitch, and act in semi unpredictable ways when it's played way beyond the limits it was intended to be played.
why are computers such fucking losers?

Once I got into programming, the thought of an AI takeover became laughable to me, because computers are legitimately retarded.

Don't show them to miyamoto, show them to the five programmers of the game.

Attached: file.png (253x225, 14K)

An English guy named Giles Goddard programmed the Mario head at the beginning of the game. He'd probably be the easiest guy to reach who worked on the game.

Yeah, but he ultimately didn't do much with the core game's logic.

because it's a finite engine trying to emulate an infinitely dense universe.

He's the only one not locked behind a language barrier, I imagine.

Okay, we show him the videos, then what?

>oi guvnah dem willy wonkey nips don't know how to program like the queen's best

THIS IS THE ACTUAL FINAL BOSS

Gen 1 still had some stupid problems in normal gameplay, like Focus Energy and the 1/256 accuracy thing, but it's true nothing was as bad as X/Y

I think there's been a decline in testing due to the capacity for patching.

I don't know, I'm just saying he's the only member of the dev team that's remotely reachable.

Attached: 1490164444650.png (612x612, 629K)

I don't care what you think of Gen VI, enemy AI literally didn't even consume PP in Gen I, that shit is inexcusable.

I've heard that people ask Scott about his mario videos during job interviews as well, just because his problem solving process is very in-depth

He's a madman that needs to be stopped.

Attached: pe106DR.jpg (1651x2928, 814K)