Видео

this reminds me of the exploration Veritasium and Vsauce did a few gears back exploring random vs truly random. in short, no macroscopic natural process is truly random since knowing the initial conditions to a system and its properties theoretically makes it perfectly predictable. really hard to predict, but in theory not *truly* random. the only truly random systems we've discovered are quantum processes which have no discernable way to predict an outcome thanks to the ol Heisenberg Uncertainty Principle: the more you know about one part of a quantum process, the less you know about its other parts. which is one reason quantum computing is a big field of research. almost like a holy grail of encryption awesome video! now i wanna know how modern game consoles took things further than just using bigger numbers and faster processing

Assembly yes c++ nope python yup cobol dope

dude, this is really really cool of you to do. thank you

Shout out to one of the greatest shapes to ever exist in human history.

pseudo comes from the Greek world ψεύτικο which translates into fake Either way that was informative video

doesn't the background color use the "transparent" colors of the palettes?

This was awesome

I had zelda...you didn't have to do this..

so no like test or demo at the end of it running, no point then.... you should at least make a deal or running it like and the final result... run program, compare it to other contemporary examples, and one modern one for a laugh

Such a scmbag company. I hate that the internet loves them.

That styrofoam filler is more important than you might realize. If the PCB sticking out of the bottom of the cartridge were to be impacted in a bad way, that's damaged goods. That's lost profit. Putting a bit of styrofoam (which was all the rage back then) solves the entire problem. The plastic shell mostly protects the PCB from the sides on all 5 covered sides, except that slot opening. The styrofoam does the job well enough, while being cheap enough to not care about lost profit from adding the styrofoam. The box being bigger because of it all, is just a nice addition for marketing.

Why would you compute pi using 6400 bytes when you can hard code 960 digits of pi in 480 bytes or less?

But, if this RNG algorithm can’t start with 0 or it will get stuck, surely that means it can also never output a 0, since the output is the starting point for the next iteration? So wouldn’t the d20 code want to range from 1-20 instead of 0-19?

I did it for every single game, even the ones without saves, because I was used to do it for the ones that "required" it. I mean better do it for all than risk forgetting for the ones that matter.

haha pee pee you

Which is why the older games are better then the games of today.

Off the top of my head, I know that the two R.O.B. games generally had the converters inside. Were there any others?

Dumb.

I recommend you look at George Marsagela's paper on the xor shift algorithm. This could be implemented at any bit depth and produces a very large sample space while using no arithmetic operations simply bit shifts and logical xor. This is absolutely producable by using a memory location and a battery backed ram. It's a shame the paper wasn't published until 2003, this kind of algorithm would have been formative If used in the 1980s for video games or cryptography.

Yes.

Are you wearing a bra?

As a classic Tetris player, I find this video really interesting. Thanks to the seed being randomized before the game starts, we can have Tetris GYM and play with the same seed in two separate games.

5:02 nice easter egg :3 可愛い猫<3

I remember as a kid holding reset for the first few times and then trying it without holding reset and it still worked fine so I stopped holding reset. I never lost a save, so it seems I was lucky. :-D

It’s really mind blowing how programmers did all this math in assembly, with no float point calculations and essentially had to write their own math rules. Programming took a different kind of genius back then, especially to pave the way and come up with all of this as a standard for things like movement.

Und deshalb hatten unsere Konsolen immer diesen restbug und wir haben die Spiele ausgeblasen wie verückt

Great video, fascinating stuff

Interesting!

A lot of bla bla bla, to have more time on video, I just left in the middle of the video, please try create a content without repeating the same information over and over to create more time on videos :) our lives are busier 😂 thank you!

5:42 the lick

👾

Funny that the way I tested my RNG implementation was by writing the output to the $4011 register continuously at a decently fast rate to produce a sound with it. Since I pretty much remade what the noise channel does by default I heard the same white noise that repeats over a few seconds. But I never thought of rejection sampling, I was satisfied by masking off the bits and then looking for some value(s) to get chances of x over 2,4,8,16, etc.

I like xorshift better

EVERYTHING has a pattern. Even dice It all comes down to math. If you have the exact same momentum, starting points, movements and release/ shake times every time, you will get the same rolls.

I'm working on my own game in Godot, and seeing how older devs did things sometimes has modern world implementations. Game Maker for example is almost entirely state machine based.

Not all dice are "fair", and may provide some numbers more often than others. A good lfsr is of maximal length: providing every possible output before repeating, but some have been notably shorter - 96 values not 255 in one example - usually to the delight of the hackers and speedrunners..

2:47 6.25 / 2 is 3.125 🤯

I don’t quite understand how the socket makes any difference? Unless the pin holes themselves need to be soldered regardless of whether the chip is there or not, that seems like it was completely unnecessary.

Z = 0 Loop(Y) Add Z, X, A X*Y=A

A lot of later games did utilize the entire space, I wonder what tricks they used to keep them small on the Japanese market

Here in 'merica, we like BIG carts and boxes. If only they'd included a gun with the system, it would've been perfect.

quick question about the algorithm, if you need to store what base each digit is in, what makes the algorithm better than just storing all the digits in base 10? or is this just a proof of concept that the nes can indeed calculate the digits of pi 'by hand'?

Don't EVER listen to people telling you asssembly is "not that hard". It's fucking hard and you gotta be a real smart person to use it. Really.

0:11 - all the digits of pi?

I love it when games have rng systems that are easy to guess and exploit. GBA Fire emblem's rng system is derived from the cursor's movement and it's movement arrow's direction, which means that with savestates you can rig anything just by wiggling the cursor around a bunch Edit: it's actually.. not that. Pathing a unit's movement is done when you move the cursor and this requires a call to randomness so your unit realistically pathfinds around obstacles and units which causes a shift and changes every subsequent random number generated. Shoulda googled it

1:07 that empty() function bothers me. Just return (s.size == 0)

I tried this but got different numbers? From 87 [01010111] shifted to 43 [00101011], since we removed 1 from end, did 43^184 (or [00101011] ^ [10111000] in BIN) with result of [10010011] = 147 🤔 next from 147 [10010011] shifted to 73 [01001001], last removed bit was 1, so [01001001] ^[01001001] gave result [11110001] = 241 .. So instead of sequence 87, 192, 201, 55, 32, 173, ... I got: 87, 147, 241, 192, 96, 48, 24, 12, 6, 3, 185, 228, 114, 57, 164, 82, 41, 172, 86, 43, ... and so on. Am I doing something wrong?😮 The sequence repeats itself after 255 values as there is no 0 (as explained in the video). Two annoying things about this generator: - even numbers get just shifted to right and therefore create very recognizable patterns of descending values in the sequence.. - sequence repeats itself with same numbers, so after a specific value you will always have the same next value as in previous run Maybe this can be fixed by getting not only "random" seed, but also "random mask" for XOR operation?

I remember there was a computer experiment to overload ram to the point it spits out garbage numbers and values that are truly random. Don't know the validity of it though

So what does a NES do when you you feed it an endless string of 0s for it's seed?

In fact, it's just Americans who don't accept what the Japanese do and want to Americanize everything, they do this with games, films, cartoons, etc.