670

submitted 1 year ago by kolorafa@lemmy.world to c/games@lemmy.world

188 comments fedilink hide all child comments

This should be illegal, companies should be forced to open-source games (or at least provide the code to people who bought it) if they decide to discontinue it, so people can preserve it on their own.

you are viewing a single comment's thread
view the rest of the comments

[-] doctorcrimson@lemmy.today 0 points 1 year ago

I wouldn’t describe it as complex, just the bare minimum of what is required to actually preserve data with no loss. All physical mediums may degrade through physical processes, but redundant systems can do better.

I think you didn't read correctly on the statement about the most complex system failing. I'm not saying that is the most complex system, I am saying the most complex system will fail.

It isn’t hard to seed a torrent. If a group of people want to preserve a file, they can do it this way, perfectly, forever, so long as there remain people willing to devote space and bandwidth.

LMAO at the idea of comparing every bit of every portion of every seeder's copy with each other simultaneously and then cross referencing every parity file to be doubly safe, and then failing to see the chance of loss of parity during transmission of said files even after that. I will admit it would take a lot longer for a torrented file to degrade than some other forms of file distribution, but it's not going to last for a thousand years, mate.

[-] chicken@lemmy.dbzer0.com 1 points 1 year ago

I am saying the most complex system will fail.

And I am saying complexity has little to do with it and also that a system can exist that will not fail.

it’s not going to last for a thousand years

Specifically why not? What is unrealistic about this scenario, assuming enough people care to continue with the preservation effort? All nodes must fail simultaneously for any data to be lost. The probability of any given node failing at any given time is a finite probability, independent event. The probability of N nodes failing simultaneously is P^N. That is exponential scaling. Very quickly you reach astronomically low probabilities, 1000 years is nothing and could be safely accomplished with a relatively low number of peers. Maybe there are external factors that would make that less realistic, like whether new generations will even care about preserving the data, but considering only the system itself it is entirely realistic.