How exactly is this supposed to be a good replacement for HTTP? Don't you need to have everyone download and keep everything stored somewhere and be online? It sounds like it would only work on stuff that is really, really popular.
How exactly is this supposed to be a good replacement for HTTP? Don't you need to have everyone download and keep everything stored somewhere and be online? It sounds like it would only work on stuff that is really, really popular.
>>58684766
>>58684766
See https://ipfs.io/#why
>IPFS aims to replace HTTP
You clearly don't understand IPFS at all let alone how the internet works.
The same people who are hosting content now will continue to host content. The benefits come around when that host can no longer host it anymore. The users of that site will still be able to use that site since everyone else who uses it is hosting pieces of it.
>>58684837
Seeding. Just like torrents, ipfs websites need at least one seeder.
>>58684914
You clearly don't understand how bittorrent works if you think anyone seeds public torrents.
>>58684949
Thats literally the point. If no one cares about it it will just die off. Not unlike how a site could die off currently. It stands a better chance of sticking around with IPFS then it does currently though.
And this is just one aspect of IPFS. This isn't even going into the security and privacy benefits it has.
At worst, it will work as badly as http. Right now companies rent out hosting that works as a single mega-seeder. With IPFS this would still work, but you could use cached versions from other seeders. Stuff like wikipedia with lots of static content would benefit a lot.
>>58684949
But they do. I am seeding several public tracker torrents as we speak.
Seems like freenet, but.. less free.
Because you can pay someone to host your shit or do what you always did and rent a server and host it yourself. When someone else cares about your shit they'll pin it, when someone visits your shit he'll help you distribute it and reduce bandwidth until his ipfs repo gets garbage collected.
still the future coming through
>>58684997
what's the difference?
>>58684949
>>58684971
>>58684991
nobody is going to be seeding that one portuguese fansub of some obscure anime some guy made 10 years ago
[spoiler]just like with torrents[/spoiler]
>>58685138
And yes, I'm am fucking stupid to hit enter before remembering I'm on /g/
/ipfs/QmdNfADsSWJe8qYrhvovzUNztqBT9tecvmduStQoENaBDb
>>58684997
Freenet forces you to allocate nGB on your HDD and will use your computer to seed stuff you don't care about. Also it's slow as fuck.
isn't zeronet similar? which one was better for handling dynamic content?
gnunet goes in the middle
>>58684609
I can only see it working for high-bandwidth stuff like videos or downloads. VPS and Cloud hosting is dirt cheap these days and 99.99% of websites are going to stick to the server-client model.
>>58685138
The developer will, that's exactly what you have now anyway - the people who develop a website seed it. IPFS, at its worst, will be as bad as HTTP?(S) in terms of availability. The only case where a website will be abandoned will be when zero people seed it - which would happen anyway with HTTP.
>>58684609
Perfect thread to ask. I was going to make a new one, but I saw this one existed.
What is /g/'s opinion and analysis on zeronet?
https://zeronet.io/
I think that the concept of it is extremely promising and will inevitably serve as the replacement for centralized Internet over the next 10-20 years. Not so sure about the actual application of zn specifically however.
>>58686345
It uses a blockchain which sucks for various reasons. It's good that people are building solutions to distribute the Web but blockchains are too much trouble to be a good solution.
>>58684941
Technically like regular websites