IPFS

>route it over Tor or anything else you want.
the god awful connection speed will make that irrelevant.

if only there was something like ipfs that also made you anonymous, because people are gonna be too scared to use it since you're uploading copyrighted shit using your regular IP and normies don't use VPN.

can someone explain the difference and advantage of IPFS compared to Torrents?

You know what would be good to get people interested in it? To write a at least a single damn sentence in the OP to describe it.

Attached: 26b04c1cb0089dc56e18cabc76afcd2e904a3310e02a7b7fe3942f19c216dbc9.png (716x592, 493K)

You literally cannot, if you route direct then you're identifiable and if you obfuscate it a la i2p/tor then connection becomes bad.
The best alternative I can think of would be to completely avoid storing any information on where something came from, which I don't know if IPFS does or not, but even that isn't perfect in many ways.

>see pic
>or go ti ipfs.io/

IPFS is like a global torrent for everything where everyone is part of the swarm and you only ever download what you need, never the entire "torrent" since that would encompass literally everything available in that network. And it uses advanced deduplication mechanisms to ensure that the resulting clusterfuck actually works efficiently. Add on to that some good caching schemes and the result is basically a distributed p2p file system.

Practical example: I have a movie. I add it to ipfs. The actual file js still on my machine. Now you, with no knowledge of me, search for that movie. You see the file that I added in ipfs. You can then watch the movie. There would ideally be no difference between watching it the first time and rewatching it immediately after, but because internet speeds are often limited, you'll probably notice the difference between streaming the remote file and playing back the local cache. From the point of the file system, though, all you did was request a file in both cases.
Oh and next person to watch the movie will grab it from both of us.
By default your cache will expire, but you can "pin" things to keep them indefinitely (the equivalent to keeping a torrent up to seed, I guess).

There's also a pub-sub system because the biggest problem with these schemes has always been the fact that the content is static, so you can share files but you can't really replace the server-client internet architecture. I don't remember how their pub-sub is implemented exactly but the idea is to make dynamic content well supported, and IIRC there might already be a working imageboard.

(You) too should read what I wrote for an explanation

Very flawed

bump

Attached: 1538247380937.webm (640x798, 1.13M)

Plan on dedicating at least 20 TB of data to IPFS.

IPFS as decentralized CDNs work; though only in the sense that your files only last as long as people (including yourself) want them to last.

js-ipfs implemented Circuit Relay (NAT piercing, to allow browserfags to connect to the global swarm) earlier this march.
They're working on DHT now but no word on how that's going to turn out.

Incidentally, their DHT implementation seems to suck so you can't really host a large amount of small files. Scale horizontally, not vertically.

Authenticated pubsub when?
Deprecation of floodsub when?
Also friendly reminder that cloudflare now runs a super fast gateway.

Oh wow another excruciatingly slow place that will be full of cp and people that like to say nigger

IPFS is generalized torrents. You have a million tiny bottlenecks all running at the same time.
Don't you know anything about chunks and downloading from multiple peers in parallel?
It's only as slow as your slowest peer, for however long it takes to download that chunk.

Imagine being this WRONG over hearing the world blockchain one too many times.

>Authenticated pubsub when?
Key signed pubsub? Sign me up!

I'd like to choose the algo to sprinkle data out when I'm feeling the need for speed.

Attached: 1529021139121.jpg (615x546, 58K)

>Key signed pubsub? Sign me up!
See github.com/ipfs/go-ipfs/issues/5383
github.com/libp2p/go-floodsub/pull/97

Holochain

>authenticated pubsub
As opposed to regular pubsub? What does this accomplish?

So the current pubsub implementation unironically sends insecure messages from the publisher to the subscriber?
What's the point?
Why couldn't I just embed a PGP signature or encrypt my messages and distribute a public key?

So, does IPFS work* now?

* Tera- to exabytes of big porn shares and site dumps. Where are they?

It's trivial to DoS by impersonating users with the current implementation. You could implement your own signing but that's a pain and it would be better if that was part of the protocol.

github.com/ipfs/archives/issues

> github.com/ipfs/archives/issues
How is that related? Doesn't even seem to have *any* porn collections.

So IPFS is still useless then? Every useful thing so far had lots of porn.

>no guaranteed anonymity
Nah, no thanks.

learn2vpn

See >IIRC ipfs has no built in anonymity, because it aims to follow UNIX philosophy. Last I checked they were working on tailored i2p support for that, but the advantage of this approach is you can just as well route it over Tor or anything else you want.

I don't mind using a VPN if it keeps pedos away desu

It is meant to be layered with something else.

Do you disregard the TCP because it doesn't guarantee you can't tell who the packets come from?

>not implementing a packet protocol based on zero knowledge proofs, built on third party certificate authorities

Here's IPFS in action, /hgg2d/'s entire catalog of translated and noteworthy Japanese porn games:
pastebin.com/4eu4QdAN

smart contracts lad

ipfs + blockchain is a great combination

Gonna pin that.
pastebin.com/raw/d7knhF0W

Here's my music folder.
Updates served as I can bother to fix old shit collections and files.

he said distributed, not decentralized playing at being a cohesive service.

Distributed applications have more obvious benefits. You have full control over all of your data, and therefore what you give people access to.
And it doesn't come at the cost of waiting for the entire network to reach consensus.

I seriously can't see a use for ipfs. So far, no one else has found a use for it either.

Yes. IPFS was meant since its conception as something you put under the hood of other things like applications.
You know, more of a daemon or a library than an actual program for end users.

maybe an addon/extension like Decentraleyes,
but instead of using own stored stuff it'd pull things from IPFS repositories made for it?
no, it's not as slow as your slowest peer because P2P networks aren't sync, they are async. This means you can download stuff from one peer at 2MB/s while downloading from other peer at measly 40kB/s.
because engineering and getting devs to opt for it takes time.

What does IPFS do that Freenet hasn't been doing for years?

>Also friendly reminder that cloudflare now runs a super fast gateway.
Wait, what? Cloudflare is running an IPFS gateway? Why?

Be less slow, be better designed (it's not a monolithic application, it's primarily a protocol that you can layer with other stuff to build applications with or tailor to your needs), be capable of dynamic content through the pub-sub system (Freenet is completely limited to static pages which severely limits what kind of stuff it can be used for), give more freedom on what you're storing (you aren't forced to allocate a certain amount of disk space for caching random stuff for other people), be in active development and be not full of CP

Honestly the pubsub stuff alone is a worthy improvement. Freenet can basically only be used to host static files. IPFS could be used to build dynamic applications, meaning that in an ideal world you could one day abandon TCP and HTTP and the rest of the modern web stack and more entirely to a decentralized internet.

en.wikipedia.org/wiki/Solid_(web_decentralization_project) isn't 10% as ambitious as Urbit, but from what I understand it can work with IPFS.

>capable of dynamic content through the pub-sub system
I wasn't aware of that. Is it described well anywhere?

Those are some great titles.

Attached: QmP8Wjiu6CpcqNif1TRwTctLmok1BW59px1zWPPZ2HgptC.png (517x745, 51K)

No it isn't there can be different things done about it.

Does ipfs finally have that fundamental thing it was trying to solve? Like load Web pages from the closest source which has the resources of the site?

my nigger, I too hate to say nigger. What a nigger world do we live in. Imagine how bad it is that niggers can call other niggers "nigger"

Does IPFS support multipart download for a given file just like torrents?

How long will C:\Users\%user%\.ipfs\blocks be polluted with 1 trillion files? this is bad for my SSD

Attached: please do something.png (425x520, 229K)

i don't want some chink's cp on my machine

>Does IPFS support multipart download for a given file just like torrents?
I'm a brainlet so not sure what you mean but if it helps I can tell you that each file is broken into small chunks that are individually hashed, which allows both intelligent deduplication (if you grab a large text file, edit a few lines, and then add that new file to IPFS it won't duplicate all the data that's identical, only the chunks that you modified) and of course distributed downloads (which is a large part of the point of distributed file hosting).

Ok, questions from someone who doesn't know anything IPFS.
How does searching work?
Is it resistant to copyright holders who would want to take files down?(I guess they do see who hosts the file, don't they?)

>How does searching work?
it doesn't have anything for that built in (afaik)
you could make a search engine for it presumably by scraping DHT (like how sites like digbt work by scraping bittorrent DHT)
>Is it resistant to copyright holders who would want to take files down?(I guess they do see who hosts the file, don't they?)
you can't really take a file down, the only effective method would be if all ipfs nodes shipped with a blacklist, but that won't stop people disabling the blacklist, since it's 'clientside' (in quotes because this is distributed network, there's no distinction between client and server, if one person hosts the file without the blacklist, the file remains accessible)