/ipfs/ - IPFS Thread #2

I agree, now let them make a platform thats not shit to build it on before they try to get a bunch of shekels out of it.

They fix bugs, conduct experiments, add features and reach out to vendors. I'm not sure what you're expecting. What area are they neglecting for a distributed filesystem? From my perspective this is the most active project for distributed filesystems. All I ever hear about ZeroNet are the limitations and security concerns, bittorrent doesn't have the flexibility to compete here, Upspin may as well only exist on Google's campus, and anything else is just a whitepaper with no working implementation.
IPFS is usable right now, in multiple languages, and has preliminary browser support. When is the last time you remember browsers adding protocol support like this? That sounds like vetted progress. The libp2p project (what IPFS relies on) were posting about the possibility of writing a kernel module for Linux, that would mean inserting their networking at the OS level.

You're going to have to point out the slack if you want to say they're not making progress. As someone who really wants something like this and has investigated a lot of the alternatives that existed before IPFS, this one is the only one I've seen that has reasonable progress, in a field FULL of hypothetical competition (the majority of projects like this never make it past concept let alone YEARS of regular commits).

If you're saying that Filecoin should be stable before they release it, that's what they appear to be doing. I don't know what you think is happening.

Maybe you're conflating Filecoin and IPFS? They're separate projects with separate teams. The only thing in common is some of the developers overlap and the parent company is Protocol Labs for both.

These things should be done in parallel, otherwise you end up in situations where IPFS is released and people don't adopt it because there's no CDN-like equivalent. For a p2p network, adoption/popularity is obviously important. They're most likely trying to finish everything at the same time.
IPFS on the desktop, server, browser in combination with Filecoin for non-tech businesses who want to host data worldwide.

For reference, Filecoin is a blockchain, the only thing it likely has in common with IPFS is libp2p, having 2 applications of libp2p during development is valuable in itself since it will likely have different needs and encounter different problems to fix before release.
IPFS is just a decentralized datastore with content addressing. You could implement a blockchain on top of it but I'm willing to bet Filecoin is Ethereum based or very similar.

No. IPFS should have more basic shit working before they build filecoin.

Totally wrong
layer on top of IPFS
filecoin.io/filecoin.pdf

fucking newline

oh no no no no

>>>/cuckchan/

Why? I can see the benefit of developing them in parallel, I can't see a reason not to do it.
What I'm asking is, how is Filecoins existence detrimental to IPFS, especially when we both agree it's a necessary component to have early on for adoption.

Implementing something on top of IPFS is a perfect way to actually test the stability and limitations of the base. How else would you do it?

Being built on top of IPFS does not mean it is similar to IPFS, it means it is an extension of it. They're implementing things on top of IPFS that IPFS doesn't have itself. IPFS obviously isn't a blockchain and doesn't have a token of its own, but you could easily implement them with merkledags and libp2p. That's what they're doing.

I think we're just having a discussion, user. I'm not interpreting the replies as "roasts", they seem legitimate. As someone who is interested in seeing decentralized networks take off, I'm very interested to hear the negatives of IPFS and what alternatives we have for it, but from everything I've seen and heard I'm banking on IPFS for now.
People are saying something is wrong but not explaining it well, I'm trying to discuss it with them to figure things out for me and anyone reading. From my current perspective a lot of the flak seems unfounded but I'm giving people the benefit of the doubt.

More importantly, this should be done before they freeze all the APIs and decide on defaults. Now is the only time to do that. Otherwise you have an implementation meant to build on top of without that having actually been tested first. I don't just mean technical testing either but conceptually too.
etc.
If they dogfood their own product now, it should benefit everyone currently and later.

lel

Because they are working on too much shit and IPFS needs more work.

Because its the same developers and company.

Exactly. Which is why they need IPFS to be good before they build more shit on top of it.