Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Neocities is implementing IPFS – distributed, permanent web

ipfs.io

Post reply on HN
Print view View original

131–140 of 249 posts

Re: Neocities is implementing IPFS – distributed, permanent web

#131

Post by phatbyte »

paublyrne wrote: post #95
sdoering wrote: post #78

Earlier quoted context omitted.

Isn't the right to be forgotten something, that esp. relates to search engines? They (esp. Google) have to delete the links to this content, but the content can stay on the web forever. So this isn't per se a problem. But what happens, when I act as a "mirror" in this net and have mirrored some questionable content (who ever might define questionable is up to anybody to imagine). Will I get a shake down from the auth…

I don't believe that law should be shaped by technology in principle. It should be shaped by what the majority think is right for society. But the right to be forgotten is a flawed concept to me, and I think any tech that makes its implementation more problematic will do us a favour.

But problem is that it is. It's a EU law, so while I agree with you, I sense the IPFS will clash with this law. Let's wait and see, progress can't be stopped though...

Re: Neocities is implementing IPFS – distributed, permanent web

#132

Post by phatbyte »

paublyrne wrote: post #95
sdoering wrote: post #78

Earlier quoted context omitted.

Isn't the right to be forgotten something, that esp. relates to search engines? They (esp. Google) have to delete the links to this content, but the content can stay on the web forever. So this isn't per se a problem. But what happens, when I act as a "mirror" in this net and have mirrored some questionable content (who ever might define questionable is up to anybody to imagine). Will I get a shake down from the auth…

I don't believe that law should be shaped by technology in principle. It should be shaped by what the majority think is right for society. But the right to be forgotten is a flawed concept to me, and I think any tech that makes its implementation more problematic will do us a favour.

But problem is that it is. It's a EU law, so while I agree with you, I sense the IPFS will clash with this law. Let's wait and see, progress can't be stopped though...

Re: Neocities is implementing IPFS – distributed, permanent web

#133

Post by david_ar »

radlad wrote: post #67
Rhapso wrote: post #54

Earlier quoted context omitted.

You can free space, but how that happens is not explicit. The more people a file is requested by, the more places it is stored. If the file ceases to be requested, and all the participants decide to remove it from their caches, the file is gone. DHT records are slightly more persistent, but without a sponsor they fade as the network's churn removes the nodes that once hosted them. It is very hard to delete things, an…

Hmm, that was sort of what I figured, but that seems to me as though calling it a "permanent" web is a bit of a misnomer. There is still a place for projects like the Internet Archive.

Brewster Kahle (IA founder) has expressed a public interest in IA acting as a seeder for distributed websites, so IA definitely still has a place.

Re: Neocities is implementing IPFS – distributed, permanent web

#134

Post by phatbyte »

This looks great, congrats to everyone involved, this is really exciting stuff. I got a couple questions I didn't quite understood.

1 - What prevents someone from altering a JS file and serve it to other peers?

2 - Is it possible to obtain all the version of an object and see it's history of changes (like Git)?

Re: Neocities is implementing IPFS – distributed, permanent web

#135

Post by david_ar »

Coding_Cat wrote: post #120
hosh wrote: post #100

Earlier quoted context omitted.

The idea is that people who care most about keeping it up will pin it, or pay to keep it pinned.

If this gets implemented in browsers (native or as an extension) I would hope that bookmarking would serve that purpose. Seems like to logical choice. Perhaps it is also possible to tweak the caching algorithm to keep track of files that are rare on the network.

I believe incentives for seeding rare files are in the future plans for bitswap

Re: Neocities is implementing IPFS – distributed, permanent web

#136

Post by heydenberk »

enkiv2 wrote: post #128

Just to point out -- hypertext systems prior to the web almost universally expected permanent stable addresses (and this was, for instance, required by the various Xanadu specs). Enquire (Tim Berners-Lee's immediate predecessor to the Web) had permanent stable addresses, too. The web didn't have permanent stable addresses because the web was a hack, intended as a way to explain Enquire to people who didn't have a bac…

Or, conversely, make sure your crappiest demo is as robust and compelling as the web.

Re: Neocities is implementing IPFS – distributed, permanent web

#137

Post by dgoldstein0 »

PhaseMage wrote: post #65

I absolutely love the concept and the underlying moral reasoning for why we need IPFS! One of the first questions that came to me: Is there any way to add a forward error correcting code, like an erasure code to IPFS? I didn't find any discussion of this in the IPFS paper. This seems somewhat critical to be able to compete with modern centralized storage systems which are likely to use FEC extensively to provide the…

Based on my reading of this article, IPFS doesn't specify how you store the content - just that the content you provide back matches the requested hash. So in theory this could easily be implemented under the hood.

Though perhaps I'm misunderstanding your question - if you want multiple IPFS nodes to coordinated on storing parts of an erasure encoded file? that would be useful too. Again I think it's possible - you could build a small distributed system to host files this way. The most interesting variant though is if IPFS could have support for something like this natively - so nodes run by different users can each pitch in a little. Conceptually it's possible, but probably the protocol doesn't support it.

Re: Neocities is implementing IPFS – distributed, permanent web

#138

Post by oskarth »

enkiv2 wrote: post #128

Just to point out -- hypertext systems prior to the web almost universally expected permanent stable addresses (and this was, for instance, required by the various Xanadu specs). Enquire (Tim Berners-Lee's immediate predecessor to the Web) had permanent stable addresses, too. The web didn't have permanent stable addresses because the web was a hack, intended as a way to explain Enquire to people who didn't have a bac…

> (Let the entire existence of the Web, with its myriad flaws, be a lesson to you: your crappiest demo may well become the thing you're known for twenty-five years later, so make sure your crappiest demo doesn't suck as bad as the Web does.)

The web is amazing, and you are using it right now. It has flaws, but so does anything that achieves success at the level the Web has. How many technologies have a direct impact on billions of people's lives?

There are subtle reasons why "crappy" technology often wins. One is that worse is often better (http://www.jwz.org/doc/worse-is-better.html) and the second is that people are more prone to complain about popular technologies, which create the illusion that their alternatives would obviously be so much better.

EDIT: Note that the web is not just a technological phenomena, it's a social one. If you care about the real world, it doesn't make sense to look at it as a technological innovation in isolation. Anyone can dream up The Perfect Protocol in a research paper, it's an entirely different beast to actually do it and make it work for billions of people.

Re: Neocities is implementing IPFS – distributed, permanent web

#139

Post by noonespecial »

enkiv2 wrote: post #128

Just to point out -- hypertext systems prior to the web almost universally expected permanent stable addresses (and this was, for instance, required by the various Xanadu specs). Enquire (Tim Berners-Lee's immediate predecessor to the Web) had permanent stable addresses, too. The web didn't have permanent stable addresses because the web was a hack, intended as a way to explain Enquire to people who didn't have a bac…

So what do you call the demo that was delayed a couple of weeks to perfect before releasing? I don't know either.

Re: Neocities is implementing IPFS – distributed, permanent web

#140

Post by charriu »

phatbyte wrote: post #134

This looks great, congrats to everyone involved, this is really exciting stuff. I got a couple questions I didn't quite understood. 1 - What prevents someone from altering a JS file and serve it to other peers? 2 - Is it possible to obtain all the version of an object and see it's history of changes (like Git)?

To 1: If you alter the contents, it's hash would change, thus it's address would change, if I understood correctly.
Post reply on HN