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

Decentralized Web Primer

gitbook.com

Post reply on HN
Print view View original

21–30 of 97 posts

Re: Decentralized Web Primer

#21

Post by gtirloni »

devttyeu wrote: post #20
gtirloni wrote: post #15

I don't have any experience with IPFS so I just downloaded the client and ran `ipfs daemon`. It's been using 500-1500Kbps down and 200-500Kbps up for the last 15min or so. It also has 330+ open connections. Is that expected? It seems a bit aggressive for idling. No message on the console indicating what exactly it's doing either.

By default every node acts as DHT server. If you have easily reachable node this is fairly typical load. You may opt-out from serving DHT requests by adding --routing=dhtclient parameter to the daemon.

That did it, thank you! For now that's more manageable to start learning about it.

Re: Decentralized Web Primer

#22

Post by domainkiller »

hossbeast wrote: post #19

One thing I've never understood is how IPFS would serve dynamic content. Like imagine building amazon.com on IPFS how would that work? Every single page served would have Different hash, because the content is always changing.

Wouldn't a single-page-app connected to a remote datastore work?

Re: Decentralized Web Primer

#23

Post by loup-vaillant »

staunch wrote: post #18

"The Decentralized Web" is the WWW system Tim Berners-Lee invented and that we're all using right now. It relies on the DNS and HTTP protocols. This new IPFS-based decentralized web is interesting but it's not even a measurable percentage of web traffic today. Far from being "The". What we need more than anything is to actually realize the beautiful dream TBL had for the WWW. HN is one great example but there should…

No. The web as it works right now is naturally centralised, for a simple reason: the bandwidth costs on a server is proportional to the size of its audience. This is why we need YouTube to begin with, instead of millions of people posting their videos on their personal page.

There are other factors of course: firewalls, asymmetric bandwidth, security issues, technical ignorance… But the client-server model does play a significant role.

If you want a truly decentralised web, you need to modify a couple things: get rid of HTTP, and use a P2P protocol instead. Generalise IPv6 and get rid of NAT. Give people symmetric bandwidth. Make secure software that is simple to implement, simple to use, and hard to misuse (it's not that hard dammit, just look at qmail).

Re: Decentralized Web Primer

#24

Post by marknadal »

hossbeast wrote: post #19

One thing I've never understood is how IPFS would serve dynamic content. Like imagine building amazon.com on IPFS how would that work? Every single page served would have Different hash, because the content is always changing.

The next step after DAGs is CRDTs.

They have a research page on this: https://github.com/ipfs/research-CRDT

One of them is us (https://github.com/amark/gun), David reached out the other week, now we're trying to find time to discuss with Juan how to best integrate.

Re: Decentralized Web Primer

#25

Post by ad_hominem »

staunch wrote: post #18

"The Decentralized Web" is the WWW system Tim Berners-Lee invented and that we're all using right now. It relies on the DNS and HTTP protocols. This new IPFS-based decentralized web is interesting but it's not even a measurable percentage of web traffic today. Far from being "The". What we need more than anything is to actually realize the beautiful dream TBL had for the WWW. HN is one great example but there should…

The traditional WWW has shown its weaknesses with websites that exist at the boundaries of legality like Sci-Hub and the Pirate Bay, or that probe the edges of free speech like Stormfront or forums that support Islamic terrorism.

Fact is unless you're a Tier 1 ISP yourself, you're going to be beholden to an ISP to host your content. Your domain name will be beholden to a registrar. Your TLS certificate will be beholden to a certificate authority. You're going to be subject to every one of these organizations' terms of service, and history has shown with the examples above that they will exercise that authority to pull the plug on you if it suits them (especially if they're facing a mob of public opinion or the state puts a gun to their head).

Re: Decentralized Web Primer

#26

Post by oskarth »

> To build the HTML, PDF, epub and mobi versions of the book with one command, run ./build-book.sh

Is it built as a one page anywhere already? Seems weird not to include this by default.

Re: Decentralized Web Primer

#27

Post by dustingetz »

throwaway2016a wrote: post #8

This looks really interesting and I look forward to reading it but IPFS is not the only player in the decentralized web game. One only needs to start typing "IPFS vs" in Google. Perhaps "An IPFS Primer" would be a better title?

yeah this reminds me of the 90s when netscape wanted to "be" the web, and 2017 where facebook wants to "be" the internet

this whole winner takes all mentality is taking the world in a really scary direction

Re: Decentralized Web Primer

#28

Post by saurik »

staunch wrote: post #18

"The Decentralized Web" is the WWW system Tim Berners-Lee invented and that we're all using right now. It relies on the DNS and HTTP protocols. This new IPFS-based decentralized web is interesting but it's not even a measurable percentage of web traffic today. Far from being "The". What we need more than anything is to actually realize the beautiful dream TBL had for the WWW. HN is one great example but there should…

> It relies on the DNS and HTTP protocols.

DNS is fundamentally centralized.

Re: Decentralized Web Primer

#29

Post by hossbeast »

domainkiller wrote: post #22
hossbeast wrote: post #19

One thing I've never understood is how IPFS would serve dynamic content. Like imagine building amazon.com on IPFS how would that work? Every single page served would have Different hash, because the content is always changing.

Wouldn't a single-page-app connected to a remote datastore work?

Alright but all of the complexity, and all of the content served, is in the data store, and now what is IPFS even accomplishing?

Re: Decentralized Web Primer

#30

Post by xrd »

marknadal wrote: post #24
hossbeast wrote: post #19

One thing I've never understood is how IPFS would serve dynamic content. Like imagine building amazon.com on IPFS how would that work? Every single page served would have Different hash, because the content is always changing.

The next step after DAGs is CRDTs. They have a research page on this: https://github.com/ipfs/research-CRDT One of them is us ( https://github.com/amark/gun ), David reached out the other week, now we're trying to find time to discuss with Juan how to best integrate.

That's exciting that Gun is looking up adopt CRDT. Is that what you are saying? A real open source competitor to Google Firebase.
Post reply on HN