Live data from Hacker News

Re-decentralizing the Web, for good this time

ruben.verborgh.org

61–70 of 295 posts

Re: Re-decentralizing the Web, for good this time

#61
post #14

Earlier quoted context omitted.

Data centers and (paid) cloud vendors don't look at your data and the police can get into your home as easily as your VM.

Nope, they can't. There are reasonably-strong legal protections against police entering your home, because it's the main place where you actually have a legally-acknowledged expectation of privacy.

Aren't there similarly-strong legal protections against police entering a data center and accessing servers?

Re: Re-decentralizing the Web, for good this time

#62
post #20
post #8

I think the part ignored by so many is the need to decentralize the computers into the home. I'm not talking meshes or shared resources. For the majority of use cases, we don't need distributed storage, compute, etc. Just start making these self-hosted "servers", "data pods", etc as easy to install as desktop software and make it clear that they are inaccessible when the computer is off. People that aren't already wi…

Back when mobile devices were rare, skype used p2p and was very popular and novel. But then an increasing percentage of nodes were mobile with limited power, limited bandwidth, and much less storage. Even being a member of a DHT (a nice way for distributed peers to keep track of each other) is a prohibitive from a battery life and bandwidth perspective for phones. So Microsoft moved skype to a centralized service and…

I don't know, if people could buy a 'personal cloud box' that is as user friendly as an iphone, for less than a 100 dollars, a substantial number of them might possibly do that.

Re: Re-decentralizing the Web, for good this time

#63
post #40

The web cannot be decentralized without putting an end to SSL. As long as certificate-issuers are the arbiters of commerce and browsers push users to trust unsecured websites more and more, malicious governments will be able to silence people by revoking their certs. There are stronger alternatives. We need to make a push to begin using them.

Your argument might sound stronger if you didn't use the name of a dead protocol. All use of SSL has been prohibited by RFC since 2015.

I'm pretty sure there aren't better alternatives.

Re: Re-decentralizing the Web, for good this time

#64
post #11

Earlier quoted context omitted.

Try Sandstorm. Home servers are a very difficult sell (see $500 Helm) compared to VMs running in a data center and IMO the privacy difference is mostly illusory.

/me goes to sandstorm.io, click install, skip past the paid version on to self-host, see it only works on Linux, close tab ^ That is your expected reaction by normal desktop users. I mean literally download an exe and up pops up your feed ready to add your friends, or favorite businesses, news sites, link aggregators, etc given their onion ID (yes, onion ID is annoyingly large, especially v3, but discovery/identity c…

I think the ideal use case for Sandstorm is either the service model or best-case the "your family computer guy runs it for a ton of people you know". If I stand up a Sandstorm server at home this year (which is likely), I'll probably allow any friends or family to use it.

The "normal desktop user" should probably not be running their own self-hosting setup, because they will fail at backups and reliability and performance.

Re: Re-decentralizing the Web, for good this time

#65

The web is decentralized, and we already have control of our data. The problem is, people keep giving it away. I'm fairly confident that 98% of the population of the earth doesn't give a crap that their data is collected, or that they don't "control" it. This whole "decentralized web" thing is just privacy nerds trying to convince us that we need this, when really no regular consumer is asking for it.

Technologically, yes. But centralization takes many shapes (as I explain in the article: https://ruben.verborgh.org/articles/redecentralizing-the-web...).

I encourage you to read the article, where you'll see that I'm arguing from a permissionless innovation perspective, not so much privacy.

Re: Re-decentralizing the Web, for good this time

#66
post #20

Earlier quoted context omitted.

Back when mobile devices were rare, skype used p2p and was very popular and novel. But then an increasing percentage of nodes were mobile with limited power, limited bandwidth, and much less storage. Even being a member of a DHT (a nice way for distributed peers to keep track of each other) is a prohibitive from a battery life and bandwidth perspective for phones. So Microsoft moved skype to a centralized service and…

I don't know, if people could buy a 'personal cloud box' that is as user friendly as an iphone, for less than a 100 dollars, a substantial number of them might possibly do that.

These products have existed for a while and don't really make a dent.

Re: Re-decentralizing the Web, for good this time

#67
post #40

The web cannot be decentralized without putting an end to SSL. As long as certificate-issuers are the arbiters of commerce and browsers push users to trust unsecured websites more and more, malicious governments will be able to silence people by revoking their certs. There are stronger alternatives. We need to make a push to begin using them.

Your argument might sound stronger if you didn't use the name of a dead protocol. All use of SSL has been prohibited by RFC since 2015. I'm pretty sure there aren't better alternatives.

People tend to use "SSL" to mean "CA-based TLS" these days, as in "SSL certificate".

Re: Re-decentralizing the Web, for good this time

#68
post #16
post #8

I think the part ignored by so many is the need to decentralize the computers into the home. I'm not talking meshes or shared resources. For the majority of use cases, we don't need distributed storage, compute, etc. Just start making these self-hosted "servers", "data pods", etc as easy to install as desktop software and make it clear that they are inaccessible when the computer is off. People that aren't already wi…

I've found secure scuttlebutt (www.scuttlebutt.nz) really impressive. By using public servers called pubs as meeting point they can ignore lots of issues like NAT punching and discovery. I don't like that it seems to be based on an immutable log but it showcases well how much is possible with offline first, sync when possible software (just like syncthing)

I have been meaning to try SSB, I have had some success using DAT[1] to transfer large files between friends.

1: https://datproject.org/

Re: Re-decentralizing the Web, for good this time

#69
post #44

This is a load of ... You can't descentralize the web for 2 reasons: DNS and SSL. And then you have the IP organisation, the name escapes me right now.

Both DNS and TLS PKI (nothing uses SSL) are decentralized by design.

Who decides I can or cannot get foobar.tld?

Re: Re-decentralizing the Web, for good this time

#70

"In order to regain freedom and control over the digital aspects of our lives" Nothing proves his point more than:

How does running analytics on my site impair my ability to talk about what Tim Berners-Lee wants to do for freedom and control? ;-) Yes, I track how popular what content is on my site. Motivates me to write more. Please feel free to block trackers; I do that as well.

Have you considered Piwik or a similar self hosted analytics solution?

I don't care if people run Analytics, its just when they go sharing all that data with a third party that it gets troubling.

Post reply on HN