Tim Berners-Lee Wants the Internet Back | Hacker News Reader

Tim Berners-Lee Wants the Internet Back

bloomberg.com

24 pointsthm10 comments

I like the idea, but I wonder how this works in practice. If I grant ACME Inc. access to my “profile picture,” for example, won't they just copy the file and store it on their own servers (be it for performance reasons, tracking, ease of development, whatever)? How does this keep me in control?

It only works when the majority of good actors adopt it as a standard. Effectively an agreement to abide by it. Or, as the article notes, it's enforced as a part of policy. In Europe it's being pitched as a method to help assure GDPR compliance. Of course that likely falls apart if what would be good actors don't agree to utilize or adhere to the standard. It probably needs to be picked up as a W3C standard or something to see broad use.

I don’t see how you can prevent others from collecting your data without laws and enforcement, which seem to be hard to pass and maintain (see: GDPR and right to be forgotten. Sometimes they work, but sometimes they don’t, and they have their own issues).

The technology is already here for people to collect and publish their own data. Most people in the US can purchase their own VPS, host their own website, use federated media and self-hosted services. Lots of sites and institutions let you request your data, though sometimes the process is obscure, and you can scrape any data about you that is visible to you. The main obstacle here is that it seems the vast majority of people aren’t interested, I wouldn’t expect a layman to even know what “self-hosted” means.

I had the privilege of spending 1.5 years (2004-06) at MIT CSAIL under his guidance, where my research focused on the Semantic Web—an idea aimed at making the web more intelligent. With that experience in mind, let me share a couple of thoughts:

1. Even after all this time, I believe that the vision of a web of data and knowledge at a global scale remains incredibly compelling. It may still represent the next major evolution of TimBL’s original invention.

2. However, it's increasingly clear that this leap forward will likely be driven by machine learning rather than traditional methods like RDF/OWL.

>His answer is a digital wallet, ... wallets allow you to control exactly who sees what.

Sounds great on paper, but can you imagine how hard this will be for the majority of people ? Many people still click on LOL cats that gives bad actors access to everything. So I can see people's wallets off in the wild for all to see after the first year.

This sounds similar to how gnupg works where you send out your public key and maybe assign an expiration date. Many tech people cannot even handle this, can you imagine "grandma" ?

https://archive.ph/rNMXA

That is a great idea, but it will have be an unrelated competing format to the WWW or it won't work.

i use it, it is compatible with www.

www runs on servers and is available over http

solid runs on servers and is available over http

the best metaphor is, you know when you log into another application using facebook? that’s all over http, or www (not a real protocol, but a paradigm for indicating http compatibility of a server)

solid is exactly like that, but not everyone is there yet, so it feels broken, when really it is just lonely.

Honestly, whenever I heard any opinion by Tim, he seems like he doesn't really have an understanding of the inherent difficulties of introducing standards in the real world, and then I think he got lucky that HTML since it came around is a much simpler and emptier digital world, with more technical users.

It’s good to develop standards even if the vast majority of people won’t adopt them, so the small minority who care build on one good protocol instead of many bad ones.

Ex: ActivityPub. You have many different people making social networks, and they all can interact with each other. These “many people” are a tiny fraction of those who use HTTP, but their numbers are enough to form many thriving communities.

At the same time, ActivityPub and its services (Mastodon, Lemmy) have serious flaws in discoverability, permanence, and inefficiency. If someone were to invent a new protocol (perhaps AT, perhaps “ActivityPub 2.0” or “ActivityPub+”, perhaps something else) with a way to interop with and migrate old data, it could be very beneficial to those communities.

Furthermore, one day the world may change enough for a new standard to take hold. And perhaps it will be sudden, and/or developing a “good standard” takes a long time, so we wouldn’t have had time to if we didn’t start on a random day like today.

> It’s good to develop standards even if the vast majority of people won’t adopt them

I'm not sure I agree that his standards are that good though. I mean, there's a limited number of standards that the community in general can pay attention to, and you still have some responsibility in not channeling energy into something that can't be upgraded to meet the unexpected needs of the future.

Reply on news.ycombinator.com