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

Time to rebuild the web?

oreilly.com

Post reply on HN
Print view View original

21–30 of 449 posts

Re: Time to rebuild the web?

#21

Post by harperlee »

goatlover wrote: post #14
naranha wrote: post #5

SPAs have made almost all desktop applications obsolete. DevTools have deprecated view source. Through progress and innovation the web can do a lot more than 20 years ago. Sure the web is being abused for tracking by Google and Facebook but that was also possible 20 years ago with a single img-tag.

> SPAs have made almost all desktop applications obsolete. Meh, Office and Adobe products still have widespread use, as do various code editors and related tools, along with specialized applications like SPSS or Matlab. Also, things like Slack are desktop, even though they utilize web technologies. And there remain plenty of PC games which still get made. Some things you don't want to run in a browser.

Isn't it office already a "local webapp", under the hood? You can open any document in-browser through SharePoint server, have (almost?) full functionality, all documents are XML... There are lots of little signs that Microsoft is rewriting everything to be seamlessly used through a browser, and that when you open it locally you just don't notice that you are using an "office browser". Even Visual Studio has also web tech under the hood, I think.

Re: Time to rebuild the web?

#22

Post by pvg »

nickjj wrote: post #9

It's weird that the author mentions the 90s and how the web was not meant to be walled gardens, but then decided to omit AOL from the story. AOL was one big ass walled garden back in the 1990s. It was pretty popular back then. The funny thing is, back then, it was popular for 2 groups of people. First were younger kids who wanted to stir up trouble on the network (AOHell, etc.). The second were old people. My uncle i…

AOL and similar services connected to the internet but were separate from it. They predated wide availability of consumer internet service. There's not much of a parallel to what the author is discussing.

Re: Time to rebuild the web?

#23

Post by Hedja »

It feels like this is already a work in progress using Beaker Browser and Dat.

- Beaker lets you browse entire websites (dat archives) and fork them.

- It lets you create and serve your own sites directly from the browser and seed it from a server (like a torrent).

- It lets other sites create templated sites under your name for user generated content.

- Visitors by default temporarily seed your website which may reduce single point of failures, hug of deaths and costs.

With this peer-to-peer torrent-like approach, the web can become distributed again and feel more like a "web". There's still a lot of work left and maybe Beaker itself isn't the best implementation for this idea, but it's a good start.

https://beakerbrowser.com/

https://datproject.org/

Re: Time to rebuild the web?

#24

Post by bhaak »

quantumwoke wrote: post #16
bhaak wrote: post #11

Earlier quoted context omitted.

That makes no sense. RSS and jsonfeed are just containers. Both may contain real or fake news. Everybody can put up a feed in RSS or jsonfeed without any trouble.

jsonfeed is a much simpler and more hackable protocol than RSS. It's much easier to build scalable and sustainable aggregators that filter fake news in jsonfeed without the cognitive overhead of RSS. Without hackable tools we end up with the walled gardens of Facebook and Google Reader.

You sound like a buzzword machine.

A simple RSS feed can easily be hand written. But of course, you don't do that because there are so many mature tools to create or manipulate an XML feed already.

Whereas with jsonfeed there is not much yet because it is so new.

If you can hack JSON, you can hack XML as well. Even programmer novices can do this.

Edit: Reworded the last sentence that could be interpreted as a personal attack. Wasn't meant as such, I was going for the impersonal you.

Re: Time to rebuild the web?

#25

Post by pjc50 »

naranha wrote: post #5

SPAs have made almost all desktop applications obsolete. DevTools have deprecated view source. Through progress and innovation the web can do a lot more than 20 years ago. Sure the web is being abused for tracking by Google and Facebook but that was also possible 20 years ago with a single img-tag.

Not sure about that, I have 20 things open on my desktop none of which are an SPA. I don't think the heavier desktop apps are going away any time soon.

> DevTools have deprecated view source

This is not good.

Re: Time to rebuild the web?

#26

Post by adrianN »

naranha wrote: post #5

SPAs have made almost all desktop applications obsolete. DevTools have deprecated view source. Through progress and innovation the web can do a lot more than 20 years ago. Sure the web is being abused for tracking by Google and Facebook but that was also possible 20 years ago with a single img-tag.

I can't think of a single desktop application that has been made obsolete by a SPA.

Re: Time to rebuild the web?

#27

Post by jaekwon »

progval wrote: post #20
jaekwon wrote: post #8

> While many of the technologies I'd use already exist, rebuilding the web around blockchains and onion routing would require a revolution in user interface design to have a chance; otherwise it will be a playground for the technology elite We've been working on proof-of-stake and blockchain scaling/interoperability infrastructure since 2014. It all starts with a classical BFT algorithm which provides simple light-cl…

> And DNS/name-resolution on Tendermint can actually solve Zooko's triangle. How?

Zooko's triangle: Human-meaningful, Secure, Decentralized.

Tendermint provides BFT security in the likes of Bitcoin proof-of-work, except without the mining. It allows for massively replicated public decentralized databases. (e.g. BFT as opposed to Raft's FT). The Tendermint algorithm guarantees that as long as less than 1/3 are malicious, the blockchain will continue to function very well. And if there ever is a double-spend attack, we can figure out who to blame (and it will necessarily be 1/3 of bonded stakeholders).

The human-readable part is relatively simple. You can easily program a name-resolution system using the Cosmos SDK. Using the KVStore implementation there, you can commit the value of a name in the Merkle tree, and get a Merkle proof all the way up to the block-hash. And the block-hash is signed by 2/3 of the validator-set by voting power.

There's more to it than that, but that's the gist of it. Now we can really, really securely know that a name resolves to a certain IP, and know that it would be extremely expensive to attack this system.

NameCoin almost solved it but you need something like this to keep persistent state and provide Merkle proofs: https://github.com/tendermint/iavl . You can with the simple Bitcoin UTXO Merkle tree, but then you can't provide fast updates or revocation easily.

p.s. anyone know why my parent post is getting downvoted? goes up and down, up and down ;)

Re: Time to rebuild the web?

#28

Post by jacquesm »

nailer wrote: post #19
jacquesm wrote: post #6

Earlier quoted context omitted.

Developer tools are better mostly because webpages are 100's of times more complex now than they were in the past whereas the typical amount of information on the page is still roughly the same. Understanding how a modern webpage source in combination with a browser results in what you see is a pretty complicated affair, in the past you just did 'view source' and it was obvious what went where. The web is now much mo…

The point I was making was that web pages in terms of HTML aren't more complex than they were in the past. An HTML page now often has a smaller amount of elements and element attributes than the equivalent page created using older tech because every element doesn't have styling hacks and behavior added to them. HTML (stuff like GAE or BEM aside) is purer than ever.

Unfortunately removing the 'styling hacks and behavior' has split the pages up into many different components that all interact in quite complex ways. The HTML being 'purer' has not made things simpler at all. Now there are many places rather than just a few where style and functionality can be modified. And that can obscure intent quite quickly. As opposed to this is bold text which any fool (this one included) could understand.

There are obvious benefits to separating layout, style, software and content. But it isn't categorically true that this will always result in something that is easier to understand by a human.

Re: Time to rebuild the web?

#29

Post by pjc50 »

realpeopleio wrote: post #13

I agree that having a few walled gardens is a problem, but having more smaller walled gardens probably won't change much. Bad actors can still sign up, post, spam, harass, etc. on several small walled gardens. What we need is a way to better identify and handle bad actors. The internet in the 90s seemed more fun and more open. Perhaps it's because the only people really participating were not interested in abusing si…

> Each person can create a single account on RealPerson.io

That web page doesn't explain how they guarantee uniqueness of users or pay for the required ID checking?

Re: Time to rebuild the web?

#30

Post by martin_drapeau »

When I ask friends and family who use Facebook daily, privacy and security never come up as pain points. Media is blowing everything out of proportion.

Replacing Facebook with RSS or Feedly is nonsense. My friends and family have no idea what those are. Facebook makes it easy for people to connect with old friends and family members through active human interactions: like this, follow her, read that. By doing so you do tell Facebook about you and others. Its interactive and solves a pain point - keeping in touch and interacting with others, your community. Feedly does not do that.

Post reply on HN