Hacker News Reader

> Yes we still have work. Infrastructure is still a big issue. If we want to reduce centralization, we need better infrastructure and equal access to it. That one is very political.

If you want to work for an open internet, this is the key thing. For open to be the default, it should be dramatically easier to implement a new thing using p2p or on a blockchain than with centralized services.

Not just easier, but preferable. When big companies (or rather new companies that can get big, as the old ones probably won't change anyway) start preferring P2P infrastructure to conduct their business and offer their services, we'll know there's no going back from a decentralized Internet.

> Patents and IP have lost their bite compared to the 80's (continuing to do so) and piracy is easier than ever.

How do you figure? There are probably more money lost from patent lawsuits or extortion than even before now. Also, didn't software patents only become a thing in the 90's? And soon we're going to have to deal with gene patents and all sort of things like that, as the gene editing revolution begins.

The TV analogy is more about people are not using internet but only Facebook and Google. I think it's more that way than the consuming / creating angle.

The way I see it, piracy was easier 12 or so years ago.

The rise of P2P, in the days before regulatory sanctions and when user tracking was less sophisticated, you could get nearly anything on P2P.

>People care more about commenting and creating content. They care about participating more than they ever did and this is increasing.

I'm not sure if that is a good thing. Have you ever been to reddit?

The entire place is practically overran with memes and inanity.

The trick with reddit is to carefully curate your subreddit list and continually clean house as the communities evolve (usually in a negative direction) over time. It's a lot of work, and the same goes for most other community-driven sites. My Facebook feed is garbage because I don't use it often enough to train the algorithm (or maybe the algorithm just sucks). My YouTube front page is only passable because I've taken the time to subscribe to a handful of channels and had made heavy use of the old (non-Spotify-only) Last.fm to listen to music I like on there. My Google+ home page consists of two people's posts plus spam because I haven't spent time managing my network. And so on.

Now that the web has more content creators than ever we also need better recommendation, search, and filtering tools. Hell, even though it's heavily moderated I still wish I could teach Hacker News not to show me the 50% of articles on here that I have zero interest in.

It's not the content that's the problem, it's the systems we use to find and consume it. Somebody wants to look at all those memes, it's just not you or I.

If you look into our biggest open platform today, the web, you see that we have almost won. We got the whole power of our PCs sand-boxed and mostly standardized at our fingertips with ECMAScript, HTTP and company. Most of the old proprietary demons are defeated with so many companies switching from native to "the web".

But with the rise of iOS and Android, new problems are rising. And if you look at the better performance and usability of apps, they also got real substance behind them.

The web has done more to centralise computing and take it out of the hands of ordinary people than any other technology. People don't even control their own data anymore, never mind having control over the software they run.

It's hard to see that as a win.

The web has done more to decentralize networked computing than any other technology, because the web is basically decentralized. Only the agreements (IETF and W3C) are somewhat centralized, but even then we are seeing a proliferation of REST APIs to encourage decentralized data management. These usually aren't exactly "RESTful" of course because we are so busy with our own startups we aren't really working towards interop as a priority.

The only inherent reason we have popular centralized services is that there is limited economic incentives of funding or adhering to a decentralized social (or business, or health, ...) network standard - the Web would support it, but we've got to WANT to build the agreements to make it work.

Tl;dr the Read side of the web is decentralized; the Write Side of the Web is currently an oligarchy but not permanently so.

The web separates endpoints into two classes: the "land-owning" pays-for-a-domain-name server class, and the disenfranchised client class where users live.

Interactions between users are necessarily mediated by a server, which is why the economics encourage centralization. You don't need "major parties" to "adhere to agreements" if users just use the software they like to interact with each other. The problem with the Web is everyone using software on third-party servers to write rows in third-party databases because everything else is technically infeasible given the architecture of Web protocols.

You need a DHT or some other kind of decentralized matchmaking service in order to actually connect users to each other. At that point using HTTP makes no sense either, and users don't have the time or resources to write complicated HTML or hire web devs to write it. Web technologies are not really useful for users-talking-to-users without server-based mediation.

And once again we stumble on the Internet's weak point we could never fix -- the DNS.

Google wanted to fix the problem and, despite their best intention, became evil in the process. Now instead of racing for a .com name, people are fighting for search terms, bending to the opaque, machine enforced rules of a single entity driven by advertising revenues.

"The web separates endpoints into two classes: the "land-owning" pays-for-a-domain-name server class, and the disenfranchised client class where users live."

That's nonsense. Anyone can get a domain name for $5 and run a server for less. Users in the early web - usually not developers! - ran their own servers just fine.

The question is why they would want to? Especially given how big the web has become. They mostly would need applications that automated these details away for a broader purpose.

In fact many stil do run servers in their home in limited contests: their PVR, their backup appliance, their network printer, their Nest thermostat, heck even their Xbox with Twitch. It's all about the software application that warrants having your own server and makes it easy to maintain/run.

"Interactions between users are necessarily mediated by a server, which is why the economics encourage centralization. "

Its more than the interactions require software that needs to be written and evolved and maintained. A decentralized social network arguably has many benefits, that would lead to everyone buying something (software to run in their devices, or an appliance for their utility closet) to hook into it.

"You don't need "major parties" to "adhere to agreements" if users just use the software they like to interact with each other. "

The web IS all about agreements (protocols), as is the Internet. Decentralized execution of said agreements lead to emergent effects. Sure, if someone writes one piece of software, everyone by definition has agreement.

But the world of the network is about N versions of software written by different people and still achieving their ends.

Today we have cases of agreements enforced by single copies of software (Facebook) on a decentralized network with many many underlying differences in software (web servers, TCP stacks, firewalls, routers, databases). The only reason Facebook exists are the prior agreements made on those lower commodified layers, and economic value moves up the stack.

So why haven't we written the software and agreements to commodify Facebook? Because these things take time for the economic incentives to work. I believe it will happen, just a matter of time. The Web's architecture encourages decentralized sharing and manipulation of information, it's just incomplete.

" The problem with the Web is everyone using software on third-party servers to write rows in third-party databases because everything else is technically infeasible given the architecture of Web protocols."

What's the infeasible part? Where you see infeasibility, I just see "work in progress", or "incompletion".

"You need a DHT or some other kind of decentralized matchmaking service in order to actually connect users to each other."

Okay, there is a good idea for part of an implementation...

"At that point using HTTP makes no sense either,"

... And a bad one.

As an example - BitTorrent actually relies on HTTP and the web to bootstrap discovery to a tracker before moving to a an (optional) DHT. The web is an essential component of the puzzle.

"users don't have the time or resources to write complicated HTML or hire web devs to write it."

Why on earth would they need to? I can buy an appliance at Best Buy today that runs Wordpress for me. Why can't that evolve into something more sophisticated and compelling , that removes the need for centralized services?

I think you're way too hung up on this concept of a server being a socio-political construct rather than a transient architectural role in an interchange. Clients can be servers and vice versa. End users already run servers today, they just don't know it. The real socio-political construct of the Internet, the nexus of control, is the protocol. That's how you blow up centralized services.

You're right that it's certainly within the financial means of anyone to run a web server; "class mobility" is not terribly difficult here. The problem is, as you state, that virtually nobody wants to have to administrate a server!

In-home devices which accept connections are not relevant as they are inaccessible without giving your home network a well-known name. Any device which is not accessible from the Internet at large is not meaningfully a part of the Internet--it's simply on a network. Of course, you can set up port forwarding and buy a domain so you can access these devices from the Internet, but that fundamentally changes the situation. This is really the important distinction here: "servers" have known names and addresses on the Internet and "clients" do not.

When you want to communicate with your friends, you do it by name or address; in the current model of the Internet as something 99% of people access via 3g on phones (with all kinds of network-side firewalling) or via their home router's dynamic IP and NAT, most users do not have any well-known address. The only way they can find their friends is by coordinating to both use some third party site. This is what it means for the current architecture of the web to encourage centralization: the most efficient way to let everyone find all their friends is for everyone to congregate at the same third-party location, i.e. Facebook.

We haven't moved away from that because the decentralized alternatives have tended to suck, requiring running your own server all the time at a fixed address; actual progress would involve eliminating servers entirely and replacing DNS with a distributed name-lookup service that's actually designed for this use-case, and then having your mutual friends play the role of mailservers when you and your peer are playing phone-tag and not simultaneously online.

I agree with your point about search/indexing, but a lot of that has to do with an economic model to fund such a service ... the advertising-driven free models leads towards centralized control and loss of privacy. One interesting way to look at this is the move Apple is making towards more search and cloud services that are funded by device sales, not by advertising and analytics.

Decentralized services have tended to suck because you always need at least a federation of centralized servers for search/discovery. But it's possible to do so in a relatively end-to-end secure way, if it can be funded without requiring spying/analytics.

Anyway I'm just saying that there's a lot of possibility out there, and the barrier really is one of coming up with a big enough application incentive to force change across multiple areas of the current deployed web.

> Anyone can get a domain name for $5 and run a server for less.

No they can't. They don't know how, in part because nobody teaches them.

No they can't. They don't have enough upload bandwidth for even a modest web server. Our links are asymmetric, remember?

No they can't. Residential IP addresses are blacklisted by most mail services. What good is a mail server that can't get its mail accepted?

No they can't. Many ISPs forbid their users to run a server —by contract.

No they can't. In many regions, most people are put behind a big NAT. They don't even have an internet connection (which by any reasonable definition requires a public IP).

Even I can't run my own server, even though I'm technically qualified. Instead, I loan some processor time from a provider I like to run a VM on their servers.

"They don't know how, in part because nobody teaches them."

Or they don't have software that does it for them.

"They don't have enough upload bandwidth for even a modest web server. Our links are asymmetric, remember?"

Last I checked, BitTorrent was still a thing. IOW, build a killer app and that will change.

"Residential IP addresses are blacklisted by most mail services. What good is a mail server that can't get its mail accepted?"

Then you can't run a mail server unless you have a cloud server that isn't blacklisted.

"Many ISPs forbid their users to run a server —by contract."

Which is a gray area that will fall apart the moment a killer app exists. ISPs don't block Twitch, they don't block Skype (which is P2P), they don't block BitTorrent, they don't block my NEST thermostat.

"In many regions, most people are put behind a big NAT. They don't even have an internet connection (which by any reasonable definition requires a public IP)."

This is barbaric and not my experience in North America.

"Even I can't run my own server, even though I'm technically qualified. Instead, I loan some processor time from a provider I like to run a VM on their servers."

Sure, that's a common workaround and I don't really consider that "centralized" computing in the manner of a Facebook or Twitter.

My point is that everyone having a server for decentralized internet doesn't have insurmountable barriers, it requires software that people want to consume that will change the way the market works. Uber is doing it for a way more regulated industry (taxis).

Well, a usable Freedom Box would indeed solve most problems. Can't wait.

It's not the "web" which causes this class separation - that happened because of NAT.

NAT is what requires a 3rd party to mediate communication. We have some hope that IPv6 will roll back this imprimatur[1] a bit. Unfortunately, a lot of people still confuse NAT with a stateful firewall that push for NAT in IPv6 where it isn't needed.

The core benefit of the internet is that it is media access. Anybody can publish because the network treats all hosts as peers. There are entire categories of software that has never been written because they don't work behind NAT, and centralized solutions became popular as a workaround.

[1] https://www.fourmilab.ch/documents/digital-imprimatur/

Almost won?

If you have to talk to the application over HTTP, then it's not your application: you don't control it, and you don't control the data it works on.

From my perspective, we're very close to losing the whole thing.

> If you have to talk to the application over HTTP, then it's not your application

Huh? Just to give one counter-example, I talk to my Kodi mediacenter over HTTP (using https://github.com/jez500/chorus) and when I'm at home the packets never leave my apartment or touch a device I don't own. The whole stack is open source to boot.

I guess you're talking about Facebook-style walled gardens, but HTTP give you more control over those (vs some proprietary protocol), not less.

I'm talking about the web, of course, and it was a blatant generalization in order to make a point briefly. Congratulations on finding a counterexample, but we're talking about unrelated phenomena belonging to entirely different scales.

> Patents and IP have lost their bite compared to the 80's

isn't that what the tpp is supposed to fix?

No. Patents have lost their bite by expiring. TPP is not going to revivify any expired patents.

right but isn't it going to give the existing ones bigger teeth?

A nice refreshing dose of optimism. Thanks