Hacker News Reader

> home servers and decentralised identity. Home servers would need to be as close to zero configuration as possible, whilst still remaining secure.

People aren't going to buy more black-box hardware they don't directly interact with. Consumers already struggle with routers.

I agree with the core concept, but I think we'll need mobile servers instead of home servers... a simple "Internet" app that installs on anyone's laptop, desktop, or cellphone. Something with a distributed/peer-to-peer file system for ubiquitous content, and peer-to-peer RSS.

The harder question is (complex) querying. How do we avoid the centralization of Google, while still retaining the functionality? I don't see how such a new service could survive if there's a regression in search.

People struggle with routers because they're terrible. Buggy, inconsistent, and their UI is full of detailed technical questions they don't know the answer to.

I imagine the box of the future is a home cloud box with a small touchscreen on it for the easiest bootstrapping ever. Enter a name and that's the new dynip subdomain and said. Enter a password and that's the password for wifi and for accessing the group shared content online, and the encryption key for cloud backup. Dont ask the user what encryption they want, just set up wpa2.

Physical panel is always admin. Let the panel admin create users and admins with their own credentials for all the usual cloud services - docs, email (at their subdomain), some kind of social networking/IM node, with easy-to-download apps for various other services like minecraft or music streaming or whatever, etc.

Complete 1-stop wifi and cloud server.

I think we are looking at this the wrong way. We are looking at it like techs. These kind of problems need to be looked at from the customers point of view (ok, I know that Facebook users are the product, not the customer but the analogy works)

Ask yourself one question: What problem am I solving with this?

Can you honestly see the Facebook-masses buying a blackbox device, however simple, just to connect to others in a decentralized way? Would your mum use it? That's your target audience.

I firmly believe the future must have a better way of doing things but we need to look at the pain points, the current problems that the current tools are attempting to solve and then coming up with something better and easier (let's keep network effects out of the equation for now).

I don't know what that is but it must be as simple as clicking something that says "Log in to <Product Name>"... not another device, however cool or simple it is.

Perhaps a router that gives you a simple question during setup, e.g. Would you like to enable SocialCloud? A major problem I could see with something like that is that I imagine most people get their routers from their ISP and I have no idea how big the router market is... I got mine from Sky and have no intention of replacing it.

> "What problem am I solving with this?"

Consider what would happen if you asked the same question about technologies that are popular today before they were popular... What problem does WhatsApp solve? What problem does Instagram solve? What problem does YouTube solve? Instant messaging apps, photo sharing websites and video sharing websites all existed before WhatsApp, Instagram and YouTube, yet they all took off in a big way. Home servers can take off despite the competition so long as there are enough early adopters willing to take a punt. Network effects can kick in after there's a small, dedicated group getting use out of the technology.

> "Can you honestly see the Facebook-masses buying a blackbox device"

They don't have to buy anything. You can run a home server on computing devices you probably already own.

Apple proved your point with Rendevous on Mac OS X. I actually got burned by its easy setup when I bought a Mac laptop for air-gapped use. They used to advertise WiFi as an optional feature, which this ad didn't have. Turns out, it did have WiFi, activated when I turned it on, and Rendevous already had it setup in the background on an open connection. I remember being confused at just how up to date the Apple Store looked. Very, briefly confused followed by eye roll and sigh that an $80 device was already compromised for use case.

Nonetheless, things like it and Time Machine illustrate just how much use one can get out of certain features if the UX is bulletproof or nearly so. Routers might be done as easily depending on the circumstances. I'm almost certain ISP's could pull that off. Give them username and password to use on a HTTPS site that downloads the right configuration into an OpenWRT router they supply with any configuration software built-in. It then prints locally-generated password and other configuration data on form for customer to safely store. Any recoveries can be done with it or through ISP.

What you think?

>I imagine the box of the future is a home cloud box with a small touchscreen on it for the easiest bootstrapping ever.

Already done with the Almond+ and others. The problem is that at best, the firmware behind the pretty screen is terrible. At worst? http://www.theregister.co.uk/2015/05/13/home_router_botnet/

IPFS already does this and exists. It is decentralized, it lets you put your own content in it whenever you want, and anyone can use it.

IE, you can simply drop an html page into your own IPFS directory, and other users of IPFS can browse to it (assuming you put global read permissions on it). It is no more complicated than a daemon and a virtual file system.

That's sort of amazing. Thanks for posting this!

Querying is indeed the real issue. I've between reading up on delay tolerant networking; it seems pretty tough to hit the right balance between making metadata generally available and crushing the network with metadata updates. Centralized directories make it easier, though still not trivial, to make content findable on a network.

There have between a bunch of proposals (eg, grokster) out there which have started off as radically decentralized, but eventually took on a layer or three of super peers to manage that metadata transfer, after the fully decentralized model failed under real traffic.

(Good datasets also seem to be a problem here. The available ones are pretty small or have a lot of caveats, which makes it tough to really see how a solution will play out is the real world.)

Take a look at Urbit, that's more what I had in mind.

http://urbit.org/

Also, if the idea takes off, I'd expect to see the functionality being built-in to home routers. However, that's not necessary for the popularity of home servers to grow.

Mobilising a server doesn't simplify it.

If it's simple enough to be mobile, it's also simple enough to be static and self-hosted. And there are advantages to that.

Replication, redundancy, and consistent presence, for starters.

You're basically describing IPFS protocol.

Very interesting idea about simple home servers used to eliminate the middle men. I have to admit, this actually seems like a viable option if it could be done right. Good idea that I think deserves more thought and should be looked into. I'm curious what the opinions of others are on this idea.

This is what I'm working on at Optik -- but forget the idea of a server. All we need is a device, a writing program, and a sync program. Everyone already has these tools in their pocket, they just have a bad interface and therefore nobody knows how to use it in this manner.

Identity is solved by real-world trust. Facebook has proven most of us are separated by less than 5 degrees.

The only thing I don't like about it being in a phone is what happens when you change phones? How's that transfer made easily? How can v you guarantee everything will be removed when they get rid of their old phone?

Also, I have only about 15 apps, 300 pictures, and 10 sorry videos and I'm constantly running out of space even with half the stearate being on an SD card. I can't imagine puerile having enough storage to run something like this on their phone.

We could easily fit more storage (and more battery) on phones, there just isn't the market demand for it. With such a decentralized system, demand may increase.

Most knowledge as we use it now is ephemeral and can be somewhat centralized safely. The permanent knowledge or at least the trusted hashes of that knowledge can travel with us.

Trouble with doing it all through phones:

- Data plans. Everyone connecting to your blog through your phone would use up a hell of a lot of data.

- iPhones are notoriously hostile to P2P apps, so that rules out about half of all smartphone users.

There are other issues but those are the most significant imo.

Aren't the main reasons everything went centralized: bandwidth and reliability?

Maybe this can be solved with good caching.

But what you don't want is your site / post going viral which will bring down your internet connection.

This can be solved with for example IPFS. When you download a file you also make it accessible to others.

This means that reverse scaling effect kicks in. The more popular the file is the less people access it from original source.

Which is why you want content-addressable decentralized storage, so the content gets distributed either peer to peer or through public caching infrastructure that's operated by ISPs, just as IP router are nowadays.

I'd say the main thing is people's inability to reason without serialisability.

Without the serialisation of actions that centralisation gives, it's much harder and hence more expensive to create things.

Ever thought about whether the assumption that having control over what your see on your Facebook feed might be incorrect?

If we're talking about the ideal world that may be true, but the real world is far from perfect. If you ask 100 people do you want to live a meaningful life, probably 99 of them will say yes. But do they? Probably 1 out of those 100 will live a meaningful life. If you ask people do you want to be told what to do, or do you want to do what you want, most of them will say they want to do what they want. But in reality most people just want to be told what to do because making decisions and being responsible for it is not an easy thing.

Coming back to Facebook feed, there's a reason why people keep using Facebook even though many people hate it. Sure it's not ideal but there will never be an ideal world. I think the reason why most "decentralization" advocates never succeed is exactly because they're being too ideal (read naive) about this.

That said I think the pendulum will swing back someday in the future surely, it just won't be by these guys. It will be from some random technology which didn't even aspire to "disrupt" the web.

> "But in reality most people just want to be told what to do because making decisions and being responsible for it is not an easy thing."

As I said before, content could be curated by apps. You could choose the apps that present the information in the way you like. The difference is, if a better app comes along you can switch to it without losing your past data because all the data would be application-agnostic, there's a greater degree of separation between the raw data and the presentation of that data.

I think it's just about network effects. Everyone is on Facebook because everyone is on Facebook.

Although network effects work the other way as well, when people start leaving social networks they collapse exponentially just like they grew exponentially.

Interestingly, Urbit seems to target squarelyat this space.

Yes, I was very interested when I learnt what Urbit did, as I think it ticks a lot of the right boxes. The only weakness for me is how it handles distribution of identity, but that's not a fundamental issue, it can be changed whilst still keeping what makes Urbit promising as a home server platform.

More people thinking about how servers! Great :)

I like the concepts of Namecoin & Keybase, IPFS and Tahoe-LAFS, etc, Tor and I2P, CJDNS, PHB's Mathematical Mesh, and so on...

Having your own server or cluster that your own devices connect to for anonymization, identity, syncing, federation, etc, would simplify so much.