Hacker News Reader

Comment by jokoon | Hacker News Reader

Whatever the solution is, I really believe data must be on user's devices, not on proprietary servers.

Open source won the battle of software. The next battle will be about data itself.

It surely implies very hard problems when it comes to standardization and how data is exchanged. It might involve something like flatbuffers. How data is exchanged, what rights you have on it, how it is made secure, there are no ubiquitous idea or software that can reinvent the web because network programming is just hard and it wont change soon.

What I think could really be relevant is a database that syncs itself like the kinds of bittorrent sync and syncthing. Once you have atomic data that is spread across users, nobody needs to rent servers, and the data belongs to the users. That is a true and real way to reinvent the web, it also solves some of the controversial problems of the internet: advertising and surveillance.

View on HN

Nobody wants their data on their own device. Nobody knows how their devices even work, let alone wants the responsibility of managing them.

What they want is what they're (sorta, kinda) being given. My device takes my data and puts it somewhere where all my other devices know how to get it. I don't wanna know how it happens, I just want magic.

We, who know how our devices work, and don't want anyone else managing our data for us, are in a vanishingly small minority.

The big companies are putting their money on clouds and centralization. And that's where it'll all go, because they have all the leverage. Twenty years from now, I'd love to reflect on how wrong I was to say this. But right now, I doubt it.

Speaking as a not-so-technical (never run a server) reader myself, what people like me want is not whether or not the data is on our own devices (that's an implementation detail) but the real issue of real reliability, privacy, and basically not having all the centralized bullshit that makes me feel helpless.

None of the big systems run by the big corporations are actually reliable long-term. They change stuff all the time, come and go, change terms, and generally mistreat everyone. So we just learn to be mostly passive and kinda just play the game. It's a weird and totally undemocratic form of extreme bureaucracy, and we have no idea how to do anything more than superficial.

So yeah, people don't want to take responsibility for all the data, but they do want the people responsible to actually be trustworthy, reliable, and serve the public interest. Is that just a complete fantasy, totally unrealistic?

The problem is that I'm not sure anyone has a way for you to get reliability, privacy and not having to deal with centralized bullshit without caring about that implementation detail.

Because if you leave it up to someone else, they can make a lot of money by removing all but reliability from what you want. So they do.

Well, you can, but you have to be willing to pay for it. That's the key problem. Users love the idea of a free service that stores all their stuff, and that's pretty clearly not sustainable from a business perspective, but it makes it hard to compete with a real service.

They'll keep their stuff on the "free" service right up until it crashes and takes all their crap with it, or the TOS is abruptly changed to delete everything older than 12 months or that's connected to an inactive login or whatever, and it's gone.

Yeah, I was counting those types of "free" services as falling in the "centralized bullshit" category. Bit of a fuzzy use of language there admittedly.

Going decentralized doesn't protect you from other people as much as you'd hope. Check out Bitcoin's scaling crisis and the DAO hack. Decentralization just takes the power away from a profit motivated corporation and gives it too a loose group of developers with their own agendas.

You don't get control unless you really do it yourself.

> Open source won the battle of software.

It has not won critical battles on the hardware, firmware, productivity, or gaming fronts. Your hardware is still holistically proprietary with baked in proprietary CPUs, the firmware is still entirely proprietary and ripe for backdoor exploitation, video games are still by-and-large proprietary and even decades old titles remain so and die in cultural death for it, and companies are reaping insane profits off Autodesk / Photoshop / Vegas etc being proprietary but their consumers "cannot live without them" and thus there is no way to liberate them.

Unfortunately you are correct. However there are many people who aren't OK with this status quo and a good percentage of them are actively working on trying to change it. There are MANY projects that aim to create an universal IoT platform out of Raspberry Pi-s and although this is 1000x times behind building a fully competitive i7 CPU non-proprietary alternative, it's still a good start.

The corporations are shamelessly playing the "they cannot live without our products" but this is already turning on them. It might not be in our lifetimes but I believe that the people in general are slowly but surely becoming aware that in fact they _can_ live without corporate / proprietary products.

At least that's what I hope for.

This is something that I'd like to see as well. I recently read about IPFS in much more detail than I did before and boy am I excited for that. Unless this technology is heavily sabotaged before it can even gain steam, I believe it'll be a backbone of this new truly decentralized network.

We do have to figure something like GFS2 + encryption, only open-source and NOT licensed / patented by any corporation.

Organizations like Storj I think get it and are a good start. They even offer you small amounts of money if you rent your own disk space for their IPFS / BitTorrent like system.

> Open source won the battle of software. The next battle will be about data itself.

Not on desktop, mobile platforms, powerful IDE tooling at the Xerox PARC level, games.

Reply on news.ycombinator.com