Sign In
Close
My initiative for a decentralized web by Mircea
I'm still trying to take my mind off the shit that went down those past couple of days, which just now seems to be dying down at last. I'm not yet in the mood to work on anything honestly, but I should have a few good things before the end of the month so don't worry. Till then I figured I'd put up one of the questions I had planned for my watchers and Patreon supporters, which I've been interested in hearing your opinions of ever since the idea came up. This should give us something less depressing to discuss, and maybe people can also offer useful tips on how I should go about this.
Here's what this is about in a nutshell: I've been following a superb project called IPFS for the past couple of months. IPFS is a decentralized networking protocol which will revolutionize how the internet works, similar to Substratum or Zeronet but as a file system rather than a plain product. It mixes the world of websites with torrent technology and other cutting edge software; Whenever you access any file or website in your browser, you automatically seed and share it for other users to download the data from you, of course for a limited amount of time unless you choose to pin that content. The project is still in its infancy and it will be a while until it becomes mainstream, but once that happens it will be one of the biggest upgrades coming to the internet. Hosting through IPFS will offer several huge advantages, with the following being most notable:
- Hosting costs are reduced or wiped out altogether, as anyone can make a website without having to worry about a server to put it on. IPFS is a worldwide file system in which every computer viewing content shares it to others, thus it's distributed globally. All you need it so have enough viewers regularly accessing your stuff in order to seed the content reliably... if not you do have to pin and host the content until you do.
- Loading speeds will increase as demand for bandwidth will decrease. With the existing internet, you download the data from a central server whenever you access a site... with IPFS however, you instead get the content of a website from other users who have recently viewed it! If any of them are close to you physical location and have a decent connection, you can expect much faster speeds than any official server would normally offer. Imagine you just watched a Youtube video and your computer caches it for a day, then your next door neighbor watches the same video during that day... instead of downloading it slowly from far away, your neighbor will download it from your machine in a matter of seconds.
- Censorship by any third party is rendered technically impossible. Normal websites can be taken down or forced to moderate themselves because they're hosted on a central server... pull the plug on that server or intimidate the admin running it, the whole site goes down. In IPFS content will continue to exist as long as at least one computer on the planet still seeds it... thus the only way to take something down is to go after potentially thousands of people across the entire planet, which is beyond impossible for any single government or institution.
I suggest looking up IPFS and seeing the in-depth demos and descriptions explaining how it works. You can get started from their official website if you are interested:
Now onto my actual idea: I wish to do my part in supporting this movement, by building my own piece of software that will run on top of IPFS. The plan is to basically create a platform which will serve as a free alternative to both social media (Facebook, Twitter) + art websites (Furaffinity, Deviantart) + video platforms (Youtube) + other. It would work like this: The platform itself consists of a general purpose interface (JavaScript file) which simply receives a link to a settings file (your profile) and translates it into a readable HTML page. Every user creates a profile by generating a json file containing their settings and links to any media, that file is literally your entire profile! Each profile you create works like a site of its own, which communicates with other profiles through this interface to offer certain functions (search, notifications, replies, etc). Profile and post tags are used to both follow and block content using a scoring system, and you can also use external lists maintained by third parties... this will make sure that everyone can watch or avoid anyone and anything they please. The site will be flexible and modern, offering customizable layouts and colors and banners and more.
This initiative is primarily intended for those of us who want to be free, at least in the online world when it's not possible outside of it. Once such a platform will be constructed, we will have the internet we've always dreamed of: A place where no government or corporation can silence us, where we can truly express ourselves without fear of being persecuted, where we follow our common sense instead of a TOS or AUP based on someone else's vision of right and wrong. As every profile is treated like a separate site, you are the one and only moderator of your page... there is no such thing as admins deleting your posts or banning you, no one can have control of the network itself. This does NOT mean there is no protection and anything goes: It only means you are in charge of deciding what you want to see for yourself. My approach will make it possible to block any tag or profile you don't want to connect to, and further more I'd be designing it to allow external files to be used as blacklists so that third parties can dedicate themselves to keeping track of offending accounts in case you trust that their lists are accurate for you.
Unlike other projects, I decided to ask for feedback before getting started on anything. That's because this is different from what I normally do, and I might have less time for game development and animations if I also dedicate myself to this task. The main question is how much you care for this: Do you want an alternative to Facebook or Furaffinity or Youtube which is open-source and where you are in full control of your data? Mind you, once IPFS goes mainstream, others will probably attempt something similar to my idea here... however I feel I have the opportunity to start one of the earliest projects in this regard.
Journal Information
- Views:
- 146
- Comments:
- 2
- Favorites:
- 0
- Rating:
- General
Comments
-
-
That all being said, I think that it'd be interesting to see this sort of thing developed (there's a need for something along those lines, I think), though I can't speak for the people who are using your game or watching your animations.
Link
9df90f5f
As IPFS seems to be something of an improved version of Freenet, with some relatively minor changes in purpose (in particular, putting less focus on the “secrecy” aspects) so this seems to be a heavy parallel of the various Freenet social applications, in particular Sone. With this as something of a guide, I have several assorted thoughts on the technical aspects of this:
• I recall the Freenet social system as using CAPATCHAs to prevent spam from going into the system. If this would be to get popular, spambots more than undesirable but fully authentic users would be a problem. Last time I looked, the Freenet CAPATCHAs looked like they’d be easy to solve automatically. (And, the only widespread apparently effective CAPATCHA service I see today is Google’s RECAPATCHA, which seems to function by gathering as much information possible on the testee and by being secret in how it works, which is not something you could have in a system that is private or open.)
• I’d be interested to see how you’d do search and replies. I recall that Sone does the latter, though I do not know how, and there does not seem to be any concise documentation for this. There's the question of e.g. whether you'd want to let a person moderate the replies that go on something they write, which is probably tied to the whole problem of reputations and sending them in the first plsce. With search, I'd think that you could either have some system of published indices (or just directories) in each interest group or else for each page or item representing a post, user, etc., you could look at the connected items, maybe filter or weigh them based on tags, and then keep doing this recursively until you had enough results.
• It's inherent to IPFS rather than to your thing, but I am concerned by the use of IPNS. Especially with social uses, it seems to be the only way to update something frequently, but it is ephemeral and tied to a specific node. You probably could get around this if you were clever enough (e.g. each reference to a person or item making a link both to an immutable copy and a mutable copy of that thing automatically), though.
It seems to me like the main problem with these things is social rather than technical, however. To use a thing, people have to be motivated to use it. It is obvious why people are motivated to use other services:
Facebook and Twitter are easy to use (to the point that they often hide detail in order to make their use easier) and have a lot of people and interests already on them. They have designed their services with a lot of services to talk to other people and to make you want to keep using their services. They are both controlled by central organizations, who can advertise, control their public image, etc.
Youtube is easy to use, and has a lot of people and content already there. It is easy to get entertainment there. Even purely as a video host, it is easy and well-made. You do not have to pay money to put videos on Youtube. Even IPFS development itself uses Youtube to host their videos. It is controlled by a central organization, which can advertise, control its public image, etc.
Bitcoin is a popular thing more like IPFS than Facebook or Youtube in how it’s structured, and people use it. But their motivation is obvious, and that is that it is (arguments over whether it really qualifies notwithstanding) money. People like to have (as shown by this recent bubble) and use money. Even though it is not so much as centrally controlled as they are, Bitcoin, like Facebook, Twitter, and Youtube, has something to entice large amounts of people to use it, effectively an advertisement, in its potential as an “investment” (again, arguments about this notwithstanding), and in all the talk about this that gets generated. This has created a change in the public perception of Bitcoin; while there used to be a perception that the only use of Bitcoin was buying illegal things, there is now one that it is relatively “normal” to have and use (though, as a society, we’re far from paying for our lunches with it).
Tor is more like IPFS in that a “don’t touch me” mentality is even more inherent to its use than Bitcoin (where, in practice, most users wouldn’t care about a proprietary client taking up a majority of the network, or flaws allowing control of what information gets passed around, on principle; only if these things threatened what they use it for, buying and selling Bitcoin, would you see an effect on its overall population). I have a harder time gauging what people use Tor for than I do with Facebook, Twitter, Youtube, or Bitcoin, mostly because there’s not as much use of it, and so I don’t encounter people talking about it if I’m not looking at something related to it in the first place. But, I think it is reasonable to say that most people who use Tor are accessing the Web, either to get around website blocks, to prevent their Internet provider from spying on them, or to make themselves anonymous on the Web. In contrast to IPFS, there is an enormous amount of information, people, etc. available through Tor. I am perhaps cynically motivated to think that most people using Tor do not do so on principle, but rather because they find a website blocked by their service provider, country, etc.
What this is leading to is that most people do not see the things that IPFS offers, mostly resistance to information getting lost, either intentionally or unintentionally, as goals in themselves. People will not be motivated to put material that isn’t at risk of censorship (The Great Ocean Liners comes to mind) on IPFS on nothing but principle. (And, even given the choice between the two, assuming equal difficulty, etc., IPFS doesn’t offer the potential to profit from advertising money.) This means that IPFS will presumably remain the domain of two groups: the minority of people that considers longevity and resistance to censorship to be goals in themselves, and also of those people that are putting out material they think might be otherwise censored. And so because IPFS does not have the advantage of having a lot of people already using it and a lot of interesting content already on it, there is no easy motivation for people to use it if they can just use conventional services, anyway. Though this problem is its own cause, it is, I think, the major reason why you won’t see adoption.
As I’ve said, Freenet presents an example of a mature service that is basically similar to IPFS. Although Freenet seems to consider privacy and anonymity goals much more than IPFS, it is similar, I think, in what motivates people to use it, and so in what you’ll find on it. As I recall, from my brief time looking around there and from what I’ve heard about it, Freenet mostly hosts bland, public things of interest to the sort of people who will use it out of principle or because they think it seems interesting (information on computer security, copies of all the XKCD comics, etc.) and (completely separate, as far as I ever saw, from that first group) material that would probably get taken down if it was hosted on the regular Web. This is hardly the sort of thing you’d want there to convince most people that the benefits of using it justify the difficulties.