[This is an original thought-piece by YK Goon]
I found out about BitTorrent browser late, but got pretty intrigued by the potential. Depending on how it pans out, a peer-to-peer browser could be a game-changer.
There's not a lot to read about the project at this point, so I can only imagine what it promises to do without official confirmation.
Here's what I think matters for anyone making a bet on it.
There Is No Server
Web pages don't get hosted on devops' carefully-manicured servers, they are found on your faceless friends' computers. This means tomorrow's Techcrunch article may not need to be hosted by a glorified Wordpress hosting-provider, and there is no such thing as server-scaling to handle popular demand.
Docker fleet management be damned.
It's About Content, Only
The entire concept is designed for content only, not applications. Meaning you're only meant to download, never upload. If there somehow is some form of uploading, I can't imagine where it'd end up in. In application terms, it's awesome is delivering for reading but got nothing to offer for writing.
What that implies is an interesting website has gotta be done with a static HTML generator, and regenerate every so often. Updates may not be seen in real-time, and the BT protocol would determine how fast changes get propagated across the network.
If somehow there's avenue to write/upload from clients, the entire prospect of server-less architecture is made null and void.
There's only one way P2P internet will get mainstream: it gotta integrate with Firefox.
I don't mean making use of Firefox as the engine (like TOR-Browser), I mean have Firefox baking the BT protocol in their engine.
Mainstream users shouldn't have to care if a URL starts with https or magnet. They just care that whatever they click on shows up. If a Facebook post brings them to a magnet link, users shouldn't have to copy and paste it into a separate BT browser.
I single out Firefox because it's the one agenda-less power player. If the political ideals of a P2P web is strong enough, Mozilla has no reason to not do it.
Webkit may also do it. But downstream players (Apple, Google, Opera) might have different idea about the politics of P2P browsing. If Webkit forces it, the project may fork, funding will dry. That's too much to stomach for an unproven protocol.
I'm sure I don't need to address Microsoft in a browser discussion.
A counter-example is TOR-Browser. It's built on Firefox alright, you can surf onion sites with minimal configuration. But it's a separate and standalone application, no one really cares enough to even know about it if it's not baked into the official Firefox itself.
What's the killer app?
For now it's about making content that doesn't go away. There's a need for the easiest and most enjoyable publishing tool that pushes to a torrent website. Essentially a tweeting client that generates a static page each tweet, have it seeded on the phone, and made available to everyone via a torrent link.
Imagine publishing a self-hosted Instragram post and showing it on Facebook.
Still, why would users bother unless they have something to hide? Perhaps they want something that doesn't count on them having a server. But if Twitter is no longer showing fail whale, why bother?
What about a server-less YouTube. Imagine an Android app that publishes your video as a torrent link. Another regular HTTP site plays your torrent video like a normal YouTube clip. Only this time there's no way for your video to be taken down (given enough seed).
Lack of Database
Without facilitating database in the BT protocol, there's no way to build real applications.
Sure, a static web app could be hosted in torrent and talks back to a central database through HTTP. But what's the point in that? You might as well host it in S3.
Given this huge gap, maybe that's an opportunity for a project on server-less database? If so, how would that even look like?
A torrent web link looks like this:
For a company intending to build a substantial asset on the torrent-web, how would they do it?
No commercial potential, no content. No content, no network.
What about TOR?
This has a different implication than TOR-Browser. TOR-Browser is about anonymizing clients, and to a much lesser extent the server. There is still only one single point for server content to exist. Onion server go down, content goes away (just ask your neighborhood darknet shop owners).
BitTorrent's promise (or what I imagined of it) is your content will persist as long as there are seeds, fulfilling the true promise of internet-never-forgets.
A few things stands in the way of P2P-web getting mainstream.
Regular internet users hardly cares, they are pretty happy getting their Buzzfeed nonsense from Facebook. There's no real reason to install a separate BT browser to look at a wasteland of poorly constructed static sites. TOR-Browser has darknet, BT needs something else.
That leaves artists and entrepreneurs to have a good reason to build on top of BitTorrent. The notion of server-less architecture is mouth-watering, but it needs to be more. More than a glorified CDN, we need to be able to build real applications on it. The idea of a server-less database doesn't sound so far fetched now.