Note: this is my first draft and I think a dumb idea ;)
Off the back of the demo I created for Serverless Sync in Web apps I got thinking about a number of other different ideas that you could build.
I am not a smart man, nor a learned man and what I am about to talk about I am pretty sure other people have thought about, but I want to at least jot my ideas down.
I was thinking about social networks, impact and reach in the context of distributed networks. Specifically, I was thinking about BitTorrent and the lifetime of content on the network based on the premise of If a tree falls in a forest, but in the digital age: If you create some content but there is no one to see it, does it exist?
BitTorrent is incredibly interesting. The torrent file defines an identifier to an immutable piece of data and the trackers can index them to help other peers discover who might have the data. Therefore, if you just have a torrent file but no peers, the content doesn't exist and the torrent file is just a null pointer.
Now on to what I was thinking.
You have a site that can't accept any input other than that generated directly inside the sandbox of the site — The first version of medium.com was like this, without an API you could only create content in the site — the content that you generate will only be hosted on your machine and not on a centralized server. Every time that you create some content it gets added to a reverse chronological list of content that you have created.
Seems pretty simple. Now, let's assume that the data only lives for the lifetime you are on the site. If you close the browser or refresh the page the data is no longer there.
Without out local storage or remote storage, everything is ephemeral. It dies with the host.
Now, if we connect two machines together so that data created on one synchronises with the other it gets a little more interesting. Individually, either could refresh or close and when they came back online the data would re-synchronise. However if both close then all the data is lost.
There has been some research that studies popular files popular files and it turns out they replicate incredibly well, but at somepoint (looking at the current state of torrents) it looks like everything disappears.
Why am I even thinking about this?
I quite like the idea of BitTorrent (still) but for legitimate data scalability issues.
I like the idea of domain-constrained content created in an app deployed through BitTorrent.
I like the idea of emphemerality, kind of like how SnapChat had a short time slot available to view the content, I like the idea of you can view the content but if no one has it any more it is gone.
What's the idea?
I started thinking about creating a self contained podcast network via the web
and BitTorrent. I was wondering
if you could build a network or stream of all the created content
generated locally by a user on a site that gets syndicated via the network of people
following that person and therefore lives entirely on everyone elses own machines.
Right now, we would need a cental host to store the list of pointers to the content that
is created. This is pretty much how trackers worked in the past when wepeople
were pirating movies and music in the early 2000's. And if you take a quick glance it
is essentially a tracker, it records a list of torrents that have been created
through the site and users download the content and it's not a million miles away
from Spotify doing it's p2p distribution of content.
Ideally you don't want to have a central tracker, so you would need to maintain a list of the latest content created locally or on the network and have that be discoverable to everyone.
I currently have no clue how to do this. The logistics of getting files to people seems like it is done already, however getting the list of files to people is harder. I initially thought about creating a single addressable torrent file, but you can't update that and have keep consistent hashes....
Personally I am quite worried about this idea. Whilst it is neat, you could never easily vet or see the content that was created through it so it sounds like the ideal place for criminals :\
Thanks for listening to this random stream of thoughts! I'm off to play a bit.
Note: I thought about Diaspora etc, but they seem to centralize around a set of servers or owners rather than distributing across each participant in the network (I could be wrong of course)
I lead the Chrome Developer Relations team at Google.
We want people to have the best experience possible on the web without having to install a native app or produce content in a walled garden.
Our team tries to make it easier for developers to build on the web by supporting every Chrome release, creating great content to support developers on web.dev, contributing to MDN, helping to improve browser compatibility, and some of the best developer tools like Lighthouse, Workbox, Squoosh to name just a few.
I love to learn about what you are building, and how I can help with Chrome or Web development in general, so if you want to chat with me directly, please feel free to book a consultation.
I'm trialing a newsletter, you can subscribe below (thank you!)