I'm 19 so the bar for "impact on my life" is relatively low, but probably choosing to cold-email a bunch of startup CEOs whose emails I guessed at the start of 2019. It got me my first SWE internship, which helped me get full-time positions afterwards.
Reminds me of a firm I worked for. They decided to adopt Google Drive for the team of about 20 employees.
Every couple weeks an employee would mistake the synced folder for their local disk and rearrange files or delete files and then the files would disappear for the rest of the company.
Then we’d have to go into the backups and restore the files and send a message out to everyone to hold the files they were working on before saving them to the drive again.
I heard of Urbit a while back and still cannot for the life of me figure out what it does or what it's for.
The comments here are helping me figure out how it works, but it still leaves me with the question of what can I do with it that I can't accomplish with my good ol' OS or internet or (whatever the closest conventional analogue for Urbit is).
It's because it's a system trying to pitch itself as a product. The system being a top to bottom rewrite of the stack in such a way so as to sidestep the client/server relationship entirely. A lot of services rely upon positioning themselves as the server, as the big computer you have timeshared access to, and they monetise your usage. For things like photo storage, or basic communication, or permissioned access to your files, this is pointless. Any computer could do it, but the internet is itself based upon asking a server for something and getting it. And running a server sucks.
Any other peer to peer solution is partial, and therefore not able to compete with the internet as is. Urbit basically plans around an identity system that prevents spam and abuse; a hierarchical packet routing structure for those identities that doubles as a de facto governance model (due to having a vested interest in the network, the higher up you go); a kernel designed to freeze, and its entire OS on top a series of event logs that mark down computations and new states; a functional language for this "internet where every computer is a database", and the encrypted networking protocol that uses UDP while still ensuring packets always find you.
So if you wanted to, say, have a group of people set as a peer list that others can subscribe to or join, or build or use applications that lets that peer list join chats or see a set of files based upon some arbitrary marker (like giving you $5/mo?) ... you don't need a million services to spread the load, one task per service, each person joining each service. You can just use your own computer. It's a personal server platform for a peer to peer internet. It's an internet designed to resist bad actors, and to resist AOL, to resist Facebook and Google — an internet that facilitates and preserves a Usenet-esque model of small communities doing stuff together on their computers.
(Obvious disclosure that I work on the project, again.)
What you just wrote needs to be at the top of any marketing related to Urbit. It’s the first thing I’ve read that made me excited about the project.
It also sounds similar to PGP’s web of trust model, which failed horribly. But you mention that you can get paid for renting out your resources. If you can somehow make that dream into a reality, then yes, it could be a big deal. Right now everyone tries to solve this partially: you can get paid to rent out your GPU (vast.ai) or your disk space (filecoin) but you seem to imply a more unified model, where payments can be continuous and effortless based on arbitrary metrics.
The whole “planet” and “galaxy” stuff is just confusing. Sure, metaphors are useful, but a basic explanation of the problem statement and the proposed solution would be better. Your explanation here has been the first time that someone put Urbit into terms that I could concretely go “okay, I see the point.”
