~sarlev: What first drew you into the idea that we needed to throw away and rewrite the entire networked computing stack?

~disden-talhes: Do I even agree with the premise of the question? I think it’s funny and fun.

I don’t necessarily want to disagree with the premise in general, because it could work. But the vision is so large. It’s such an incredibly large idea. I think it’s hilarious to even begin to mess with. When I discovered that this idea existed, I thought, “Well, that’s ridiculous.” I find ridiculous things interesting to play around with.

It’s as big a project as you could take on, especially in the internet age. We have this internet that everybody says is great, works great, and is insanely big. And then: actually, the whole thing needs to go away and start over. That’s not where most people are right now.

~sarlev: Do you have a favorite part of the system, or a way you think Urbit succeeding would change how people interact with their computers?

~disden-talhes: I really like the idea of identity. I like the idea of signing your packets—having something at such a low level be a signature. You can do so much with that. It makes so many things easier if you can get that signature to a high enough level in terms of legality, or how we socially agree on what those signatures mean.

That’s how I see the foundation of this thing. It’s a presence issue: this is my signature with you over some matter. It’s communicative. A signed document doesn’t quite go far enough.

Ultimately, I’m big into the concept of claims, and claims require identity. Online, the best you can probably do is a signature like this. It just makes sense.

~sarlev: Urbit promises persistence alongside identity—the idea of a forever computer. Is that compelling to you? What does it mean to have something where you can say, “This will last, and it’s mine”?

~disden-talhes: Yes and no. History, right? Provenance. It’s compelling in the sense that I think provenance is important. History is important. If we really drill into it, this is what identity is.

What concerns me about some parts of Urbit is that need for ownership, and the level of keeping every bit of data. It feels very Tower of Babel to me in some ways. In the struggle to keep an immutable record, you can have issues. I know this isn’t popular, but I’m a big believer in third-party arbitration. I think these things are really necessary for society to function. Sometimes that history can fly in the face of that to some degree.

But that doesn’t mean you throw it out. It’s still important.

~sarlev: What did you want to do with your personal server when you found Urbit?

~disden-talhes: My goals and ambitions had very little to do with having a personal server. That wasn’t necessarily what I was rearranging the symbols around.

Especially with currencies and things like that, I tend to think in terms of claims rather than numbers. One of the best things for personal servers would be contracts, basically, if I were going to be technical about it. I think it’s a great way to do contractual stuff. But that wasn’t necessarily what my personal goals were or are here.

~sarlev: What drew you in? Did your relationship with computers change when you found Urbit?

~disden-talhes: It definitely changed. Until I found it, I had years when I wasn’t even on the internet. I was basically internetless. Early in my life, when I was young, I was on computers a lot. Then, because of the nature of my life and work, I had zero internet. I had a computer that wasn’t connected to the internet for multiple reasons. I did a lot of audio and video stuff on it, mostly audio, but it was not connected.

Then there was life, kids, and all these things. Eventually, I had more free time and got back on the internet. A lot had changed since the AOL days. I stumbled upon Urbit then, but I wasn’t going through technology looking for it.

I was very interested in the concept of truth. This sounds like a stupid thing to spend time on, I guess, but I was spending a lot of time investigating: what is capital-T Truth? What is truthiness? What is true? That leads you to certain other things.

The concept of deterministic computing, or solid-state computing, makes a claim about what is true, kind of like Bitcoin makes a claim to what is true. I found Urbit in that pursuit.

~sarlev: Was there something in the legacy computing world that felt like a fundamental problem, but isn’t a concern when you look at Urbit?

~disden-talhes: When I was young, the internet was magical. Now some of the cryptography and AI stuff is like magic again. Maybe there was something in the middle that wasn’t very magical, and I missed it. Given a choice, I would have spent the whole time on the internet. I wouldn’t have had to play so much catch-up. But I don’t know what I don’t know. I don’t know how bad it was in between.

Community is something I definitely missed out on. When I was young, I could find very weird things I was into and people who were also interested in them. That went away, and I forgot about it.

Finding people who are interested in the same weird niche thing matters. How many people do you bump into during a day, and how many of them will also land on that specific thing?

~sarlev: Speaking of specific things--you've been working on a thing that is specific to the way you want to use Urbit. What is Noltbook?

~disden-talhes: Ultimately, it’s a chat app. It has what I would call minimum viable chat, which is quite a bit more than what other people would call minimum viable chat.

It isn’t that complicated. It is a chat app, but it has a certain set of features that need to exist together as a baseline. Everything else can be stacked on top of that. It isn’t that these things don’t already exist on Urbit. In Noltbook, we have a %pals system, and %pals has existed for a long time. It’s not necessarily about what each thing does. It’s about bringing everything under one roof without a bunch of dependencies: this is the package, period.

To some degree, it’s also an art project. It has the intent of inflicting violence on people. It’s a weapon. That’s the other end of this in terms of its being specific to Urbit.

~sarlev: As AI lowers the cost of custom software, are we entering an era of more bespoke, opinionated tools for small communities?

~disden-talhes: I think so. Outside Urbit, I built a small web app that I use for my actual work, and other people use it too. I made it in a weekend. It’s amazing. It does exactly what we need it to do. I can’t imagine not having it. We would either pay for something that did more than we needed or lose money. That little web app is exactly what you’re describing.

Noltbook is more than the specific technical things it does. A big part for me was bringing them together in one place, as opposed to what mostly goes on in Urbit now: “I make this protocol, and I make this other protocol.” I think that plan has major flaws, unless we find some of the possible solutions.

~sarlev: What can people do with this digital-weapon art project?

~disden-talhes: Noltbook is straightforward. We envision everything in Noltbook as a note. You can write notes to yourself. One type of note is a Markdown file, which we were forced to make as a sort of humiliation ritual, but it’s in there.

You can write notes to other people and share notes. If you think about sharing or passing a note, it’s just a DM or a group note. You have your DMs, personal notes, and group notes. Everything outside your personal notes also gives you the ability to make voice and video calls.

We’re trying to make notes robust and interesting—not just text. We want to share all these other things, including other people’s programs. We want to take what we can do in it and pass it around straightforwardly and easily, especially applications. Everybody likes making little apps right now, and they want to share them.

We want to think of a note as not just passing text, but passing voice, video, and applications.

The whole thing also includes contacts and %pals. This isn’t just someone you follow; it’s not quite the same thing. That all gets handled in Noltbook. There are also note types that are purely gossip and run on %pals.

~sarlev: You gave me one of the easiest Urbit onboarding experiences I’ve had. How did you make it so simple?

~disden-talhes: It never had to be this hard. For starters, these are social problems, not technical problems. They are social and emotional issues. We can cut the Gordian knot. Almost all problems are Gordian-knot problems when you really think about it, and you can just cut them.

When you came on that day, you were using Theseus. Theseus is a planet virtualizer, so you can run planets inside planets. Theseus is a fork of Pyro, which was a fork of Aqua. With Aqua and Pyro, the virtual ships were only on a network under the planet that virtualized them. With Theseus, they can reach out and go onto the mainnet.

You were on a virtual moon hosted under my test planet. The planet where I build and do most of my testing, ~magnus-magtel, was running a virtual moon inside it on the Gall agent Theseus. You basically had web access to that and hopped onto the ship.

It doesn’t have to be a virtual moon. It could be a regular moon or planet. The idea with Theseus in the future is that we could save quite a bit of compute and also have a streamlined management system, because it all runs on one ship. You can segment and keep track of things easily instead of running all these other ships.

Anybody could do this. Anybody could spawn multiple planets and use them to let people hop on easily.

~sarlev: Does Theseus require changes below userspace?

~disden-talhes: Right now, it uses a modified Vere. But, there are a couple of ways we can do this. At the time you were on, we were running a sidecar to go outside. You basically pop out of Urbit and then pop back in, so you could still do that.

I think modified Vere is probably a little easier, though it’s more invasive to Urbit. No one else on the network knows the difference either way. It would be nice if we could get an upstream change, but we’ll see. We’ll try.

We would need either a small change to Lick specifically or—which I think would be funnier—our own vane. Lick is made to hit other hardware, such as a printer, and another ship is basically another piece of hardware.

It is slightly different because you’re opening all these UDP sockets. If you have twelve ships, you need twelve sockets open. There is an argument for giving it its own vane, which I think would be much funnier. I’m pushing that route, but I’ll take the change to Lick.

~sarlev: What’s next for you with Noltbook? Are there other projects whispering in your ear?

~disden-talhes: Glurff is the next thing we’ve been working on. Ultimately, Glurff is Noltbook; it’s just a different way of experiencing it, as a virtual office space.

You have a little character. There are multiple rooms. You can walk around and use those rooms for conferences. One big benefit of Glurff is that every room where you hold a conference can be one of your Noltbook groups.

The chat can all be saved to your Noltbook group. For calls, one person could be calling from Glurff while the other person is simply in Noltbook. You can call in from these different interfaces. Ultimately, it’s still Noltbook; it’s just a different way of experiencing it.

Theseus needs a lot of work. Once I get Noltbook and Glurff to the point where I can take a whole week’s worth of little bug and error reports and settle them, I’ll be able to get back to working on Theseus. That is already a lot for me to handle—besides the fact that I don’t know what I’m doing. I only have so much clanker access and other things.

~sarlev: You’ve built something tactile: digital skeuomorphic maximalism. Why is putting all these pieces together important?

~disden-talhes: There was a composability crisis, not just on Urbit but everywhere. One benefit of my approach—and part of this project—is saying: no, we’re putting everything in one place, because this is how you get composability.

Composability requires rules, and rules require a ruler, unfortunately. You need someone to say, “No, this is how things go.” Then you either follow suit or you don’t. I don’t necessarily expect people to follow suit with what I’m doing, but that’s one thing I want to demonstrate: this is how composability works.

I understand that when you start making rules and laws, it eventually makes composability harder. You squeeze some things out. But that’s how this works. People are afraid of the ball of mud, but otherwise you have the desert of sand. You can’t scoop anything together; it all falls through. You can’t build anything out of it. The hyper-composable world is nothing, really. There are no pieces to put together.

The bricks don’t necessarily sway in the wind. They’re not as flexible as the straw hut, but these are trade-offs.

~sarlev: If you weren’t working on Urbit, what would you be doing with your free time instead?

~disden-talhes: But I am working on Urbit.