MichaelJ

https://geyser.fund/project/gitcitadel

Building the library of Alexandria

Public Key

npub1wqfzz2p880wq0tumuae9lfwyhs8uz35xd0kr34zrvrwyh3kvrzuskcqsyn

NIP-05 Address

buttercat1791@gitcitadel.com

NIP-57 Address

gitcitadel@getalby.com

Profile Code

nprofile1qqs8qy3p9qnnhhq847d7wujl5hztcr7pg6rxhmpc63pkphztcmxp3wgpp4mhxue69uhkummn9ekx7mqpz4mhxue69uhkummnw3ex2mrfw3jhxtn0wfns0r5hzj

Publishing to

nos.lol

nostrelites.org

theforest.nostr1.com

Author Public Key

npub1wqfzz2p880wq0tumuae9lfwyhs8uz35xd0kr34zrvrwyh3kvrzuskcqsyn

Show more details

Last Notes

2025-05-12 17:41:08

- reply

You must have passed the filter, because I can see it.

2025-05-10 20:22:30

- reply

I'll test with that publication as I work on improvements, then.
Do you know your average internet speed, by the way? I can artificially slow down the network in my browser, for testing purposes, and it would be good to have a benchmark for that.

2025-05-10 19:50:57

I'd love to hear how loading performance is for anyone with a slower Internet connection. The work we did should provide a smoother experience even on slow networks.
#nevent1q…uk68

2025-05-10 14:08:43

- reply

Start 'em young!

2025-05-09 20:26:17

- reply

It's like a lot of tools: once you set it up right it's powerful.

2025-05-09 04:30:51

- reply

I heard that his previous role had him travelling widely and meeting a lot of people, so he may have been a familiar face to many of the cardinals.

2025-05-08 18:29:54

- reply

I was astonished it's an American.
Cool Pope name. Leo XIV.

2025-05-08 12:46:02

- reply

And God is in control always

2025-04-23 13:17:55

- reply

We can remove it or I can fix ToC in my next PR 😉

2025-04-17 12:59:08

- reply

This seems like a great GitRepublic feature.

2025-04-14 13:45:09

- reply

Not deep enough for hellthreads 👀

2025-04-14 13:11:05

- reply

Yeah, we're calling our solution for git on Nostr "GitRepublic."
@nprofile…nfns has already done some of the legwork to figure out how to package and distribute .deb packages, so we'll be using that format.

2025-04-10 14:06:49

- reply

Plus if you want ML features down the road, Python has over and above the best support for ML toolchains.

2025-04-10 14:06:02

- reply

Python is a good language for this sort of thing; it clears away distractions so you can focus on the actual parsing part.

2025-04-10 03:33:20

- reply

It's a big umbrella

2025-04-10 03:32:52

- reply

TBH, it's hard to hype it up when it's still a big construction zone. We're all waving our hands saying "you should see the architectural plans, they're great!" but all people see is a forest of timber and mounds of dirt.
Like you, we all have 9-5 jobs, so this is our slow-and-steady hobby project.

2025-04-01 23:18:28

- reply

Yeah versus Alexandria, which is aimed at desktop users.

2025-04-01 23:17:39

- reply

Yeah the LLMs don't know a ton about Nostr yet. It's good to be on the bleeding edge.

2025-04-01 16:09:48

- reply

Yeah pretty much

2025-04-01 16:09:29

- reply

Awesome, looking forward to exploring those possibilities.

2025-04-01 16:09:03

- reply

Yeah a good SDK ought to be able to optimize those issues away.

2025-04-01 13:24:54

- reply

It's also true that connecting to one relay is lightning fast, but waiting for data streams to resolve across several relays can quickly slow things down.

2025-04-01 13:23:31

- reply

I've used Claude a bit to help me explore and learn C++ language features, but I don't trust it to write C++ code as much as I trust it to write TS or Python.

2025-04-01 13:22:18

- reply

If you came in with an architecture/component diagram and a clear set of requirements, you could probably get pretty far from scratch just by prompting. In that case, you've already frontloaded the design and code organization work, and now you're just handing that design to a junior dev (Claude) and saying "go do this."

2025-04-01 13:19:41

- reply

I loved the article, thanks for sharing!
One thought I had as I was reading is that the author focuses on the potential WebAssembly offers to make _browsers_ more powerful. However, we could turn this model around and think about how WASM can make desktop apps more powerful.
A desktop application could contain a runtime for WebAssembly binaries, thus defining a sandbox for other developers to write extensions to said desktop app _in any language_.
I think extensibility will be a key feature for Nostr app integrations.

2025-04-01 13:00:07

- reply

Yeah raw storage is the cheapest resource these days.

2025-04-01 12:59:33

- reply

Are you working with WASM too?

2025-04-01 04:32:06

I've been spending a lot of time learning about WebAssembly lately, and I think it will be a critical component of Nostr's future.
WebAssembly makes fast computing possible at the edge. What does that mean for Nostr? Since Nostr is decentralized, and not all Nostr relays are created equal, it falls to the client app developer to implement inbox and outbox management, content recommendation algorithms, and so on—types of tasks that have traditionally been delegated to central servers.
JavaScript running on a browser simply won't be quick enough for many use cases.
This is one reason why @nprofile…m3j7's planned Nostr SDK, named Aedile, will be built from the outset for a WebAssembly target. With the right technology, Nostr can share the benefits of cloud computing while maintaining flexibility and keeping control in the hands of the end users, that is, all of us.

2025-03-23 15:02:30

- reply

My theory is that a lot of engineers and science-y types get into art because it takes their minds from the abstract (where they usually live) down to the concrete. That's part of why I started learning guitar a few years ago.
It could also have to do with the affinities between science and art/music, in that they're both mathematical.

2025-03-22 03:17:03

- reply

My prediction is that vibe coding will give us "fast fashion" but for software.
There will be a sharp dividing line between the overnight apps that are only meant to work for a few months or years, and the handcrafted artisan apps that will power all the apps and last for a decade or more.

2025-03-22 03:11:36

- reply

Autistic -> artistic 😀

2025-03-22 03:05:52

- reply

Oh and maybe @npub176g…q836

2025-03-20 13:04:25

- reply

One advantage of the Git+Nostr integrations that people like @nprofile…dsw5 have been building is that it removes what was a single point of failure.
Today, if GitHub were to go down, everyone's work would grind to a halt. With Nostr, we can make it much easier to discover and manage repos across a network of smaller gitservers.
With that in mind, the more the merrier!

2025-03-19 13:06:39

- reply

Excellent point! We'll need to make the onboarding easy and fun, to retain users.

2025-03-19 13:01:47

One of my goals with #Alexandria is to build an app that I can explain to people who have never heard of it without once mentioning Nostr.
It should stand on its own merits as an app.
#nevent1q…8a0t

2025-03-19 02:06:31

- reply

Yeah code is on the gitserver, PRs are on the relays.

2025-03-18 15:34:46

- reply

Good point. Ngit can already handle a good deal of the collaboration, and our team will build more collaboration tools as we work more on git stuff. We just need a place to stash code that's not GitHub and that doesn't require reinventing git in the form of Nostr events.

2025-03-18 14:11:23

We host Alexandria and our other projects on a private git server using OneDev.
Would other #nostrdevs be interested if we began providing a public git server as a (probably paid) service for others to host their projects?
Our goal is to break dependency on GitHub, and this would be a step in that direction.
#nevent1q…cch8

2025-03-15 04:06:00

- reply

Claude tends to perform best at programming tasks, both in formalized benchmark tests and in anecdotal developer experience.

2025-03-03 21:41:11

- reply

Just told my coworkers my wife and I are expecting, and one was like "This means you're a real Catholic now." 🤣

2025-03-03 21:34:25

- reply

Reminds me of when one of my coworkers and I realized we went to the same mass time at the same church, and we'd never noticed each other there before.

2025-03-03 18:44:54

- reply

It's like color -> colour, we've got letter inflaytion

2025-03-02 01:52:11

- reply

Fair enough. I guess sometimes I find giving part of my brain music to focus on frees the rest to concentrate.

2025-03-02 01:48:48

- reply

Lol I know the feeling.
Try listening to some music, or else put your phone face-down and sit there till the groove kicks in.

2025-03-01 18:54:49

- reply

The world is realigning. Russia and China no longer want to recognize the US as the leader of the international order, and it seems to me that, under Trump, America is about to step back into a smaller sphere of influence.
Is that a good thing for America? For Europe? For the world? It's not clear. What is clear is that the long liberal years of the 20th century are fast drawing to a close.

2025-03-01 17:35:46

How come stuff like this doesn't happen nowadays?
#nevent1q…8de4

2025-02-26 16:17:37

- reply

True, but the redundancy lets clients cross-check for errors. For instance, if a "w" tag says "BTC" and "bitcoin" but gives a Lightning address, the client would notice that right away and could act accordingly.

2025-02-26 15:54:44

Slack is experiencing some outages. Remote workers everywhere get the morning off.
(It's me, I'm remote workers.)

2025-02-26 15:40:41

- reply

I like the idea of using a single-letter tag for indexing. I'd suggest just using @nprofile…ehxa's original suggestion, but breaking it up across the tag, so `["w", "BTC", "bitcoin", "address"]`.
A client could assemble the values into the form the user needs for various purposes.

2025-02-26 13:24:50

- reply

I'm not mean to it, but it's not a person. 🤷‍♂️

2025-02-25 04:25:46

- reply

I still feel like I'm just figuring out the capabilities of the current models 🫨

2025-02-25 04:10:18

- reply

It's kinda funny to me how short the timescales are that Anthropic is working on.
For instance, on their site I read, "since June of 2024, Claude has been the leading LLM for coding." June of 2024 was basically yesterday.
Anthropic wants Claude to provide "breakthrough solutions" by 2027. I wonder if we won't become myopic, with everything happening on such short timescales.

2025-02-25 04:06:09

Wake up babe, new Claude just dropped.
https://www.anthropic.com/news/claude-3-7-sonnet

2025-02-24 22:25:19

- reply

Yep. We'll have our core relay set usable by default, and anonymous publishing, so that users don't need a preexisting keypair to read or write publications.
It's possible to read already without logging in, and it will quickly get better when we add search.

2025-02-24 22:21:11

- reply

Item 4 is already in our backlog—we plan to add anonymous publishing.
As to the others, we can create tickets for those.

2025-02-24 21:45:53

- reply

I've tried Nikka before, it's quite solid, as I recall.

2025-02-24 21:32:53

Anyone have recommendations for a good Japanese whiskey?
#asknostr

2025-02-22 18:04:02

- reply

I'd recommend Playwright for E2E tests in JS/TS projects. It's really powerful and easy to use.

2025-02-22 18:02:44

- reply

VS Code for its extension support. More recently Cursor, which is a fork of VS Code with first-class LLM integrations.
I think @nprofile…ddwx may have some good recommendations though.

2025-02-22 14:50:48

- reply

Robert Fagles and Richard Lattimore made the two most commonly-read translations, I believe. I read Lattimore in college, it has some memorable turns of phrase in how it renders the Greek.
Take a look at this site to compare some passages.
https://www.iliadtranslations.com/compare/fagles-vs-lattimore

2025-02-22 04:45:38

On behalf of the @npub1s3h…75wz team, I want to give a big shout-out to @npub1636…qpl4, who stepped in with his TailwindCSS expertise and spruced up the UI in #Alexandria for NIP-62 publications. You can see his work below.
I'm particularly grateful for his attention to detail on the light mode variants. I dare say Alexandria now has one of the most elegant light mode UIs in the Nostr space.
Take a look at these latest developments at next-alexandria.gitcitadel.eu, and let us know what you think!
https://image.nostr.build/6d1701569819d2c31853a2d5ca7d44850110ce90ead6e2305ba88338ab9d62ac.png
https://image.nostr.build/b3d44ba3c5db2698d43bfcf09c0fb97799494a1cb49ca765fdbc5db821ef183c.png
https://image.nostr.build/4607546d89d98d2cb54634535293ad46dea8e75239b081f4086a5d87efbda9d0.png
https://image.nostr.build/39f7b26322ed9e06a359ad74fbf7a7fa3e733e8cf42fe7b28e9d6cac51dddb92.png

2025-02-21 14:10:46

- reply

It's worth slowing down, refactoring, and retesting as soon as you notice an architectural flaw. The early stages of the project is the best time to do that, because the more code you have, the more time, energy, and risk is required to fix a bad design pattern.

2025-02-21 14:08:13

- reply

One of my favorite resources lately has been https://refactoring.guru. It's a great place to survey code organization and design patterns to understand what is available to help solve the problem at hand.

2025-02-20 22:10:17

- reply

Happy to add you, if you want to collab with other relay devs.

2025-02-20 21:58:53

- reply

We have a Matrix chat with @nprofile…u7te, @nprofile…nggv, and @nprofile…aqk5.
We've mostly been focusing on the needs of the #Alexandria project, but it's a cool spot for talking about relay concerns.

2025-02-20 17:02:39

- reply

In the US, as long as you're not in a coastal elite city, you could probably buy a house and raise kids in a lower-middle-class lifestyle on that. You'd have to be careful with the budget.

2025-02-20 16:59:26

- reply

Depending on where you live, $100k is barely enough to ante up to the table, especially if you're trying to support a family just on that income.

2025-02-19 21:05:07

- reply

☝️All of this.
Mozilla Development Network has great learning and reference resources for web development.
Typically hosting a basic website involves transferring a few code files onto the web server. Your server hosting provider should have some tutorials to help with that.
https://developer.mozilla.org/en-US/docs/Learn_web_development

2025-02-19 20:15:26

- reply

A lot depends on what you want to build?
Want to build Nostr familiarity? See about contributing some bugfixes or small features to existing projects.
Want to build a website with Nostr features? What sort of site do you want to build? What Nostr features will it have? Answer those questions and you can make yourself a roadmap.

2025-02-18 15:26:39

- reply

Yeah I agree it's outside scope. But if we wanted to add versioning, I think metadata would be the way to do it.

2025-02-18 15:25:58

- reply

In addition to traceback.

2025-02-18 15:05:26

- reply

Individual events in the tree, I think.

2025-02-18 15:05:00

- reply

Potentially. Just throw a bunch of tags on there.
Though that wouldn't work if we wanted to tie metadata to each level of the file tree.

2025-02-18 14:07:45

- reply

Versioning could be handled by the metadata event @npub1ecd…s735 suggested 🤔

2025-02-18 14:05:08

- reply

What kind of information would it be useful to have on this distributed filesystem concept? It doesn't need to exactly mirror Unix. Should we have a metadata event that contains relay hints, for instance?

2025-02-18 14:03:41

- reply

I think, at least for the first draft, it will be a read-only drive for everyone except the user who creates it. Handling write permissions would require some way of delegating event signing or deferring original events signed by one user to a replacement signed by another user, which means the event is constantly changing ownership/authorship.

2025-02-18 06:04:19

- reply

Boosting for the name 😂🧙‍♂️

2025-02-18 05:45:21

- reply

From the article:
"The `a` tags MUST be an unordered list of root 30040 index events."
A couple of notes:
1. The diagram shows a symlink connected to the drive event. How do we make symlinks discoverable for a client trying to browse a drive? It might be better for the drive to reference the symlink, and then the symlink references a "leaf" event in the filetree.
2. This specification says "unordered" list. Does this mean the client can render the child events in any order it desires? That differs from the behavior defined for 30040 events, so we should make sure the expectation is perfectly clear.
3. To use the Drive analogy, a typical drive would only have one root (e.g., `/` on Linux, or `C:\` on Windows). Is the 30043 considered to be the root of the filetree, or is it the flag that identifies the root? If that's the case, the root is the 30040, and each drive should have only one root.

2025-02-18 01:45:22

- reply

Dang.
Interestingly, I did try putting a `const newValue = value` inside the callback body in your example, and even that didn't work.
It seems like the Svelte compiler should undefine functions in component bodies, to be thorough, since those functions lose their captured context. Then everything in a component would be tied to its lifecycle, not just some things.
Maybe doing that with functions has other problems, though.

2025-02-17 22:47:24

- reply

Maybe the most idiomatic way to implement this pattern with modals in Svelte 5 is to use snapshots. That may suffice to capture the variables the closure needs into a persisted context.
https://svelte.dev/docs/kit/snapshots

2025-02-17 20:29:51

- reply

Taking a look at this...it's a really odd pattern, but it also doesn't surprise me that it's not working.
The `value` prop is: a) only scoped to `Component1`, and b) never passed to `Component2`. Since the `state` object on `+page.svelte` is redeclared, `value` doesn't live there, either. It looks to me like Svelte is doing the natural thing and letting `value` be destroyed with its owner (i.e., `Component1`).
Interestingly, the `closure` function seems to outlive its component. As your example solution suggests, copying `value` into the closure's scope ensures `value` has the same lifespan as `closure`.
How come this worked in Svelte 4? I'm trying to understand what Svelte/JS is doing with the references under the hood that would explain the behavior.

2025-02-17 19:37:41

- reply

Sick! So we can store blossom objects, and structure them however we like using NIP-62 (and extensions) events.

2025-02-17 19:06:33

- reply

Could we use 30040s to index blossom files? I don't see why not.

2025-02-15 03:08:44

- reply

If a struct is like the blueprints of a car, then the blueprints would say it needs to have an engine, and they might specify a few kinds of engine it might have. Maybe the car designer isn't an engine designer. He'll tell the engine designer "Go make an engine for this car," and trust the engine designer to do his job.
We'd say the lead designer's blueprint, in this case, is like the struct and its traits—they say that there has to _be_ and engine. The impl is like the engineer who actually designs the engine and makes it fit the vehicle.

2025-02-14 21:14:26

@nprofile…je4n what AsciiDoc parser are you using, or did you write your own?

2025-02-14 20:22:37

- reply

Perhaps, though I want to try making software better, first. That is my trade, after all.

2025-02-14 19:00:56

Everyday I weep at the slow, inexorable enshittification of software.

2025-02-12 14:38:51

- reply

Looks delicious! Mind if I ask for the recipe?

2025-02-11 23:27:13

- reply

I was looking at GUI options for C++ and Rust, earlier, and Slint looks promising to me. I haven't used it myself, though, so I can't say more on it.
https://slint.dev/

2025-02-11 18:47:00

- reply

Yeah and with all that privacy there's no paper trail when your ecash mint decides to walk off with your sats. 🤷‍♂️

2025-02-11 18:39:20

It seems that the problem with ecash is that it requires you to trust a 3rd party like you would a bank, but without any of the insurance or legal protections for your assets that exist with a real bank.
#nevent1q…fz3g

2025-02-11 15:15:40

- reply

You could kill two birds with one stone by writing a program for mathematical visualization.
When I was a CS student I made a Java GUI program that did fractal visualizations. You could do graph visualizations as well, perhaps illustrate different graph algorithms.

2025-02-10 00:40:42

- reply

That's exactly what we're imagining! Share or discuss individual sections.

2025-02-08 18:42:49

- reply

We call that Rosie the Robot

2025-02-08 15:14:45

- reply

I totally understand! We're an open source project, so if you ever want to contribute some code here or there, let us know and we can assign you a ticket!

2025-02-08 15:08:39

For anyone who wants serious buy-in to the future of project #Alexandria, this is your chance.
#nevent1q…xf00

2025-02-07 23:10:44

- reply

Too right for the lefties but too left for the right-wingers, I get the feeling.

2025-02-07 04:23:42

- reply

I would suggest doing it the other way around, actually. Ask the LLMs and the Internet as many questions as you want, but write the code by hand. That will cement the concepts better.

2025-02-07 02:24:26

- reply

I copied this over here for npub19ma2w9dmk3kat0nt0k5dwuqzvmg3va9ezwup0zkakhpwv0vcwvcsg8axkl to work on.

2025-02-06 15:56:30

- reply

This will be a great test case for improving our stylesheets!