fiatjaf

Then you can do stuff like
echo "#surely you're joking, mr npub1l2vyh47mk2p0qlsku7hg0vn29faehy9hy34ygaclpn66ukqp3afqutajft olas.app is broken again" | nak publish
It will add the hashtag, turn the npub1 code into a nostr:npub1 URI, turn the olas.app string into https://olas.app, add the "p" tag (and "q" tags too if you were mentioning an nevent1 code or naddr1 code) and finally publish it to your "write" relays and to any mentioned person (or author of mentioned events)'s "read" relays. There is also a --reply flag that you can pass an nevent, naddr or hex id to and it will do the right thing (including setting the correct kind to either 1 or 1111). And there is a --confirm flag that gives you a chance to confirm before actually publishing the result to relays.

2025-05-06 02:42:47 GMT

After a brief interruption, this bot is back and catching up:
#nevent1q…s6pd

2025-05-06 01:04:08 GMT

- reply

This was very mindblowing, for sure.
I pasted that command on my terminal and it played some music.

2025-05-05 13:08:18 GMT

Bitcoin Core had 4 pull requests merged in the last 5 days.
It's pretty clear that none of these had consensus from the broader Bitcoin community.
#nevent1q…kedw

2025-05-04 00:33:50 GMT

James O'Beirne has the best takes:
https://www.youtube.com/watch?v=JathbmXKwVg
Except for wanting bigger blocks: we actually need smaller blocks and Drivechain. He is also wrong about Nostr being awful. But aside from these two he is great on everything.

2025-05-03 15:28:06 GMT

Has anyone tried https://github.com/jj-vcs/jj?

2025-05-03 15:25:57 GMT

- reply

Jokes aside it was a good idea that someone should still implement, but instead of a DVM-via-events like this we should do it in a DVM-via-HTTP-API or DVM-via-special-relay.

2025-05-03 15:21:03 GMT

The first DVM:
#nevent1q…y0er

2025-05-03 02:38:09 GMT

Hopefully no one can read this note.

2025-05-02 19:08:05 GMT

DVM drama > OP_RETURN drama

2025-05-02 15:10:54 GMT

- reply

It wasn't clear to me at first, but now I think we should reframe DVMs as bots, instead of as a standard for RPC communication via Nostr events.
"Bots" are often seen as useless, playful things, but if we call them "DVMs" it's clearer they can also be useful tools. Ultimately it doesn't matter, it's all marketing.
Finally, we can also reframe HTTP/JSON APIs as yet another type of DVM. Zap providers, for example, are an HTTP/Event-hybrid DVM. @nprofile…7rj8 's translator API is another type of HTTP/JSON DVM that follows an existing standard (which is to copy some other translation provider I forgot) and is implemented in some clients (Damus?). Primal, Nostr.Band and Nostr.Wine also offer trending feeds HTTP/JSON-based DVMs, but they're all incompatible among each other (meanwhile Snort and Iris use the Nostr.Band API while Primal and Yakihonne use the Primal API, I don't know who uses the Nostr.Wine API), we should try to get them to standardize these (or offer these feeds as custom relays, which would be better).
That would solve our problems instead of creating more.

2025-05-02 10:57:25 GMT

- reply

I forgot to say that @nprofile…wvxz is right in the sense that HTTP services could also be considered DVMs.
It's hard for me to imagine that the feed-generator DVMs wouldn't be better as HTTP services rather than Nostr event publishers, or the Vertex DVM (which, by the way, is not listed in the official DVM specs, do not follow them strictly and is being used as an HTTP service even by their own website https://npub.world/ ).
In that sense every zap provider is a DVM: you call it via HTTP and it emits a kind:9735 event.
But today apparently people feel they're not being nostric enough if they do not turn what could have been an HTTP service into a service that listens on a bunch of relays for events and responds with events. I think a new NIP with clearer recommendations would address that, by which I mean we should tell people that sometimes it's better to do an HTTP service, and we can even standardize those HTTP services if they call for it (again, like NIP-57 did).

2025-05-02 10:42:53 GMT

- reply

@nprofile…wnez would you say @npub1z4d…j7rz qualifies as a DVM or not? Shouldn't that kind of thing also be considered a DVM?
What about a thing like https://bsky.app/profile/did:plc:bpkpvmwpd3nr2ry4btt55ack? I think the way Bluesky did this "labeler" stuff is interesting in a way: the DVM is just a "bot".
The bot should have a kind:0 profile and it can even post notes sometimes, perhaps advertising their services.
Some bots respond to kind:1 commands, others respond to kind:5001 commands, some send DMs, others publish content when prompted, others do it without being prompted.
If multiple people are running bots that behave similarly -- or if we expect that to happen -- it's time to standardize their behavior, otherwise there is no point.
There are tons of bots out there, I'd say some will never qualify as a DVM because they're just shitposting (which is good), but some are posting weather data, others are posting bitcoin data. It would be nice if bots publishing weather data could be standardized as DVMs, for example. I don't know bitcoin data, but what do I know?
I don't have a very concrete plan.

2025-05-02 10:32:10 GMT

- reply

Is there any thinking person who thinks Nostr isn't much better than the fediverse at this point? Or Bluesky? Or X?
They just happen to have a little more network effect.

2025-05-02 10:23:34 GMT

- reply

"Each DVM kind can have its own flow, NIP-90 flow is a suggestion, not a requirement."
This is basically why. There is no point in having a spec that is just a suggestion.
All I want is clarity, the current model is producing way too much confusion.
Deleting everything was meant as a first step only.

2025-05-01 20:02:00 GMT

- reply

You can talk all day, I have the screenshot, and screenshots never lie: https://aegis.utxo.one/a91ab475fb596ec6f781e25106cca2b20319d4a5f036365c5784a502698abe2b.png

2025-05-01 16:20:59 GMT

- reply

There are some "DVM" use cases that make sense, but framing it as this super generic tool that encompasses all "machine" use cases is absurd and counterproductive. Also trying to have a single spec for everything hurts all the use cases at the same time. The specs aren't even machine-readable (unless you're paying Sam Altman to read them for you).
If someone wanted to write a "DVM" that did one thing that person should have done that one thing, not called it a DVM, and we could have created a standard for it if it was useful, or not, maybe that machine job was a proprietary spec that didn't need standardization, maybe it would fit in some other NIP and enhance it, each case is different.

2025-05-01 16:17:02 GMT

- reply

The fact that they don't use the DVM on their own website says it all.
Instead they use a normal HTTP API at https://npub.world/api/query and server-rendered profile pages apparently.

2025-05-01 12:08:09 GMT

DVMs are not a useful concept.

2025-05-01 12:07:18 GMT

"I love Sam Antman and will give him all the money I can because ChatGPT is so good", said all bitcoiners and nostrers.
#nevent1q…6758

2025-04-30 19:12:59 GMT

- reply

@nprofile…axjj here you have it, a bigger article about Git on Nostr that does mention many projects:
#naddr1qv…hjvt

2025-04-30 10:49:49 GMT

It's time for clients to start displaying kind 1111 in their "replies" feeds.
What clients are doing it currently besides https://coracle.social/?

2025-04-29 03:07:38 GMT

This interoperability thing is actually very amazing: https://v.nostr.build/ic7VsRonqBUrNBIX.mp4

2025-04-28 18:11:23 GMT

This is such a good talk:
#nevent1q…85f5

2025-04-28 14:32:09 GMT

- reply

First you would need an interoperable format for slides that multiple apps could implement.
Or you could just associate content with a specific website and that website could use any format they want, then they have an internal database that associated slides with a Nostr public key, which is not a "real Nostr client" but it's also something that we need (we can't make everything an open protocol at once), in that case we don't need a kind or a NIP.

2025-04-28 01:04:22 GMT

Do we have something to learn from XMPP?
https://xmpp.org/software/?platform=linux

2025-04-27 13:29:26 GMT

Nostr casually mentioned as if it was obviously known to everybody here at 26': https://www.youtube.com/watch?v=rdE00-1xoqU

2025-04-27 01:41:28 GMT

Lantern v0.3 will come with an automatic detector of Nostr events even when they're being displayed in a web client (by looking at the URL) or when they're in a completely arbitrary webpage (by looking at @nprofile…7jp9's <link> tag):
https://v.nostr.build/4K5ldYVqLh8QIToa.mp4
That means that you can use Lantern to highlight and annotate and comment on an article on Yakihonne, njump or Habla or a website like https://fiatjaf.com/ or https://nostr.kosmos.org/@raucao and your comments will show up on every other instance of that same article.

2025-04-26 10:41:14 GMT

- reply

Why impractical? It doesn't have to be a huge thing, it can be very small, and every node runs it. It doesn't have to produce any intelligent output.

2025-04-26 01:49:41 GMT

- reply

I think you typed the question in the wrong prompt window.

2025-04-26 01:48:27 GMT

- reply

Meanwhile this is my experience with Goose recently:
https://i.nostr.build/CNYN6iV3kRSKUwEL.png
(It just closed, not even in error.)

2025-04-26 01:46:01 GMT

Coracle's custom feeds are actually usable now, and pretty fast.
Even though @nprofile…3w2r claims Coracle is buggy I've only been seeing one bug sometimes and it seems to be caused by a broken NIP-11 response from purplepag.es.

2025-04-26 01:39:27 GMT

What are some websites that use querystrings for core routing (i.e. the presence of a query string changes the actual page that is displayed, not only acts as a filter or toggle in the same base page)?
The only two I could find were YouTube and Hacker News (and the OPNSense dashboard, but that is not a real website).

2025-04-26 01:35:55 GMT

- reply

Stop bloating the Jumble feature set!

2025-04-26 01:26:02 GMT

I wonder if one could make a blockchain that is a neural net thing and every block modifies its parameters and that influences how it validates and interprets the next block and so on.
Or: what are people doing when they try to combine "blockchain + AI" in order to scam people? That was the best I could come up with.

2025-04-25 22:29:40 GMT

- reply

They didn't pay me enough.

2025-04-25 20:29:21 GMT

- reply

I can't see my own bookmarks!

2025-04-25 20:28:37 GMT

- reply

Amethyst is just not rendering the comment, other clients do.

2025-04-25 18:32:21 GMT

- reply

Deleting actually kinda works on Nostr (as long as you're not blasting your notes to all relays under the sun).

2025-04-25 18:12:21 GMT

- reply

Lantern doesn't create any kind:1s. Your comment is in the "comment" tag.

2025-04-25 13:52:04 GMT

- reply

That's how I feel about most things, but surprisingly not so much with Chromium.
Not desktop Chromium at least, on Android it's intolerable. I use Firefox there, a very old version of Firefox.

2025-04-25 13:50:15 GMT

- reply

We should also add it to https://oracolo.me/ @nprofile…admp (but will it work if it's rendered dynamically?)

2025-04-25 13:43:03 GMT

- reply

What is the thing about tmux? Were you saying nak didn't work on tmux?

2025-04-25 13:27:44 GMT

- reply

Feeder supports Nostr articles too: https://github.com/spacecowboy/Feeder
And https://noflux.nostr.technology/
Also both https://github.com/msasikanth/twine and https://github.com/Ashinch/ReadYou have merged PRs that add Nostr articles support, but for some reason they haven't released those yet.

2025-04-25 12:34:00 GMT

Please try Lantern https://chromewebstore.google.com/detail/lantern/jjoijlenmgefkaeiomoaelcljfibpcgh again if you haven't yet and let me know of all the horrible bugs.

2025-04-25 03:44:40 GMT

- reply

I see. Oh, well, the problem is that it doesn't load anything for me because it's only connecting to my NIP-65 relays, which are not shared by many, so I basically only see you and @nprofile…qsav in my feed. Maybe that's it.

2025-04-25 02:32:17 GMT

- reply

Do you use Notedeck consistently? I was trying to use it here now that my Gossip is broken, but it doesn't feel super usable yet. Doesn't load quotes and has all the "pre-alpha" warnings it had 6 months ago, it kinda showed me notes from the last time I opened it (2 months ago) and then some more recent notes but not very recent. It's hard to know if it is working. Maybe I just have to get used to it?

2025-04-25 02:22:41 GMT

- reply

Just like the Brave browser protects you from ads and then forces you to see their own ads.

2025-04-25 02:22:20 GMT

Remember when Elon Musk saved free speech?
#nevent1q…xapg

2025-04-25 02:09:32 GMT

- reply

What clever relay name? And how does a relay name affect Google?

2025-04-24 22:06:20 GMT

- reply

This is a good counterpoint to https://oracolo.me/

2025-04-24 18:58:19 GMT

- reply

I'll switch my allegiance to wikitext if you can come up with a wikitext parser in JavaScript, Go and Rust, and if you get me a relay that has all the Wikipedia articles published as events.

2025-04-24 18:49:32 GMT

Someone please build this.
#nevent1q…xfss

2025-04-24 17:12:55 GMT

- reply

OK, let's do that, but only after Nostr-over-STDIO and Nostr-over-UnixSockets are well-established.

2025-04-24 12:20:34 GMT

MCP is just a JSON-RPC API with natural-language docs.
Maybe all the hype around it will make developers realize that RPC is better than REST.

2025-04-24 12:18:58 GMT

@nprofile…l8mn was using 30GB of storage (is that expected @nprofile…m90e ?), so I nuked it all, but then it has also deleted my key so now I'm using Coracle as my main client for a while.

2025-04-24 10:36:46 GMT

- reply

Don't be silly. We all know RSS is a perfect protocol that came from the heavens directly and therefore it cannot be beaten or replaced. We can only replace X, TikTok, Instagram, Facebook, YouTube, Wikipedia and GitHub, we cannot replace RSS, RSS is so great it will beat us every time. Also it doesn't have any flaws, it's not tied to the DNS system, it's not just a single giant page of XML that no one understands and therefore everybody has to use custodial providers and get tied to them. Episodes can be linked to directly and unambiguously, we don't even have to link directly to Spotify or YouTube or Fountain or refer to an episode by vaguely alluding to it. Moreover, RSS is so great that it is a single protocol with all the network effect, no one has ever created a competing format called Atom nor another called JSON feeds and the ones that tried saw zero adoption because RSS is one 100% of the pages.

2025-04-24 10:24:19 GMT

Somewhat of a long shot, but if we could achieve it would be great for adoption:
#nevent1q…u0ke

2025-04-23 13:45:48 GMT

- reply

There was a horrible bug, thank you.
It was so horrible I can't believe zapbox has been working at all.

2025-04-23 01:36:48 GMT

In the middle of this interview the host, beginning to grasp the idea of Nostr, says something like: "I like this separation of powers".
It is interesting that it's not hard for people to realize that that a big corporation controlling the "public square" isn't a good thing, but many will have an immediate impulse of trying to fix that situation by coming up with a protocol in which not one company controls everything, but multiple companies.
This seems to be the mindset behind Mastodon, Matrix, Farcaster and Bluesky. They all assume users will be immediately subject to one server, one company, but that there is room for other companies to join and compete for users, or something like that (there are differences in how this plays out between all these protocols). So instead of one "corporate square" controlled by one company you end up (in the best case) with a bunch of private squares that may or may not have communication between them.
Nostr is different because it tries to create a single square that is actually public and infinite in size where any company can open a stand, but also anyone can go and speak to whoever wants to hear without asking for permission from any of these companies, simultaneously any person or organization can find a corner (there are infinite corners) and talk only to a smaller group and so on.
#nevent1q…2myl

2025-04-22 15:42:07 GMT

- reply

The last 10 minutes are very intense.

2025-04-21 17:55:08 GMT

- reply

You can argue that Bluesky censorship is happening at the client level today, but I'm not sure that's what the Turkish government thinks. Once they understand it they might ask for a server-side censorship, and then Bluesky won't have any chances.
They might realize that once users start migrating to alternative clients. If that happens that will be a good sign that censoring clients directly isn't a good idea, so the weakest point of Nostr will actually prove to not be that weak after all.
Still I think we need more client diversity (even though we actually have plenty) and I think we will get that over time.

2025-04-21 17:52:25 GMT

- reply

The Nostr strategy from the start was to have many clients -- none of which are the "official" client -- such that the threat of one client hardcoding banlists is mitigated. Of course that could always happen and is probably the hardest part, but assuming a relatively diverse ecosystem of apps it gets harder for one app to comply with such ban requests as there will always be alternatives for users to jump to (and with different clients having different approaches to Nostr it makes sense for users to experiment with more than one as time goes by even if none censor). Also no country has ever tried to force web browsers into compliance, and Nostr client are just like web browsers, so we have that on our side.
Censoring relays, either by DNS or by asking the relays directly, absolutely doesn't work because it's in the direct interest of the person being censored to move to other relays and all clients, as long as they're doing Nostr right (i.e. implementing the "outbox model" for basic feeds), will find that person. See https://how-nostr-works.pages.dev/#/outbox for an example.
Media server censorship is solved by Blossom the same way the outbox model solves relay censorship. You signal that your media is in another server, clients can adapt and use that.

2025-04-21 17:47:50 GMT

- reply

People can also use an alternative X client that doesn't show the bluechecks. What is the difference?

2025-04-21 17:16:08 GMT

This is getting ridiculous: https://bsky.social/about/blog/04-21-2025-verification

2025-04-21 05:04:54 GMT

- reply

Run it locally while developing with nak serve and nak bunker.

2025-04-21 01:55:11 GMT

- reply

What is the problem with NIP-46?

2025-04-20 22:35:24 GMT

- reply

Thank you, I've changed it so ⚠️ reactions are ignored now.

2025-04-20 22:15:55 GMT

- reply

I don't know what variable you're referring to. What is the name of the variable?

2025-04-20 16:34:57 GMT

- reply

The client is being used by `nak` itself. Instead of taking a secret key nak can take a bunker URL and connect to it as a client.
It's also used in https://git.fiatjaf.com/song/+/master/cmd/song-pre-push/main.go for automatically signing updates before pushing to the git server.
And in some other yet places yet to be revealed.

2025-04-20 16:33:01 GMT

- reply

Wrong. It was used in a ZBD Social bunker implementation that was never released. And is used today in https://git.fiatjaf.com/promenade/+/master/coordinator/nip46.go that powers the https://nstart.me/ bunker.
I almost never write library code that I'm not using in anything more concrete.
But I agree you probably won't need it.

2025-04-20 14:02:52 GMT

Who understands this? https://eprint.iacr.org/2025/709.pdf

2025-04-19 16:58:32 GMT

- reply

What were the problems with https://github.com/jamesmagoo/nostr-writer?

2025-04-19 16:56:32 GMT

Can a normal web app make a call to a NIP-55 signer on Android?

2025-04-19 16:06:37 GMT

- reply

You mean vibecoding, right?

2025-04-19 00:26:44 GMT

The world needs this desperately:
#nevent1q…x4g9

2025-04-18 22:31:35 GMT

So both https://www.themoneyillusion.com/ and https://slatestarcodex.com/ have abandoned their blogs which worked perfectly over many years and migrated to Substack. I imagine many other blogs I never knew about have done the same.
Why is that and what does it mean for us?

2025-04-18 22:27:20 GMT

- reply

No one has to run these things. I don't run them. They have nothing to do with the current discussion.
Your client can always suggest or even pick a default relay for you. Suggestions can be based on your friends' choices or something else, or just be hardcoded (which is nothing but a suggestion from the developer whom you're already trusting to some extent anyway).

2025-04-18 22:06:24 GMT

Hopefully this settles all the discussions regarding "dumb relays" that I've seen since I wrote that line on that initial Nostr description text:
#nevent1q…9md8

2025-04-18 20:01:30 GMT

- reply

I think there is a misunderstanding on the meaning of "dumb".
Is choosing to accept notes from some person and not from others dumb or not? What about allowing some person to read some notes and not others?

2025-04-18 17:27:43 GMT

@nprofile…05wq is it possible to make it so npub.world somehow breaks the link to CSAM-publishing accounts so they don't get into the index?

2025-04-18 17:02:16 GMT

- reply

Nostr is based on the fact that relays are special. They have to be special because they have owners and they can't be free to write by anyone. Nothing would work if it wasn't for that.
Once you've crossed that line, the specialness in relays can only grow.

2025-04-18 17:00:18 GMT

- reply

Because you want to be able to connect to just that single (or more, of course) relay in order to get all your DMs.
Because you want to trust it to not leak the DMs you receive.
Even if you don't care about any of that does it hurt to publish a new event with your same "read" relays in a list? It serves as a good signal that you're ready to receive NIP-17 DMs instead of NIP-04.

2025-04-18 10:59:10 GMT

- reply

NIP-17 has very good reliability and they only work if the target has specified special relays for receiving, which helps with that. NIP-04 DMs should be killed indeed.

2025-04-18 10:56:47 GMT

- reply

@nprofile…athl what do you think of publishing that on Nostr so we can comment on it better?
I agree with everything you said, it's very similar to what many of us have been arguing for for years. You'll find the discussions if you search for "outbox model".
I just don't agree with the DM bit. NIP-17 should be followed as it is better solution. But this is a minor point and almost unrelated to the core of your article.

2025-04-18 10:53:48 GMT

I don't know how he does it, but @nprofile…t2mx keeps publishing the best curated links to interesting things on Nostr day after day.

2025-04-17 22:10:08 GMT

- reply

In that case we'll have to kill all of Nostr.

2025-04-17 15:20:27 GMT

Suppose there is a relay that only accepts PoW notes, another that only accepts notes from your WoT, another that only accepts notes once they got enough zaps or likes, another that is manually curated by a group of people, another that only takes notes from people that pay to publish, another that only takes notes classified as talking about some given subject.
Then we could have a client mode that fetches from all these, then only displays notes that have showed up in at least 2, or 3, of these -- or whatever combination.

2025-04-17 13:41:58 GMT

I have zero tabs open for the first time in many years.

2025-04-17 12:11:47 GMT

It works: https://aegis.utxo.one/35e607c6b5303d7a8abb87f265d077431abbf4c732baac4f35ffdffa1c1ebf59.png
I can see mine and @npub1qe3…zefe's annotations there, as well as replies.
#nevent1q…s0pr

2025-04-17 04:16:50 GMT

- reply

If you open this with https://chromewebstore.google.com/detail/lantern/jjoijlenmgefkaeiomoaelcljfibpcgh you might be able to see my comments on the side, please report if you can't (you have to switch the group to "From Contacts").
OK, I just found that it is not displaying the hashtags.

2025-04-17 03:49:50 GMT

- reply

Random dude uses other niche relays. People who follow random dude automatically connects to niche relays and fetch random dude's posts. No one has to install anything.
https://how-nostr-works.pages.dev/#/outbox

2025-04-17 03:48:42 GMT

This is a good one about all the ways in which Mastodon doesn't work: https://v.cx/2025/04/mastodon-exit-interview
(and also about the fakery that is Bluesky's decentralization -- even though the author likes Bluesky as a platform)

2025-04-16 19:15:07 GMT

- reply

They definitely would, that's why the protocol is organized to be resilient to that kind of thing.

2025-04-16 19:10:27 GMT

This thing looks like a massive scam.
Why hundreds of people employed to keep a database of bugs?
And then they make a giant drama once their massive government funding gets cut.
#nevent1q…jsqc

2025-04-16 10:46:41 GMT

- reply

Doesn't matter, I'm not in Turkey so I can still see it.
Anyway it's a matter of time before "deer.social" receives a letter from the Turkish government too. How many "AppViews" will exist after? And will users keep moving from one to the other, reinstalling their apps, changing their habits?

2025-04-16 10:07:05 GMT

So Bluesky has just banned some guy exclusively on Turkey only because the Turkish government asked.
https://cdn.azzamo.net/ae569fbc63f1951c4a8d2ea43c3cbe24ede9ed06b3adb483ef69e00ba9d76ad2.jpeg
(from https://x.com/aliskorkut/status/1912191854939943362, this is the profile: https://bsky.app/profile/carekavga.bsky.social)