<unnamed>

Public Key

npub1jlrs53pkdfjnts29kveljul2sm0actt6n8dxrrzqcersttvcuv3qdjynqn

Profile Code

nprofile1qqsf03c2gsmx5ef4c9zmxvlew04gdh7u94afnknp33qvv3c94kvwxgspp4mhxue69uhkummn9ekx7mqprpmhxue69uhkscnj9e3k7unpvdkx2tnnda3kjctvqy28wumn8ghj7un9d3shjtnyv9kh2uewd9hsz8rhwden5te0dphkgmrzdajzucm0wfskxmr99eek7cmfv9kqq4amh5

Publishing to

nos.lol

hbr.coracle.social

relay.damus.io

hodlbod.coracle.social

Last Notes

2026-03-23 20:37:13 CET

- reply

Cool, I'm glad to know that's the sort of thing you're hoping to address. Where is this stuff?

2026-03-12 20:15:02 CET

I love Norse mythology, we just read about the goat that stands on top of Valhalla, which provides enough mead from its udders for roughly 10 billion heroes to drink

2026-03-12 19:43:03 CET

TFW people take your "send tweet" seriously
#nevent1q…xv9c

2026-03-12 19:00:20 CET

- reply

Encapsulation assumes developers are dumb and will shoot themselves in the foot. They are, and they will, but I prefer my code to have affordances for me to do weird things. But a big part of that is the difference between being a one-man team with comprehensive knowledge of the codebase and being on a larger team with shared institutional knowledge.
My complaint is less about const, which is a good idea, than about linters wasting my time by making me switch back and forth. The number of times const has saved me doesn't compensate for the busy work of pleasing the linter.

2026-03-12 18:26:55 CET

- reply

Sure, but the language's conventions are very reliant on reassignment. And const doesn't prevent mutation anyway. I say this as someone whose favorite language is clojure

2026-03-12 18:21:16 CET

- reply

Just do `let SECONDS_IN_DAY` and call it good

2026-03-12 17:05:33 CET

- reply

I've already stubbed out a re-write of welshman in rust. I don't know how it'll turn out, but I'm definitely interested in the idea

2026-03-12 16:42:22 CET

- reply

Is `awk` not perfect?

2026-03-12 16:18:53 CET

- reply

No, another think I wanted to do but never did. If you copy the naddr of the feed, you can go to coracle.social/naddr and it should load the feed

2026-03-12 16:17:06 CET

- reply

I (and the LLMs) spend way more time switching between const and let because the linter tells us to than relying on something not being reassigned because it's declared with const. If you need const, your scopes are too big.

2026-03-12 16:16:18 CET

- reply

😂

2026-03-12 16:16:05 CET

- reply

Kind-based approvals is a broken UX pattern. I don't know what better would look like, but no way is a user going to know offhand what kind 68291 is. I also don't think it's realistic for signers to render every kind correctly either. Honestly I feel like developers of signers have sort of dropped the ball on exploring better ways to reconcile UX smoothness with user safety/privacy/control.

2026-03-12 16:13:31 CET

- reply

The necessity to go back and forth between `let` and `const` depending on whether a variable is reassigned isn't worth the mental overhead conserved by knowing that a variable isn't reassigned provided by `const`

2026-03-12 04:31:11 CET

- reply

I think I could live with that

2026-03-12 00:43:42 CET

@nprofile…cffw small requests for wisp:
- Likes should be one click, lots of times I see the emoji menu open and I get decsion fatigue and end up not reacting. I want to like and move on, maybe long press for the menu
- When I restart the app my notification tab selection is gone (I always want it to default to replies, not all)

2026-03-12 00:40:36 CET

- reply

Do you have any ideas for how to get better search coverage across more of tge network? Even if all relays supported search, you would have to make relay selections based on some other criteria than search term, which means you're limiting reach out of the gate. The only solution I can think of is indexers, which creates google all over again, and also doesn't cover auth protected relays either.

2026-03-12 00:32:16 CET

- reply

Typescript still has the let/const dichotomy. I held on for a long time, but adopted ts in 2023

2026-03-12 00:31:53 CET

- reply

Of course, your execution is better than mine probably

2026-03-12 00:31:44 CET

- reply

It's fun watching your "client developer" arc. A ton of the things you've said are opinions I held years ago, but things weren't as clear back then, so I either didn't follow through on them or went back and forth. Like, I used to use client tags, then there was talk about it being a privacy problem so I removed them.
Take this either as me boasting or as flattery as you like, but wisp is basically coracle v2. It's what I would try to write today if I started over with a social client.

2026-03-12 00:28:59 CET

- reply

what are the urls?

2026-03-12 00:28:53 CET

- reply

and use rust? With the machine's help I am finally giving that a go

2026-03-11 23:48:51 CET

`const` in javascript was a mistake (like most things in javascript)

2026-03-11 23:42:43 CET

What are some good default DM inbox relays? So far I have auth.nostr1.com but would like to have a couple more

2026-03-11 23:04:10 CET

- reply

stupid outbox model is better than smart outbox model, I appreciate this now

2026-03-11 18:39:40 CET

- reply

@nprofile…jrqt I appreciate your experiment but I think auto auth is not the enemy. Fingerprinting can happen in many other ways. It's worse on nostr, but it's also true of the entire rest of the internet.

2026-03-11 18:38:10 CET

- reply

I am not sure about this, because it undermines privacy in other ways. Now, if I want to protect my privacy by enforcing auth on my outboxes (to restrict reads to people I follow or something), I am banning myself from wisp. Better would be to have some kind of heuristic for relays that might legitimately enforce auth. If someone I follow has a relay that enfoces auth, I don't mind identifying myself. But if I'm browsing topics or whatever that changes.

2026-03-11 17:27:28 CET

- reply

https://habla.news/p/npub1jlrs53pkdfjnts29kveljul2sm0actt6n8dxrrzqcersttvcuv3qdjynqn
https://building-nostr.coracle.social
https://fiatjaf.com/-/tags/nostr
(in my very humble opinion)

2026-03-11 14:31:23 CET

- reply

Thank you for the vote of confidence

2026-03-11 01:15:48 CET

You heard it here first folks, Lopp is trans
#nevent1q…89dp

2026-03-10 23:58:22 CET

This is a balanced, intelligent, and constructive article on using LLMs for coding:
https://ankursethi.com/blog/programming-language-claude-code/
My main takeaway is that some problems are well suited to using LLMs, others aren't. I followed basically the flow described when porting pomade to go and rust, but LLMs invariably make a mess and slow me down when doing front end work. Stuff that can be explained in words and tested using a command line interface works much better with LLMs.

2026-03-10 22:43:37 CET

- reply

This should be fixed in the next version

2026-03-10 21:40:51 CET

- reply

flotilla will/can never do that to you

2026-03-10 21:05:01 CET

- reply

same, I think I'm finally old enough to understand it

2026-03-10 19:10:59 CET

In case you missed it:
#nevent1q…s3vd

2026-03-10 14:14:50 CET

- reply

Do you like seed oils, plastic, planned obsolescence, cars with screens instead of buttons, bloatware, pressboard, glyphosate, red dye, and McMansions then? Because those are are failures of craftsmanship that affect us every day.

2026-03-10 05:18:51 CET

Flotilla 1.6.5 is out on web and on zapstore!
This release is NOT the big one. That's coming hopefully this week. However, this release is substantial:
* Attempt to fix permission grant for notifications
* Make sync logic more robust
* Add unban/unallow support
* Improve support for downloading/opening protected images
* Add manual send/receive to wallet
* Show wallet status when wallet is unreachable
* Update nostr signer capacitor plugin
* Fix some safe area insets
* Update NIP 55 signer plugin (fixes Primal login)
* Refine space join dialogs and discover page
* Reopen the last DM that was open when navigating back to chat
* Get rid of ChatEnable interstitial
* Enable auth for relays we're publishing to
* Drag and drop space icons
* Add better muting support
* Add back button to settings menu
* Add page titles
* Improve scroll to event behavior
* Add in-memory search to rooms
* Fix editing messages with html tags
* Fix DM media detection
* Clean up reporting dialogs
* Improve room detail
If you experienced bugs, please try the new release and tell me about any new bugs. Voice rooms and email/password login coming very soon.

2026-03-10 02:20:21 CET

- reply

Explain how facebook business suite is as bad as pressboard bookshelves

2026-03-10 00:59:26 CET

- reply

Pretty scary for large scale alignment

2026-03-10 00:36:50 CET

- reply

I am a below averge vibe coder

2026-03-10 00:21:14 CET

- reply

In theory sure, but there is an inherent tension in higher level tools, because it distances you from your work. The key to craftsmanship is balancing depth of understanding through unmediated participation in your work with the precision conferred by better tools. (notice I didn't say efficiency, which is orthogonal to craftsmanship)

2026-03-10 00:15:13 CET

- reply

No, it's badly designed. I want to redo it but haven't had a chance. Create a feed which pulls from that list.

2026-03-09 22:35:34 CET

- reply

> You're experiencing the same as countless other artisans have experienced in the last century.
> It doesn't matter to anyone but you.
These claims are not mutually consistent. The result of the loss of craftsmanship is disposable consumer goods and the exploitation of the environment and human life. Software already was this, but LLMs are more so. The loss of craftsmanship is the loss of everything worthwhile that was being produced, and its displacement by literal garbage.

2026-03-09 15:09:20 CET

- reply

Hit the customize button at the top of the feeds page

2026-03-09 00:22:27 CET

- reply

Just been inching into woodworking over the past year. Asked my mom for a spoon carving knife so she gets this first one

2026-03-08 21:29:01 CET

- reply

If you mean did I do it with my eyes closed, yes

2026-03-08 21:28:31 CET

- reply

Just find some sticks, this one was pretty old which makes it easier to carve. Bass wood is easy to carve if you're serious about it, but kids mostly just want to make sharp things 😂

2026-03-08 16:19:18 CET

- reply

Awesome. My sone came and sat next to me and made this:
https://hbb.coracle.social/5564c00d280fe0209c11776814886c23b39e4bce44c0b7ff603053b4ab17f708.jpg

2026-03-08 16:17:49 CET

- reply

Norway Maple, just a stick from my backyard

2026-03-08 15:18:22 CET

- reply

Just a stick

2026-03-08 14:41:39 CET

I made a spoon
https://hbb.coracle.social/7b63558d4e2a008cd1029f9806a3d2b878c0551112a575d2b4a798ea6ccf3a1b.jpg

2026-03-07 21:23:55 CET

Not saying your mom jokes in front of my kids is one way in which I die to self daily

2026-03-07 16:26:21 CET

- reply

Fresh
https://hbb.coracle.social/0e560a4959be1cead77ee8acc95df39ca8e0a09fc1bcab3bd74b38d746e3677b.png

2026-03-07 16:21:26 CET

- reply

A bit of both

2026-03-07 01:25:39 CET

@nprofile…jejf I gotta ask, how is wisp so fast

2026-03-07 01:13:29 CET

switching to @nprofile…6nzx makes spending on tokens hit different, I've spent 300k sats this week 😅

2026-03-06 23:34:52 CET

Getting close to being able to release pomade to Flotilla users. Here's a boring video which showcases the UX:
https://coracle-media.us-southeast-1.linodeobjects.com/pomade_demo_20260306.webm

2026-03-06 23:07:40 CET

- reply

Nice! I will probably use this, it could also work great for moving stuff to untrusted devices using one-off sign in with nip 46

2026-03-06 20:31:32 CET

- reply

They are leftists, they will probably eagerly comply

2026-03-06 17:18:32 CET

- reply

Hints all look ok, here's the note:
#nevent1q…cvk8

2026-03-06 17:12:09 CET

- reply

It never did load, but it looks like it was only sent to mleku's inbox relays for some reason. Could be a coracle bug

2026-03-06 16:35:19 CET

- reply

The note that failed to load is the same note, in coracle it's only shown once with both replies underneath

2026-03-06 16:33:34 CET

- reply

Yeah, I switched to replies (which imo should be the default) but still the majority of space on the page is taken up by my own notes. When looking at notifications, I don't want to see what I said, I want to see what everyone else said. Coracle shows multiple replies under the same note if they came in at around the same time. Here's an example:
https://hbb.coracle.social/98dc896ca5171e608df0b27b0535ca2f0e76a45b6ebd159f604cefe2cbd7d748.png
https://hbb.coracle.social/2d1de937c7896959659623a530e2536fceeae80018d37be393d8ef3fff6869a5.png

2026-03-06 15:50:54 CET

- reply

The notifications screen is kind of noisy and repetitive. I really like how easy it is to keep up with replies on coracle and only look at zaps and reactions if you're feeling vain.
Complaining because wisp is the first client that has a chance of displacing my own client as a daily driver. Although that won't happen realistically without a "no media" mode

2026-03-06 15:38:33 CET

- reply

Is there a way to turn off the "1 new notes" button at the top of the feed? It's incredibly distracting to OCD people like me. Also, you should fix the grammar

2026-03-06 01:05:55 CET

- reply

Oh, I see, yes I ignore <15

2026-03-06 01:04:47 CET

My spirit animal is this website
https://406.fail/

2026-03-05 23:14:17 CET

- reply

I don't know how true this is, but maybe because they see it as competing for funding with bitcoin, as a distraction, or even as "crypto"? I know there was some talk about this back when OpenSats started their nostr fund.

2026-03-05 21:28:53 CET

> It’s been six years since GPT-3, and there has been no “move 37” moment for writing (as there was for AlphaGo’s creative play of Go). Not even close.
https://www.theintrinsicperspective.com/p/bits-in-bits-out

2026-03-05 19:36:45 CET

This article says a lot of what I wanted to say about LLMs, but couldn't find the words: https://acko.net/blog/the-l-in-llm-stands-for-lying/
I don't agree with his conclusion that leaning in to intellectual property rights and source citation is the solution, though that's an interesting though. But there are some great sections, particularly in the first half. Here are some highlights:
> LLMs do something very specific: they allow individuals to make forgeries of their own potential output, or that of someone else, faster than they could make it themselves.
> Experienced veterans who turn to AI are said to supposedly fare better, producing 10x or even 100x the lines of code from before. When I hear this, I wonder what sort of senior software engineer still doesn't understand that every line of code they run and depend on is a liability.
>
> One of the most remarkable things I've heard someone say was that AI coding is a great application of the technology because everything an agent needs to know is explained in the codebase. This is catastrophically wrong and absurd, because if it were true, there would be no actual coding work to do.
>
> It's also a huge tell. The salient difference here is whether an engineer has mostly spent their career solving problems created by other software, or solving problems people already had before there was any software at all. Only the latter will teach you to think about the constraints a problem actually has, and the needs of the users who solve it, which are always far messier than a novice would think.

2026-03-05 17:25:19 CET

- reply

Sounds great! Another feature request is to add api request history somewhere. On opencode I spent $20 very quickly and it was really helpful for figuring out exactly which API calls caused it. Including number of tokens in/out, dollar amount of upstream, and payment amount would be awesome. Also, I've noticed that the current tally of dollar amount doesn't show in opencode like it does for other providers.

2026-03-05 17:21:41 CET

- reply

Nice, sounds like a great way to simplify. I like routstr a lot, but here are a few other things you could to do improve the UX:
- Don't process payments synchronously in chat
- Auto-refresh balances in the background?
- Maybe eliminate the balance on a key? Although having a budget on a key is nice.

2026-03-05 03:16:34 CET

- reply

I may be wrong, I tried again with another pubkey and it didn't merge accounts

2026-03-04 23:10:41 CET

- reply

delusional

2026-03-04 23:08:44 CET

Just added this to my opencode build prompt:
> You are in a docker sandbox, which means timestamps on files are often incorrect. To get around this, always touch a file before editing it.
🙄

2026-03-04 21:53:44 CET

- reply

Yeah, pretty nice. Although it auto linked my email account, could there be an attack vector based on non-unique names?

2026-03-04 17:54:06 CET

@nprofile…uy5c https://git.mleku.dev/mleku/gitea-nostr-auth is giving me an SSL error

2026-03-04 01:52:19 CET

- reply

How is sprites different from just setting up a VPS? How's the latency?

2026-03-04 01:46:34 CET

- reply

You clearly never wrote php in 2012

2026-03-04 00:47:58 CET

Vibe coding is the death of abstraction. Why use the visitor pattern or transducers when the LLM will just scatter `if` statements everywhere anyway

2026-03-03 20:44:57 CET

- reply

That 20% matters. I try to use kimi k2.5 as much as possible, but it's just not as good as sonnet

2026-03-03 20:44:11 CET

- reply

stuff != wealth

2026-03-03 19:42:04 CET

Agents are getting more expensive:
https://newsletter.danielpaleka.com/p/you-are-going-to-get-priced-out-of

2026-03-03 19:21:51 CET

- reply

No one should use openclaw, use nanoclaw or ironclaw or something else if you must run an agent

2026-03-03 18:42:37 CET

```
docker sandbox save opencode-flotilla my-opencode:v1.0.1
Snapshotting image in sandbox ...
Reading image from sandbox ...
Save complete. To use the image:
docker sandbox create --load-local-template -t my-opencode:v1.0.1 [...]
```
```
docker sandbox create --load-local-template -t my-opencode:v1.0.1
unknown flag: --load-local-template
Usage: docker sandbox create [OPTIONS] AGENT WORKSPACE
Run 'docker sandbox create --help' for more information
```
great, thank you docker
#nevent1q…jl3k

2026-03-03 18:41:02 CET

docker sandbox is clearly vibecoded. Half the flags don't work, and half the commands advertise flags that don't exist.

2026-03-03 03:51:14 CET

Spent the day fiddling with agent isolation. At first I went down the rabbit hole of setting up a dev environment on an old macbook and acc,ssing it over wireguard, but the latency was annoying. Then I tried matchlock, which was promising but had weird build and control character issues. Finally, I went with docker sandbox, shich is good enough, although I had yo use a very dumb hack to get my config into the container. This is a massive product opportunity.

2026-02-28 02:07:56 CET

- reply

Just coding

2026-02-27 23:10:56 CET

- reply

That's pretty much how I'm setting it up. I'm not going to go full openclaw, but I am going to give the machine its own ssh key, nsec, etc

2026-02-27 23:07:53 CET

- reply

Setting up nix on an old 2012 macbook I have lying around. We'll see if it survives 😂

2026-02-27 22:30:55 CET

- reply

Yeah, I was thinking about that. I intend to give it all its own keys.

2026-02-27 22:08:21 CET

Now that I'm using agents more extensively, I'm thinking about moving my development environment to a VPS to make sure the agents don't send any important data (like my ssh keys) to my provider. Am I being paranoid, or has anyone else done this?

2026-02-27 19:59:04 CET

- reply

Very good articulation I hadn't hear before. We're already seeing a lot of those fragility points with the increasing ability to impersonate/scam people in ways that we have zero defenses against. Many more will come I'm sure.

2026-02-27 18:03:21 CET

- reply

Is AI anti-human, pro-human, or something else?

2026-02-26 20:05:25 CET

- reply

another standard with partial support, neat

2026-02-26 20:04:47 CET

- reply

Easy UX != simple UI, it means things "just work". Do all 100 wallets interoperate seamlessly with the other 100 (10k combinations)? Certainly not. That's my point — not that routstr didn't work (it did), but that it's crufty UI reveals the underlying complexity that leads to poor UX.
We can agree to disagree, I don't expect to convince you. But I do think you need to take this problem seriously in order for cashu to succeed (and I think you do, you just seem to treat people bringing this up dismissively, which is weird).

2026-02-26 19:19:43 CET

- reply

These are all obviously much better, but unless all wallets seamlessly support onchain, the various flavors of lightning, and cashu, you're always going to have friction when sending/receiving.
"scan what?"
"receive what?"
"what kind of qr code is this?"
"oh, my phone doesn't support NFC"
"my wallet doesn't do onchain, can you send me a lightning invoice?"
"it's saying something about liquidity"
"what's a 'mint'?"
"what is a 'melt'?"
"what is a 'submarine swap'?"
"why do I need to 'topup', can't I just transfer funds from my bank?"
None of these problems are impossible to solve individually, but solving all of them across a network of apps created by people with different ideas is combinatorially difficult.
In nostr we say "don't overload kinds". I think this is the problem — we've overloaded the term "pay with bitcoin". It means:
- P2PKH
- P2SH
- Native SegWit
- Taproot
- Bolt11 invoice
- LNURL
- Lightning address
- Bolt12 invoice
- Keysend
- Cashu eCash
- Fedimint eCash
- Spark
- Ark
I'm not saying each of these doesn't have its use case, or that innovation isn't good, but this complexity comes at a cost, which either the wallet developer or the user has to bear.

2026-02-26 19:06:08 CET

- reply

I have four wallets:
- Onchain onramp (strike)
- Lightning (alby)
- Cashu balance in routstr
- Token balance for an api key
It's like 2 or 3 too many

2026-02-26 17:52:01 CET

A few months ago I asked @nprofile…xvl9 to sell me on using cashu. His argument was focused on better payment UX, which is compelling.
But something I think is left out of the conversation around lightning/ecash is the funding UX, which can be very complex.
I recently started using routsr (and appreciate it), but it exemplifies this UX friction perfectly. I've written lightning and cashu integrations, and even I am confused by what the UI is telling me. This is not (entirely) the fault of the routstr developers, it's a result of the irreducible complexity of separate protocols for different technical tradeoffs.
https://coracle-media.us-southeast-1.linodeobjects.com/routstr_cashu.webm

2026-02-26 16:25:49 CET

AGENTS.md is a psyop. The success of agents is predicated on their ability to adapt to human language and conventions. Just writ a readme/contributing and tell the agents to figure it out.

2026-02-26 15:16:52 CET

- reply

Let's talk, I promise I will try not to miss the meeting again 😅