Interesting. I was leaning towards the Luma Ultra based on reviews.
24.02.2026 16:49
👍 2
🔁 0
💬 1
📌 0
Interesting. I was leaning towards the Luma Ultra based on reviews.
Announcing Oxfmt Beta
* 100% Prettier Compatibility
* Additional File Formats
* Tailwind CSS Integration
* Import Sorting
* package.json Sorting
* Embedded Language Formatting
* Node.js API
* `--migrate prettier`
* config overrides
* Full IDE support
oxc.rs/blog/2026-02...
which ones are these? xreal?
We've been hard at work on a new website with new guides, better docs, more translations, even our very own blog. We think it looks pretty sharp, too. It's now easier than ever to just build things with AT Protocol.
atproto.com
it worked eventually but it struggled a lot. it also went to tangled so not sure what happened in the end.
@standard.site trying to teach claude standard.site lexicons but completely unable to access anything, any advice?
This is quite interesting, I may explore integrating this to mataroa
London friends!! 🦞🇬🇧
I’m hosting an AI Agents Hack Night at
@nwspk
in East London on Feb 17.
Bring your laptop, your friends, and your agent swarm ;)
👉 luma.com/46cy08st
#AIAgents #AIHackNight #LondonTech #BuildInPublic #OpenClaw
welcome!
wow that might be the next revolutionary rust rewrite for devex!
Cross-posting this one from METR. If you've seen this graph before, you're probably going to be pretty interested in the new update.
If you've never seen this graph before, please just ignore this post. Just a guy and his charts, nothing to see here.
New search filters
This doesn't entirely replace the old behavior, it mostly adds to it
Handle typeahead suggestions are only shown when the input starts with @
@michaelnielsen.bsky.social Hi there! I'm thinking about an atproto-based book notes network with social features. I would be interested in your opinion! Repo here: github.com/sirodoht/alc... -- want to talk about it?
kagi.com has many tricks like that!
say more, like what kind of things that gpt 5 thinking is not as great?
This was a worthwhile use of my morning 🔉
oh i see! but you can't have an underscore as part of the domain name, right?
🤔 it has to resolve on the internet so probably no?
excited to share that we are following through on our earlier commitments and putting together an independent+neutral organization to house the DID PLC system, includes the directory service
I hope it improves!
:(
@jonty.bsky.social did you forget about this list in 2025 too? 😄
jonty.co.uk/made
Designs for Semble
1/ 📢 @cosmik.network has been awarded $1M in grant funding by Open Philanthropy and @asterainstitute.bsky.social These generous grants will support our development of Semble - a social "micro-knowledge" network for researchers on Bluesky/ATProto. Think Are.na + Goodreads for research!
what did you think of antiindividualistic individuality?
I first wrote up the idea for this 6 years ago, but it took a huge amount of work to actually turn it into reality. Thank you to Orion for the heavy lifting, and all the other Automerge contributors for making it happen!
Red graffiti that says: “SIR STARMER” IS A TORY IN A RED TIE
Blue graffiti that says: BE PROUD OF WHO YOU ARE ❤️
Orange graffiti that says: STOP LETTING POLITICIANS THAT DONT CARE ABOUT US BACK INTO OFFICE
Surprisingly polite graffiti at the local skate park
Some notes on Grok 4: excellent benchmark scores, a mid-quality pelican and a launch that was overshadowed by this week's disastrous Grok 3 system prompt update simonwillison.net/2025/Jul/10/...
2.5 hours (!)
quite frequently!
AI systems are able to pass "theory of mind" tests - Recent research shows post-2022 models achieve 70-93% success rates on these tests, comparable to a 7-9 year old child. This forges trust in ways we are not really prepared for.
