Using Talos in my homelab has made it so stable I've forgotten I actually selfhost a bunch of things. It's boring, in a very good way!

This looks delicious. 😋 No more aws.String-like helpers!

This Christmas vacation I'm tackling the backlog of my Steam library. Some of it at least. The winter sale has already managed to grow it quite a bit.

@atproto.com need some help debugging bsky backend querying my feed generator. At "coldstart" bsky produces a http5xx. From my server access logs I can see it hit the .well-known/did.json and then nothing else. After refreshing the bsky page, it starts working and properly hitting the xrpc API.

Gnome has been my desktop of choice for the better part of my life. Giving back a little is the least I can do.

2 of the 6 consumer SSDs (that I bought for my NAS) have died in less than a year. Both of them from the same brand - fanxiang. The other 2 brands - Kingspec & Crucial - are still going strong. Be smart, do NOT trust fanxiang with your data.

Put together a bsky feed for posts related to running AI locally: bsky.app/profile/did:...

Openness for Nemotron includes open evaluation! Eval transparency completes the last puzzle in the E2E repro lifecycle for LLMs. Check out our step-by-step blog on how to use the same pipeline we used for the evaluation of Nemotron 3 Nano through Nemo Evaluator: huggingface.co/blog/nvidia/...

If you've been meaning to give Zed a spin during the slower season, now's the time. Use code ZEDSEASON at checkout for $15 in LLM spend on us. Limited supply, valid through Dec 31. Already using Zed? Share this with someone who needs a nudge.

The main thing Ollama (and LM studio) has going for it, is that everyone has integrated it everywhere. IDEs, Obsidian plugins, self-hostable AI apps.

I tried to give Ollama a fair chance. I was tempted by the automatic VRAM management it does. But everything else is just inferior to llama.cpp. The modelfile concept is just weird. No support for sharded GGUFs. And why use Go templates when all of the ecosystem uses Jinja2?

I did not understand HuggingFace before. I had done some LLM courses out of curiosity, but HuggingFace didn't click. Until I started running things locally. Base models, adapters, fine-tunes, quantizations. I get it now!

At some point I need to more seriously test out ROCm vs Vulkan performance for local models I use. Haven't spent much time on ROCm currently as AMD broke it in the latest linux-firmware release. It has been reverted already, but not sure what the release cadence is there.

llama.cpp just recently merged in (github.com/ggml-org/...) support for serving multiple models, effectively making llama-swap a lot less useful. I love it. Just makes me wonder if llama.cpp is intending to compete with ollama (and LM Studio to some extent).

I wonder.. is it possible to use host @zed.dev's Zeta model locally? 🤔

I wish @zed.dev's Open AI compatible provider support would attempt to auto-detect the models from the server. It's such a chore to manage the models manually in settings.json.

After a bit of experimentation, I've settled on llama-swap and llama.cpp combo to run my local models. LM Studios was a bunch of UI that I wasn't going to use, and having the option to go and tweak every lever that llama.cpp has feels good.

ISP: We see your bandwidth usage has spiked... Me: Oh don't mind me, I'm just downloading some models from HuggingFace.

Got myself @frame.work desktop as an early christmas present. Been having a blast setting it up and playing around with local LLMs.

Switched my personal email from Proton to mailbox.org. Webmail UI is a slight downgrade, but the knobs you get for spam control, etc is wonderful. It's also cheaper. And my main reason for switching - up to 25 custom domains.

Ahh, the wonders of modern world. Junie writes some unit-tests, and then spends all the monthly token quota attempting get the tests passing. I guess the fact that it didn't just remove the tests could be considered a win.

I wrote a bunch of implementation docs for a browser extension. Then let Gemini 2.5 Pro make an implementation plan out of it, in distinct stages. And now I'm feeding these into Junie one by one, and trying not to criticize the output. Vibing they call it. I call it stressful.

Updating Windows is SLOOOW and seemingly never ending. Downloading 100mb update takes more time than downloading 500GB game on Steam. After restarting to finish the install, for 5 minutes it claims there's no more updates, and then it finds yet another one. I think I've already done 3 restarts...

Take Zed's code, throw out half (or more) and build a blazing fast Obsidian clone? 🤔

As the first task for JetBrains Junie, I had it move some test code around. The UI is super nice - very easy to jump into diffs & terminal out put of different commands being executed. The calls to underlying LLM models seemed a lot slower than claude-code, though.

My 30-day Claude Pro subscription is expiring tomorrow. Next, will give the JetBrains AI Pro (the cheaper option) a go for a month to try out Junie. Based on the amount of tokens claude-code has gone through in a month, I'm expecting to consume the allocated token quota fairly quickly with Junie.

We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers. The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.

Are working parents over-employed? 🤔 It sure does feel like juggling several jobs...

PSA: Don't forget to check your old portable devices every once in a while for deformations! I just pulled two fairly spicy pillows (expanded batteries) from old laptops that were stashed away out of sight.

Last week, I gave @zed.dev another go (on Linux). I had tried it briefly 6 months ago. Since then, it has received a lot of polish and love. So much so, that now I'm rocking it as my secondary editor. Still has some rough edges, but I can see it becoming my primary workhorse soon.

You can have two Big Things.

This one is a no brainer for me, I've been getting value from GNOME for *decades* at this point, and for decades to come I hope.

Printed out the RFC and read it thoroughly. I can say with confidence now that LLMs suck at using RFCs as a base for implementing something. Let's see how far I get with the experiment now that I know what the vibe-machine is expected to implement.

Been attempting to vibe-code a SMTP server based on the RFC with claude-code. For a week, it insisted that the maximum line-length across the board is 998 characters. Suddenly, yesterday, it started saying that everything is wrong and the maximum length is actually 512 octets.

Something I realized the other day is that the concept of being "fluent" in a language applies to programming languages in the same way it does for natural language. A native C-speaker who has to speak Rust when they aren't fluent in Rust will (subconsciously) think in C and translate to Rust.

Any recommendations which cargo doc MCP server to use with claude-code? #rustlang

Why do some have a shit time with LLMs for programming while others love it? To succeed, the latter group tacitly creates tons of scaffolding and gain weird new skills. While it works, this posts explains how doing all that is an incidental consequence of bad interaction design in coding AI agents.

If nothing else, LLMs will solve the no.1 problem developers face daily - naming of things.

Took claude-code out for a spin over the weekend. Some initial thoughts 🧵

Everyone pays for search. The difference is whether you're paying with your attention, time, and private data - or with your wallet.

It's allergy season and the pollen is so bad right now that I was sneezing non-stop. Took some meds, but apparently they make me depressed. I guess it's time to put on some sad tunes and wash the blues away.

Been experimenting with shifting my day forward by 1.5h. All so that I could "make" time for a nice long daily walk. After a week of this, I'm finally used to waking up earlier. And even though getting out the door is still a bit challenging, the brisk walk in the nature is definitely worth it.

There was a discussion on HN about moving from k8s to simple podman (and systemd) setup. After some time contemplating doing the same, I came to the conclusion that I like k8s just too damn much (warts and all).

Another great video from 3Blue1Brown. This time an introduction to quantum computing. youtu.be/RQWpF2Gb-gU?...

Earlier this week Spain experienced a country wide blackout. It gave my home networking setup a chance to prove itself. My homelab and networking is backed up by an UPS, but with the lab running, it would only last about 15 mins.

Rust 1.86 just released: https://blog.rust-lang.org/2025/04/03/Rust-1.86.0.html and it stabilizes trait upcasting - especially useful in connection to the Any trait #rustlang 🦀🎮

For the record: I passionately hate the daylight savings fiasco! It's been half a week since the turning of clocks and I'm still feeling majorly jetlagged.

As a pre-launch of the Launch Week 14, we’re announcing the initial release of Postgres Language Server! It includes: - Autocompletion - Syntax Error Highlighting - Typechecking ⁃ Linting Shout-out to @psteinroe.com and juleswritescode for the hard work! supabase.com/blog/postgre...

That's it. Solo Leveling Season 2 is finished. What to look forward to now on Saturdays?

Steam sale resistance LEVEL UP++. I think I've finally managed to break the cycle of buying discounted games and never even opening them. But does that mean... I need to start using the past tense when describing myself? "I used to be a gamer." Sounds awful 😭