Hacker News: 0xc133

New comment by 0xc133 in "AMD expects to outsell Intel in Q4 for the first time in history"

0xc133 — Wed, 04 Nov 2026 18:47:40 +0000

Yeah, about that. Those cards all have MSRPs that have moved north by over $500 since launch.

New comment by 0xc133 in "Stalwart v1.0: The Fastest Mail Server on the Planet"

0xc133 — Mon, 26 Oct 2026 17:49:37 +0000

Well, I know what I'm upgrading tonight. (I didn't see any mention of this in the blog post, but they also implemented batch-upload for the vandelay tool, so exporting your archive into a new Stalwart server is much faster now, too!)

New comment by 0xc133 in "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"

0xc133 — Fri, 23 Oct 2026 20:18:11 +0000

Anthropic released Claude Opus 4.6 on February 5, 2026. What year are you in, man?

New comment by 0xc133 in "New Apple TV 4K"

0xc133 — Wed, 14 Oct 2026 04:21:19 +0000

I bought a Sonos soundbar (the Ray) to dip a toe in the waters, so to speak.

I’ll never buy another Sonos product ever again. Terrible experience all around. App is trash, soundbar drops off wifi for no reason, songs will stop playing over Airplay and it’ll just disappear as an Airplay speaker for hours and then show back up randomly and work fine again.

New comment by 0xc133 in "Restoring 5 GHz Wi-Fi on an LG C5 by changing its webOS region"

0xc133 — Fri, 04 Sep 2026 14:33:14 +0000

Why the hell would you ever let a TV on any network?

New comment by 0xc133 in "New Mac mini, featuring M6 and M5 Pro"

0xc133 — Tue, 25 Aug 2026 22:01:57 +0000

The cheapest Steam Deck available now is $789.

New comment by 0xc133 in "Qwen3.8 27B scores 52 on Artificial Analysis"

0xc133 — Wed, 19 Aug 2026 02:49:25 +0000

How were you accessing the model? I was using ollama.com’s cloud hosting the first few times I ran into it, and then it happened again the first time I tried Fireworks.ai before I made it even 10% of the way through the context window.

New comment by 0xc133 in "Qwen3.8 27B scores 52 on Artificial Analysis"

0xc133 — Mon, 17 Aug 2026 23:37:09 +0000

deepseek-v4-flash-0731 has been awfully prone to infinite looping output in reasoning for me, and once it hallucinated in the middle of going in circles that I had instructed it to start using Yoda-speak, which… I don’t have any idea where that came from.

New comment by 0xc133 in "Ask HN: Alternatives to GitHub"

0xc133 — Mon, 17 Aug 2026 17:11:13 +0000

*Forgejo

New comment by 0xc133 in "Hax – a minimalist, terminal-native coding agent written in C"

0xc133 — Wed, 12 Aug 2026 18:10:51 +0000

asciinema is so good! Love seeing the player on the site.

New comment by 0xc133 in "Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows"

0xc133 — Mon, 10 Aug 2026 20:35:41 +0000

I was able to use rope/yarn scaling with llama.cpp to extend the context window to 256K and it seems to be pretty usable on the debugging-and-bugfixing session I have that’s 216K tokens deep. No infinite output loops, reasoning is still coherent, tool calls appear to be passing and failing at roughly the same rate as a fresh context window. Haven’t tried going to 512K or higher yet but 256K definitely seems usable to me on a single Radeon R9700.

I added these arguments to my llama-server call, using the official GGUF release from Meta’s account on HF.

`--rope-scaling yarn --rope-scale 2 --yarn-orig-ctx 131072 --override-kv muse-glimmer.context_length=int:262144,dflash.context_length=int:262144`

New comment by 0xc133 in "DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis"

0xc133 — Fri, 31 Jul 2026 14:39:52 +0000

because once they run out of tokens they can’t do their jobs anymore

New comment by 0xc133 in "Running Kimi K3 on a M1 Max"

0xc133 — Wed, 29 Jul 2026 00:37:03 +0000

I had my clanker implement this idea in a standalone Rust server that speaks IMAP and SMTP and proxies your emails to an OpenAI endpoint you configure: https://tangled.org/clee.sh/posthorn

Works in mutt; other MUAs may vary.

New comment by 0xc133 in "The death and rebirth of my home server"

0xc133 — Sun, 19 Jul 2026 18:12:39 +0000

I discovered that barely-used enterprise Intel SSDs (DC S3610, S3700, S3710) were cheap on ebay a year or so ago and I may have bought a few dozen. These things have remaining endurance measured in petabytes. Sure, it’s overkill to have a 200GB enterprise SATA disk as the boot volume for a Raspberry Pi, but this is exactly my kind of overkill.

New comment by 0xc133 in "Show HN: Getting GLM 5.2 running on my slow computer"

0xc133 — Fri, 10 Jul 2026 06:21:53 +0000

So I’ve been thinking about this problem a lot, specifically as it relates to running LLMs at home, and I’ve been using GLM-5.2 to make an SMTP/IMAP-to-LLM gateway.

https://tangled.org/clee.sh/posthorn

New comment by 0xc133 in "The Token Compression Illusion: Why I'm Skeptical of RTK"

0xc133 — Thu, 18 Jun 2026 22:56:22 +0000

I also have become a maki convert and I really like it. I ran into an issue with the dynamic model provider that I should probably make a patch for; list_models doesn’t use the ` models` output at all but instead tries to look up ` resolve`’s base URL + /v1/models, which breaks on a provider like Z.ai which doesn’t have /v1/ anywhere in the path…

New comment by 0xc133 in "Running local models is good now"

0xc133 — Tue, 16 Jun 2026 16:37:56 +0000

With yarn and rope scaling arguments for llama.cpp you could run qwen3.6-27B with 1M context… if you have enough memory to store it.

New comment by 0xc133 in "Handling the great code forge fragmentation"

0xc133 — Wed, 20 May 2026 21:23:50 +0000

> It'll be a very slow decay

Two ways. Gradually and then suddenly.

New comment by 0xc133 in "New Lifetime Plex Pass Pricing"

0xc133 — Tue, 19 May 2026 14:54:54 +0000

I’m still salty that their last app platform rewrite dropped Watch Together as a supported feature. Server still supports it, browser clients still support it, as well as the older Plex apps if you were prescient enough to turn off automatic app updates and manually keep refusing to update the Plex app.

This is a feature I use multiple times per week with friends who live states away. I can’t believe they just dropped it.

New comment by 0xc133 in "No more JetBrains products for me"

0xc133 — Mon, 18 May 2026 23:46:11 +0000

> so hyper-sensitive that a couple seconds of loading screen seen a few times a day is enough to trigger them

Reportedly the xz supply chain compromise was first noticed by someone who reported his SSH connections took an extra 500 milliseconds.

https://thegeekinsights.com/how-a-hacker-saved-the-internet/