Hacker News: Kubuxu

New comment by Kubuxu in "The economics of software teams: Why most engineering orgs are flying blind"

Kubuxu — Mon, 13 Apr 2026 13:46:09 +0000

I have TDD flow working as a part of my tasks structuring and then task completion. There are separate tasks for making the tests and for implementing. The agent which implements is told to pick up only the first available task, which will be “write tests task”, it reliably does so. I just needed to add how it should mark tests as skipped because it’s been conflicting with quality gates.

New comment by Kubuxu in "Tailslayer: Library for reducing tail latency in RAM reads"

Kubuxu — Fri, 10 Apr 2026 03:43:18 +0000

That somewhat existed in servers in the past, my R720xd had RAM mirroring mode. IDK if it used it for reducing latency, but you could take out a stick and the server would continue running as normal and report an alarm in iDRAC.

New comment by Kubuxu in "The bridge to wealth is being pulled up with AI"

Kubuxu — Tue, 24 Mar 2026 16:45:18 +0000

Yeah, after two paragraphs, my eyes started skimming it due to low signal-to-noise ratio.

New comment by Kubuxu in "OpenAI raises $110B on $730B pre-money valuation"

Kubuxu — Sat, 28 Feb 2026 10:56:21 +0000

Just try using Claude with API for an hour and you will see that the subscriptions are definitely not profitable (unless they percent off “partying but dormant” is very high).

New comment by Kubuxu in "Evaluating AGENTS.md: are they helpful for coding agents?"

Kubuxu — Tue, 17 Feb 2026 08:14:06 +0000

Reasoning, in majority of cases, is pruned at each conversation turn.

New comment by Kubuxu in "Attention at Constant Cost per Token via Symmetry-Aware Taylor Approximation"

Kubuxu — Wed, 04 Feb 2026 19:45:08 +0000

A paper on the same topic: On the Expressiveness of Softmax Attention: A Recurrent Neural Network Perspective, Gabriel Mongaras, Eric C. Larson, https://arxiv.org/abs/2507.23632

Video presentation if someone prefers it: https://www.youtube.com/watch?v=PN3nYBowSvM

Linear attention is a first-degree approximation of Softmax attention, and model performance gets better as you increase the degree of the Taylor approximation.

I'm thinking about adapting an existing model to Taylor-approximated attention. I think it should be possible with some model surgery and rehabilitation training.

New comment by Kubuxu in "Ask HN: How can I get better at using AI for programming?"

Kubuxu — Sun, 14 Dec 2025 09:36:42 +0000

Definitely agreed, that is what I do as well. At that point you have good understanding of that code, which is in contrast to what the post I responded suggests.

New comment by Kubuxu in "Ask HN: How can I get better at using AI for programming?"

Kubuxu — Sat, 13 Dec 2025 22:12:12 +0000

I don’t see how I would feel comfortable pushing the current output of LLMs into high-stakes production (think SLAs, SRE).

Understanding of the code in these situation is more important than the code/feature existing.

New comment by Kubuxu in "How the UK lost its shipbuilding industry"

Kubuxu — Mon, 10 Nov 2025 11:28:03 +0000

And DEF is ~33% urea in water.

New comment by Kubuxu in "Ask HN: How to stop an AWS bot sending 2B requests/month?"

Kubuxu — Sat, 18 Oct 2025 11:33:16 +0000

I've done that in the past (8+ years ago) with raw IP sockets.

New comment by Kubuxu in "I regret building this $3000 Pi AI cluster"

Kubuxu — Fri, 19 Sep 2025 18:05:01 +0000

For 8+ bays you just need a SAS HBA card and one free PCI-E slot. Not to mention that many motherboards will have 6+ SATA ports already.

If anything, 2nd hand AMD gaming rigs make more sense than old servers. I say that as someone with always off r720xd at home due to noise and heat. It was fun when I bought it during winter years ago, until summer came.

New comment by Kubuxu in "Disrupting the DRAM roadmap with capacitor-less IGZO-DRAM technology"

Kubuxu — Wed, 10 Sep 2025 03:00:58 +0000

Traditional CMOS D-flipflop takes 16 transistors, SRAM cell takes 6 (two looped inverters and two data-line selection transistors).

The capacitor in DRAM is usually realised as an enlarged gate of a transistor AFAIK.

New comment by Kubuxu in "PWM flicker: Invisible light that's harming our health?"

Kubuxu — Thu, 19 Jun 2025 09:04:53 +0000

It is always PWM under the hood, the question is, how much was spent (or not) on the filtering network out of the PWM. Is it closer to buck converter or is it straight up flicker at the output.

New comment by Kubuxu in "RenderFormer: Neural rendering of triangle meshes with global illumination"

Kubuxu — Sun, 01 Jun 2025 09:26:38 +0000

Nobody runs 4096 samples per pixel. In many cases 100-200 (or even less with denoising) are enough. You might run up to low-1000 if you want to resolve caustics.

New comment by Kubuxu in "Beware of Fast-Math"

Kubuxu — Sat, 31 May 2025 11:26:16 +0000

IIRC reordering additions can cause the result to change which makes auto-vectorisation tricky.

New comment by Kubuxu in "Lock-Free Rust: How to Build a Rollercoaster While It's on Fire"

Kubuxu — Fri, 16 May 2025 15:56:01 +0000

You can think of closing the channel as sending a message “there will be no further messages”, the panic on write is enforcement of that contract.

Additionally the safe way to use closing of a channel is the writer closing it. If you have multiple writers, you have to either synchronise them, or don’t close the channel.

New comment by Kubuxu in "Progress toward fusion energy gain as measured against the Lawson criteria"

Kubuxu — Fri, 09 May 2025 07:33:00 +0000

Majority of good spots for hydro were already built up, and if they weren’t, good luck with NIMBY.

New comment by Kubuxu in "Dummy's Guide to Modern LLM Sampling"

Kubuxu — Sun, 04 May 2025 19:16:24 +0000

Beam Search sampling is sometimes getting used

New comment by Kubuxu in "Getting forked by Microsoft"

Kubuxu — Tue, 22 Apr 2025 15:18:37 +0000

The question is, "If I look at this repo, who owns the copyright?" Sure, you could move the original LICENSE into a directory. Still, if the files are intermingled, you should have a prominent notice that says, " Hey, these files have mixed copyright ownership."

New comment by Kubuxu in "Getting forked by Microsoft"

Kubuxu — Mon, 21 Apr 2025 16:06:47 +0000

In case of deps, the dependency comes with its own LICENSE file.

In this case the code is essentially forked, integrated and intermingled, so that is why it should be in the LICENSE file.

If it was file or two, it would be fine to add a comment pointing to the license file in the repo, if it was a directory, or to copy it verbatim to that file. It all the copied code was in a directory then having it in directory would be fine.

In this case it looks like they took the original code and heavily modified it, so the simplest way to solve it is one LICENSE with both notices.