Hacker News: c7b

New comment by c7b in "DuckDuckGo search saw 28% more visits after Google said people love AI mode"

c7b — Thu, 28 May 2026 12:58:47 +0000

Then we should take your word over mine. My assumption was that those A/B tests will lead to products that do increase the numbers they were measuring (retention, conversion,...) at the expense of enshittified UX (up to the point of things feeling objectively broken, like notification badges re-appearing for the same items, settings that reset after user changes, search results missing,...). At least that was my explanation for how products by major tech giants like LinkedIn, Facebook, Outlook,... could end up being shipped with such flaws. What would you say?

New comment by c7b in "DuckDuckGo search saw 28% more visits after Google said people love AI mode"

c7b — Wed, 27 May 2026 22:23:14 +0000

I would assume that they've A/B-tested any such important change extensively and basically know that it won't affect their numbers for the worse.

New comment by c7b in "The bootstrapper's EU stack for under €10 per month"

c7b — Mon, 25 May 2026 21:02:04 +0000

That's highly misleading to outright misinformation.

> Passkeys don't have to be remembered

Because you need an app for the login flow. You also don't have to remember passwords if you use a password manager app.

> don't need 2FA

Not true, a second factor in the form of eg a biometric ID or PIN is mandatory.

Phishing resistance exists, but only truly so if you completely surrender control over your device and access to your credentials. Something that the same organizations who you'll depend on for Passkeys are actively pushing for through various initiatives.

New comment by c7b in "Omarchy Is Not A Distro"

c7b — Sun, 24 May 2026 18:06:30 +0000

I agree with the author's general sentiment, but I would give Omarchy credit for some design features that I find either really innovative or well executed.

I find the launch menu the most interesting I've seen out of any desktop environment / OS. It puts easy access to your apps front and center and above esthetics, and yet it still looks great. It's not a sane request, but I'd love to be able to swap the launcher/start menu in another DE for a fully customizable version of that.

Omarchy also has a very smooth way to 'install' web apps. There exist packages that do that for you, but I've tested several and wouldn't recommend any of them. They're bloated (this should really only be a couple lines of bash, like Omarchy's [0]), some use web views (I really want an actual browser under the hood so I can have my extensions), and all that I've tested leave litter on your machine after removing apps or the software. I find web apps as .desktop files so useful that I'm actually using a hacky DIY script now (which I'm considering releasing under GPL if I ever find time to clean it up).

Also, whether you're a beginner or a pro, having a sane starting config for Hyprland is just convenient. Which tells you something about it imho. My conclusion would be similar to the OP's, if you're Omarchy-curious, try Cosmic on your distro of choice. Or at least a cleaned up version with the most egregious personal preferences (like a global keybind for opening Twitter/X) removed, if anybody cares to maintain that.

[0] https://github.com/basecamp/omarchy/blob/dev/bin/omarchy-web...

New comment by c7b in "Where are all the UK red telephone kiosks?"

c7b — Thu, 21 May 2026 20:01:55 +0000

Some suggestions:

- Not sure what they're called, but I've seen a lot of fully automated outdoor "locker stations" for packet deliveries

- Power bank "banks" or charging stations for smartphones in indoor spaces like malls

- QR codes on stickers/ads in public spaces are a sort of bridge between the physical and digital worlds

New comment by c7b in "How fast is N tokens per second really?"

c7b — Wed, 20 May 2026 22:04:03 +0000

Regarding the first, parallel requests to the same loaded model seem to work pretty well, I'm trying to find time to look more into it myself, but this may be something that might already be within reach for local models.

New comment by c7b in "An OpenAI model has disproved a central conjecture in discrete geometry"

c7b — Wed, 20 May 2026 20:25:58 +0000

One could hardly ask for a task better suited for LLMs than producing math in Lean. Running a restaurant is so much fuzzier, from the definition of what it even means to the relation of inputs to outputs and evaluating success.

New comment by c7b in "How fast is N tokens per second really?"

c7b — Wed, 20 May 2026 18:48:33 +0000

Do you have ideas/suggestions for agentic workflows that only start making sense at such speeds?

New comment by c7b in "Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks"

c7b — Wed, 20 May 2026 15:33:37 +0000

So as an oversimplified PoC, I get:

    llama-parallel -m ~/models/Qwen3.5-4B-Q8_0.gguf -ns 4 -p "Fix this Python code, answer with code only: prnt('Hello World)" -pps

    llama_perf_context_print:        load time =    1181.90 ms
    llama_perf_context_print: prompt eval time =     190.57 ms /   374 tokens (    0.51 ms per token,  1962.49 tokens per second)
    llama_perf_context_print:        eval time =    3612.25 ms /   159 runs   (   22.72 ms per token,    44.02 tokens per second)
    llama_perf_context_print:       total time =    4302.84 ms /   533 tokens
    llama_perf_context_print:    graphs reused =        155

and four answers (3 of which are immediately usable), with -ns 1 I get :

    llama_perf_context_print:        load time =    1185.61 ms
    llama_perf_context_print: prompt eval time =     187.55 ms /   305 tokens (    0.61 ms per token,  1626.27 tokens per second)
    llama_perf_context_print:        eval time =     158.92 ms /     7 runs   (   22.70 ms per token,    44.05 tokens per second)
    llama_perf_context_print:       total time =     468.85 ms /   312 tokens
    llama_perf_context_print:    graphs reused =          6

Now this is probably not the right way to use it, you should probably also use vLLM instead and it's also not a good model to use for this. But there is a real effect here that others have demonstrated, that the GPU is apparently not always maxed out while handling a single request, so sending concurrent requests can yield substantial parallelization benefits. The idea with this application would be something like this: send off the same query in parallel requests, triggering parallel tool calls, and then filter the results (filter out all failing ones, rank the rest by some simple metric of code complexity). There are probably better applications as well, I'm basically just thinking what kinds of tasks could benefit from parallelization.

New comment by c7b in "Copy Fail, Dirty Frag, and Fragnesia kernel vulnerabilities"

c7b — Wed, 20 May 2026 08:57:51 +0000

If we were to start from security first, we would be asking questions like 'how can we make sure that new code is safe?'. Manual review is great, but we can likely think of some desirable invariants for program behavior that could be tested automatically, or even formally verified. Those would come at the very start. The entire mindset right now is that the existing code is probably unsafe and we'll ship fixes as we discover its vulnerabilities. Not immediately applying updates is seen as a kind of moral failure. All major OS and most software projects were developed with this mindset of crossing your fingers at launch and then changing the tires while driving. So much so that we think of it as the natural state of software. If you start from a base of verified code, the mindset shifts. Not that there are zero vulnerabilities guaranteed in the existing code, but you become a lot more suspicious of new code.

New comment by c7b in "Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks"

c7b — Wed, 20 May 2026 07:38:53 +0000

Thanks for building what I'd hoped to find the time to build (and much better than what I would have made)! One question: do you think there is room for parallelization here, eg in the retry loop? Local models generally can handle a limited number (~ 2 digits) of concurrent requests pretty well, even on consumer hardware, which can give >10x boosts in the effective number of token/s. I've been thinking for a while about workflows that could take advantage of this, and 'fix this error' could be one (if not ideal) application. Would be curious what you think.

New comment by c7b in "Copy Fail, Dirty Frag, and Fragnesia kernel vulnerabilities"

c7b — Tue, 19 May 2026 20:15:39 +0000

Linux itself, major Linux distros, npm - none of these were designed with a security-first approach. Even the things that do help with security, like package maintenance or containerization, were more incidental to other primary goals like stability, reproducibility and so on rather than being born from a comprehensive security-first strategy. They could have been, but then things would have moved slower. They even exist, like Alpine, OpenBSD, RedoxOS, but the major ones, the ones we're talking about today, were the ones who moved faster and managed to take over. That's the fundamental issue I'm talking about, the mindset shift that would be required before we could even start the Herculean effort of rebuilding much of the existing stack with different architectures, in different languages and using different development models, always knowing that, in the past, the ones who moved fast and broke things instead tended to be the ones who succeeded.

New comment by c7b in "Copy Fail, Dirty Frag, and Fragnesia kernel vulnerabilities"

c7b — Tue, 19 May 2026 17:21:39 +0000

After the npm supply chain attacks people suggested automating delays before installing updates, now we're talking about automating update delivery... I'm afraid there won't be any easy or quick fix after decades of treating security as an afterthought.

New comment by c7b in "Hindenburg’s Smoking Room"

c7b — Sun, 17 May 2026 22:34:03 +0000

I felt a bit similar about electric cars with a trunk in the front where the engine would sit in an ICE car. But that's more about esthetic expectations, like the first cars looked similar to horse carriages.

New comment by c7b in "Trade Dollars with other startups. Book it as revenue"

c7b — Fri, 15 May 2026 21:02:36 +0000

Re taxes: Barter exchanges are considered taxable revenue by the IRS and must be reported on a 1099-B form. [0]

[0] https://en.wikipedia.org/wiki/Barter

Re investors: please list at least one credible source supporting this assertion.

New comment by c7b in "A message from President Kornbluth about funding and the talent pipeline"

c7b — Thu, 14 May 2026 18:37:21 +0000

The vast majority of the endowment isn't money (dollars in bank accounts). University endowments work like private equity funds, most of the funds will be invested in assets, most of which hardly liquid enough to reasonably convert them into cash on short notice. They could try to borrow money against the valuations of those assets, but it's not sane to take on debt in order to sustain a level of expenditures that was adjusted to a much higher level of income (true more generally). Especially when the alternative of temporarily scaling back expenses is relatively easy.

How to Fine-Tune LLMs on AMD Strix Halo

c7b — Mon, 11 May 2026 19:56:20 +0000

Article URL: https://www.promptinjection.net/p/how-to-fine-tune-llms-on-amd-strix-halo-ryzen-ai-max-395-sft-lora

Comments URL: https://news.ycombinator.com/item?id=48099860

Points: 3

# Comments: 0

New comment by c7b in "Show HN: Git for AI Agents"

c7b — Sat, 09 May 2026 03:07:04 +0000

Great idea! I was really baffled when I found out that that's not how agents, or also Chat-based IDEs like Cursor, work by default (guess how).

New comment by c7b in "Motherboard sales 'collapse' amid unprecedented shortages fueled by AI"

c7b — Fri, 08 May 2026 16:55:33 +0000

Yeah, the DGX Spark could qualify as a mini PC too. The AMD chip is sold as a laptop chip I believe, but I've mostly seen it in mini PCs. And the Framework Desktop. A brand that probably carries a lot of trust among the kind of tinkerers who would consider buying a barebone motherboard in the first place.

New comment by c7b in "Maybe you shouldn't install new software for a bit"

c7b — Fri, 08 May 2026 16:41:53 +0000

Sounds like functional programming could help with that?