Hacker News: RussianCow

New comment by RussianCow in "Firefox 153.0 Beta"

RussianCow — Mon, 20 Jul 2026 19:14:26 +0000

This seems like an arbitrary line to draw. Should your photo viewing software also not provide editing tools? If it's useful, why not?

New comment by RussianCow in "Running Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPU"

RussianCow — Thu, 16 Jul 2026 00:05:06 +0000

It depends. For something high stakes or inherently complex, sure, you don't want to have to clean up the agent's mess afterwards. But for many tasks like building web UIs, the difference in output quality is going to be small enough that iteration speed will win over quality.

With a fast enough model, I can iterate on the UI of a given screen 4-5 times before Opus finishes its first attempt.

New comment by RussianCow in "Running Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPU"

RussianCow — Wed, 15 Jul 2026 18:33:37 +0000

Once you've used a model that runs at hundreds of TPS, it's hard to go back. Everything completes so quickly that you can iterate without breaking out of flow state. My biggest gripe with slow (<50tps) LLMs is that I've lost all the mental context I built up by the time it's done, which makes it extremely difficult to explore or iterate on solutions.

New comment by RussianCow in "Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k"

RussianCow — Tue, 14 Jul 2026 16:21:49 +0000

The system prompt and available tools would likely only change for different agent types. So how they're launched probably doesn't matter.

I say this as I don't actually know how Clade Code does this, since it's not open source, but I fail to see why two agents doing the same thing launched at different times would have different tools and system prompts.

New comment by RussianCow in "Show HN: Clawk – Give coding agents a disposable Linux VM, not your laptop"

RussianCow — Mon, 13 Jul 2026 15:27:06 +0000

Yes, Codex has its own sandbox which you can disable: https://learn.chatgpt.com/docs/sandboxing?surface=app

Cursor has something similar. I don't know about Claude Code but I assume it does as well since Anthropic has open sourced their own sandboxing tool.

New comment by RussianCow in "Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k"

RussianCow — Mon, 13 Jul 2026 02:16:31 +0000

Compared to the primary agent, maybe. But it's highly unlikely that all the agents have different tools and system prompts than each other, and those account for the bulk of the context per the post.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Thu, 09 Jul 2026 04:06:09 +0000

You cannot; you must use either their Devin Desktop app or the Devin CLI.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Thu, 09 Jul 2026 01:41:09 +0000

But they don't appear to subsidize them to the same degree. I've only been using Devin for less than a month, but I've been hitting the limits of the $20/month plan way more quickly than I'd expect, and definitely more quickly than with Claude Code or Codex.

So far, Cursor provides the best value for their subscription, but I have to imagine they're basically lighting money on fire. There's no way their current pricing is sustainable.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Wed, 08 Jul 2026 22:52:09 +0000

But that 10% is the most important part! Getting the plumbing wrong means you might have bugs or your code is brittle. Getting the domain-specific business logic wrong means your product doesn't fundamentally solve the correct problem.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Wed, 08 Jul 2026 20:57:42 +0000

Ironically, Devin Desktop is one of those tools. It supports any harness that supports ACP (which is most of them)—you can use Claude Code, Codex, OpenCode, etc from the Devin Desktop UI.

I'm currently experimenting with OpenSpec[0] as the "framework" and using different subscriptions for different parts of the spec-driven process: Opus via Claude Code for exploration, Devin SWE for building, and GLM 5.2 via the Z.ai Coding Plan for verification. I don't love having to mix and match harnesses, but in practice it's barely more effort than switching models.

[0]: https://openspec.dev/

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Wed, 08 Jul 2026 20:51:18 +0000

But some requirements you don't realize you have until you start building. With a fast model, you can surface those really quickly and have more time to iterate and explore different solutions. With a slower but smarter model, you just hope that what it produces after an hour is what you were imagining.

And yes, with Fable, the chance of that is higher than with SWE/Composer, but in my experience it's not so much higher that the extra time and cost is worth it. But it certainly depends on your goals and what you're building.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Wed, 08 Jul 2026 20:31:47 +0000

The regular one (not the fast variant) is free but slow. The "Lightning" variant (which uses Cerebras and gets supposedly 1000 TPS) costs $12.50/M output, $2.5/M input, $1/M cached input. So it's quite a bit more expensive than SWE 1.6.

New comment by RussianCow in "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"

RussianCow — Wed, 08 Jul 2026 20:25:20 +0000

The "Lightning" (Cerebras) variant isn't free, only the regular one, which runs closer to 50 TPS in my experience with SWE 1.6.

New comment by RussianCow in "Road to Elm 1.0"

RussianCow — Mon, 06 Jul 2026 14:58:11 +0000

That's not why Elm is an ideal language for LLMs: it's because, if it compiles, it's most likely working software. Agentic workflows have gotten significantly better over the last year, so LLMs using languages like Elm, Haskell, or even Rust have an amazing feedback loop where even lower quality models can keep trying until things compile.

New comment by RussianCow in "Command and Conquer Generals natively ported to macOS, iPhone, iPad using Fable"

RussianCow — Sun, 05 Jul 2026 16:45:31 +0000

> asking an LLM to reverse engineer and make your own plugin is trivial.

If you already have engineers on staff, a few tens (or even hundreds) of dollars per month per plugin is likely a rounding error budget-wise. If you don't have your own engineers, you're probably not going to be able to produce something as good (reliable, well thought out, etc) as a commercial offering.

I had the same gut reaction as you, but the reality is much more subtle. We work with several clients who are bought into at least one of these ecosystems, and there's no way the math ever works out in favor of building an in-house solution.

New comment by RussianCow in "Android Developer Verification: Threat masquerading as protection"

RussianCow — Thu, 02 Jul 2026 17:38:31 +0000

At least 3 times a month. I have a rental property and my tenant prefers to mail a check instead of paying extra to pay electronically. My spouse gets paid by check for dumb reasons I won't get into. I sometimes get dividends from my insurance company via check. And then several family members still prefer to use checks to pay each other back instead of Venmo or other electronic services.

I blame it on the fact that the US doesn't have a free electronic bank transfer system like the rest of the developed world.

New comment by RussianCow in "Android Developer Verification: Threat masquerading as protection"

RussianCow — Thu, 02 Jul 2026 15:44:51 +0000

> Also why do you need bank app on your phone?

Many banks gate features like mobile check deposit behind the native app. The nearest ATM is 20 minutes away from my house, so unfortunately I consider this feature essential.

New comment by RussianCow in "I used Claude Code to get a second opinion on my MRI"

RussianCow — Tue, 30 Jun 2026 06:11:11 +0000

> En, I think you’re just trying to justify your pre-existing position that this can’t work.

I never said it can't work. I just said that finding the correct medical digagnosis is different than finding a solution to a software problem.

New comment by RussianCow in "I used Claude Code to get a second opinion on my MRI"

RussianCow — Mon, 29 Jun 2026 06:05:40 +0000

Also, there are multiple "correct" ways to code something, so imperfect code that solves the problem is still useful. A medical diagnosis is either correct or incorrect.

New comment by RussianCow in "I used Claude Code to get a second opinion on my MRI"

RussianCow — Mon, 29 Jun 2026 06:03:39 +0000

At that point, cut out the LLM and just see the radiologist.