Hacker News: vasachi

New comment by vasachi in "Grit: Rewriting Git in Rust with agents"

vasachi — Wed, 10 Jun 2026 05:08:13 +0000

Surely git’s source is already in LLM’s training corpus. So this is far from clean room approach.

New comment by vasachi in "Bill to block publishers from killing online games advances in California"

vasachi — Sat, 16 May 2026 06:07:19 +0000

You wouldn’t get one if you followed the law and released the server.

New comment by vasachi in "How Claude Code works in large codebases"

vasachi — Fri, 15 May 2026 05:39:47 +0000

I find that most “techniques” are basically user hallucinations. Simple plan-write-refactor loops and trivial CLAUDE/AGENTS.md, generated by the harness itself, work nicely. Maaaaaaaaaybe write a skill or two, but usually it’s better to just write a script.

New comment by vasachi in "Access to frontier AI will soon be limited by economic and security constraints"

vasachi — Fri, 15 May 2026 05:26:08 +0000

I can’t compare raw intelligence of these models, and I certainly can’t say anything about their advances in mathematics (without repeating press releases). But, erm, does it really matter? It’s not like some engineer somewhere will vibe-calculate how much weight a bridge can hold.

Well, yes, someone probably will do that. But I’m pretty sure there will be consequences for the engineer errors in this vibe-calculations.

New comment by vasachi in "Access to frontier AI will soon be limited by economic and security constraints"

vasachi — Fri, 15 May 2026 05:09:13 +0000

I’ve used DeepSeek 4 Pro through Claude. It’s fine. Plans are similar to what sonnet/opus make. Same massage-the-plan -> massage-the-code loop. Maybe the code is a bit worse, but that’s the “months behind” thing.

The thing is, vast majority of code tasks aren’t a venture into the unknown. We as an industry for the most part build CRUD interfaces and dashboards. That can be achieved, with supervision, with frontier open-weights models quite well.

New comment by vasachi in "Claude Code to be removed from Anthropic's Pro plan?"

vasachi — Wed, 22 Apr 2026 05:53:36 +0000

The economy, not the workers :) It feels like pretty soon white collar workers will be in a “You have nothing to lose but your chains” situation. Except we are not as fit as the proletariat of the past.

New comment by vasachi in "The Bromine Chokepoint"

vasachi — Sun, 19 Apr 2026 21:00:44 +0000

Look, it is calming to think that Trump is somehow a Russian agent, and not a sign of collapsing political class, but let’s be realistic here.

If it is indeed true, then Russia managed to put its puppet as The President of the United States of America. Twice. Without triggering any red flags from a dozen or so of US TLAs. That would be the best intelligence operation in the history of everything. At that point, USA should just put the chairs away and turn down the lights.

Epstein files yes, I grant you that.

New comment by vasachi in "Internet Protocol Version 8 (IPv8)"

vasachi — Thu, 16 Apr 2026 06:10:05 +0000

IP is what, four layers of protocols lower than OAUTH?

New comment by vasachi in "Supply chain nightmare: How Rust will be attacked and what we can do to mitigate"

vasachi — Fri, 10 Apr 2026 18:01:11 +0000

Doesn't USA do that all the time with .com and such?

New comment by vasachi in "Ex-GitHub CEO launches a new developer platform for AI agents"

vasachi — Wed, 11 Feb 2026 07:52:42 +0000

Didn’t Juicero get more than a $100M? Do you think they had a clear plan? How much did Rome get? Did they have a clear plan?

New comment by vasachi in "Windows 10 Deadline Boosts Mac Sales"

vasachi — Sat, 25 Oct 2025 16:13:40 +0000

First, they released it only 10 years ago, and second, Microsoft marketed it as the last windows version. So yes, this is a own goal)

New comment by vasachi in "We reverse-engineered Flash Attention 4"

vasachi — Sun, 28 Sep 2025 07:26:41 +0000

Are… Are you pulling out the “It depends on what the meaning of the word ‘is’ is.” trick?

New comment by vasachi in "Pnpm has a new setting to stave off supply chain attacks"

vasachi — Thu, 18 Sep 2025 08:39:36 +0000

If only there was a high-ranking official at Microsoft, who could prioritize security[1]! /s

[1] https://blogs.microsoft.com/blog/2024/05/03/prioritizing-sec...

New comment by vasachi in "About the security content of iOS 15.8.5 and iPadOS 15.8.5"

vasachi — Wed, 17 Sep 2025 09:09:09 +0000

I'm pretty sure Google can buy a femtocell to simulate local mobile network of their own.

New comment by vasachi in "Universal Tool Calling Protocol (UTCP)"

vasachi — Thu, 21 Aug 2025 08:06:53 +0000

Initial commit: Jun 24, 2025. Already has a migration path from 0.x to 1.x.

Why am I feeling so old now?

On a more serious note, do any models support this?

New comment by vasachi in "Auf Wiedersehen, GitHub"

vasachi — Mon, 11 Aug 2025 16:04:04 +0000

Given that it’s going to be a part of “Microsoft CoreAI organization”, whatever that is, there will be even more AI stuff.

New comment by vasachi in "Who got arrested in the raid on the XSS crime forum?"

vasachi — Sun, 10 Aug 2025 07:40:10 +0000

Ah yes, because only Russians can be criminals. And also Russians cannot be born in Eastern Europe.

New comment by vasachi in "JavaScript broke the web (and called it progress)"

vasachi — Fri, 20 Jun 2025 09:44:57 +0000

Usual arguments for "upgrading" are:

* We can't hire specialists for older frameworks.

* We can't generate positive hipe over older technologies.

* New technologies are better, so we will deliver features faster.

In reality, it's almost always resume-driven development.

New comment by vasachi in "Ask HN: Cursor or Windsurf?"

vasachi — Mon, 12 May 2025 05:23:09 +0000

I’d just wait a bit. At current rate of progress winner will be apparent sooner rather than later.

New comment by vasachi in "Mozilla launching “Thundermail” email service to take on Gmail, Microsoft 365"

vasachi — Wed, 02 Apr 2025 20:57:11 +0000

My coworker uses thunderbird since time immemorial, and I don’t think it ever corrupted his mailbox.