Hacker News: esafak

New comment by esafak in "Cirrus Labs to join OpenAI"

esafak — Sat, 11 Apr 2026 16:22:43 +0000

What was good about it? It looks pretty ordinary to me: https://cirrus-ci.org/features/

New comment by esafak in "Code is run more than read (2023)"

esafak — Fri, 10 Apr 2026 16:20:23 +0000

That's the opposite of 'great'. Good code is that which can be refactored.

New comment by esafak in "Code is run more than read (2023)"

esafak — Fri, 10 Apr 2026 16:16:21 +0000

More important than:

"We arrived at a little model that expresses the relative importance of various factors in software development..."

New comment by esafak in "Research-Driven Agents: When an agent reads before it codes"

esafak — Fri, 10 Apr 2026 13:44:21 +0000

Agreed. This is nothing but RAG, which helps when the task benefits from more knowledge. They're just going into the details of their application.

New comment by esafak in "Model-Based Testing for Dungeons & Dragons"

esafak — Fri, 10 Apr 2026 13:32:05 +0000

Yet another specification language! And it also has a new sibling for distributed protocols: https://quint-lang.org/choreo

Any opinions on this one for software development?

New comment by esafak in "Intel 486 CPU announced April 10, 1989"

esafak — Fri, 10 Apr 2026 13:21:13 +0000

https://en.wikipedia.org/wiki/VESA_Local_Bus for the younger crowd.

New comment by esafak in "We've raised $17M to build what comes after Git"

esafak — Fri, 10 Apr 2026 13:19:13 +0000

Some people are not sensitive to quality. A car is a car, a shower gel is a shower gel, etc. In the computer world, they curiously congregate around Microsoft...

New comment by esafak in "We've raised $17M to build what comes after Git"

esafak — Fri, 10 Apr 2026 12:50:13 +0000

I am actively looking for a replacement for GitHub and would prefer something that is not based on git.

New comment by esafak in "We've raised $17M to build what comes after Git"

esafak — Fri, 10 Apr 2026 04:46:59 +0000

Don't you read the PRs?

New comment by esafak in "We've raised $17M to build what comes after Git"

esafak — Fri, 10 Apr 2026 04:24:22 +0000

If you think like that why invest in software at all; the AI will do everything?

Does AI make reading or writing stacked PRs any nicer? No, it does not.

New comment by esafak in "The Training Example Lie Bracket"

esafak — Fri, 10 Apr 2026 01:59:02 +0000

Sooner if you explain why.

New comment by esafak in "Claude mixes up who said what"

esafak — Thu, 09 Apr 2026 15:31:09 +0000

Unless it is a bug in CC, which is likely as not, the LLM is failing to keep the story straight. A human could do the same; who said what?

New comment by esafak in "Launch HN: Relvy (YC F24) – On-call runbooks, automated"

esafak — Thu, 09 Apr 2026 13:08:47 +0000

They claim a 12% lead (from 36% to 48%) over Opus 4.6 in a RCA benchmark: https://www.relvy.ai/blog/relvy-improves-claude-accuracy-by-...

New comment by esafak in "Show HN: Hippo, biologically inspired memory for AI agents"

esafak — Thu, 09 Apr 2026 04:14:13 +0000

I would look at information theory for inspiration; notions of information gain and surprisal.

New comment by esafak in "Revision Demoparty 2026: Razor1911 [video]"

esafak — Wed, 08 Apr 2026 14:01:27 +0000

I think a modern equivalent might be WebGPU. Anybody doing cool stuff that? I'd like to see the code.

New comment by esafak in "Your File System Is Already A Graph Database"

esafak — Wed, 08 Apr 2026 13:08:59 +0000

His argument is that the LLM is the query engine. By that logic you can approximate anything since LLMs can.

New comment by esafak in "Git commands I run before reading any code"

esafak — Wed, 08 Apr 2026 12:50:34 +0000

Not a model, but a product: warp.dev

New comment by esafak in "US and Iran agree to provisional ceasefire"

esafak — Wed, 08 Apr 2026 00:47:37 +0000

Look on the bright side; that picture respects terms limits.

New comment by esafak in "A truck driver spent 20 years making a scale model of every building in NYC"

esafak — Tue, 07 Apr 2026 20:20:43 +0000

Your sibling post estimated it pretty well :)

New comment by esafak in "Assessing Claude Mythos Preview's cybersecurity capabilities"

esafak — Tue, 07 Apr 2026 19:53:04 +0000

We need to promote alignment and other ethics benchmarks; we can't change what we don't measure. I don't even know any off the top of my head.