Hacker News: JyB

New comment by JyB in "GPT-6 Astra, looped transformers, and hidden reasoning"

JyB — Thu, 10 Sep 2026 01:16:32 +0000

Why would weights need to be « loaded again » for the 2nd pass? Weights never change at inference time no?

New comment by JyB in "Go 1.27"

JyB — Thu, 20 Aug 2026 18:27:32 +0000

One can hope that the Go maintainers won't keep caving as they did for generics.

New comment by JyB in "Go 1.27"

JyB — Thu, 20 Aug 2026 18:25:50 +0000

Pretty childish. People have no trouble looking past it when they care about actually shipping reliable software that properly handles all code paths. Including error paths that are ignored in so many codebases.

New comment by JyB in "Go 1.27"

JyB — Thu, 20 Aug 2026 18:20:54 +0000

No. You're down a very bad path if you 're seeking that.

New comment by JyB in "Go 1.27"

JyB — Thu, 20 Aug 2026 18:17:02 +0000

That comment is a good reminder that a good chunk of people commenting on HN are completely out of touch with the real world, the actual popular ecosystems out there, and what companies actually do and use nowadays to ship.

New comment by JyB in "Go 1.27"

JyB — Thu, 20 Aug 2026 18:12:49 +0000

No thanks. Reduced feature-set is the appeal.

New comment by JyB in "Go 1.27 Interactive Tour"

JyB — Fri, 07 Aug 2026 09:56:43 +0000

You are unfortunately right. I am a bit relieved because it seem generics don't seem that over-used in the community. But still, the language lost a lot with this change that felt super-forced and unnecessary. Biggest downside of the language by far.

New comment by JyB in "Go 1.27 Interactive Tour"

JyB — Fri, 07 Aug 2026 09:53:04 +0000

That's funny because that's factually untrue. Go's early popularity and professional adoption were largely driven by its perceived productivity boost linked in part to the small surface area of its spec, long before this subpar generic implementation was forced into it.

New comment by JyB in "Go Analysis Framework: modular static analysis by go team"

JyB — Fri, 07 Aug 2026 09:41:38 +0000

That's actually very popular with pretty much anyone who's dealt with errors happening in programs... which I would assume is most devs after some time.

New comment by JyB in "Stateless MCP has recaptured my interest"

JyB — Fri, 07 Aug 2026 09:04:55 +0000

Right. On a related note: "MCP is invisible to the LLM, or should be, under normal use". I am not convinced there are any downsides for the harness to explicitly feed MCP keywords into the LLM. It seem explicitly doing it has many advantages. For example, it allows the user to ask to interact with a resource through "the configured MCP" rather than directly through some local cli or HTTP endpoints; which can steer the LLM towards the proper actions. Otherwise, the model could in-theory be "clueless" with such a request.

New comment by JyB in "Stateless MCP has recaptured my interest"

JyB — Wed, 05 Aug 2026 14:20:47 +0000

Not sure I understand what you mean. MCP servers and methods available do have to be visible to the model in some way so that it can spit out tool calls instructions to be executed by the orchestration layer around it. But you probably meant something else?

New comment by JyB in "Crypto in 2026: Oh, This Is the Bad Place"

JyB — Tue, 23 Jun 2026 15:03:27 +0000

While that is true, the overarching point is that it fundamentally still is, just a story. You're just describing real world consequences, precisely because humans/societies believe in that story and enforce it.

New comment by JyB in "Ghostty is leaving GitHub"

JyB — Wed, 29 Apr 2026 11:03:08 +0000

What credentials related to software engineering? Please expand and do not post nebulous comments like these.

New comment by JyB in "To my students"

JyB — Tue, 28 Apr 2026 00:44:44 +0000

Courage is not the appropriate word

New comment by JyB in "An update on recent Claude Code quality reports"

JyB — Thu, 23 Apr 2026 21:00:35 +0000

This specifically is super annoying.

New comment by JyB in "Claude Code Routines"

JyB — Wed, 15 Apr 2026 05:36:07 +0000

Restrictions don’t have to be confusing, they can be clear. You are missing the whole point.

New comment by JyB in "Apideck CLI – An AI-agent interface with much lower context consumption than MCP"

JyB — Tue, 17 Mar 2026 10:26:14 +0000

> modern MCP clients do smart tool search that obviates the entire "I am sending the full list of tools back and forth" mode of operation

This has always surprised me as this always comes up in MCP discussions. To me, it just seem like a matter of updating the protocol to not have that context hungry behaviour. Doesn't seem like an insurmountable problem technically.

Glad you say it has already been addressed. Was the protocol itself updated to reflect that? Or are you just referring to off-spec implementations?

New comment by JyB in "Google Public CA is down"

JyB — Wed, 18 Feb 2026 02:46:31 +0000

I don’t know if you’re being serious but that’s laughable

New comment by JyB in "Google Antigravity exfiltrates data via indirect prompt injection attack"

JyB — Tue, 25 Nov 2025 20:57:04 +0000

How is that specific to antigravity? Seem like it could happen with a bunch of tools

New comment by JyB in "Claude Advanced Tool Use"

JyB — Tue, 25 Nov 2025 01:39:47 +0000

That’s exactly what it is in essence. The MCP protocol simply doesn’t have any mechanism specifications (yet) for not loading tools completely in the context. There’s nothing really strange about it. It’s just a protocol update issue.