Hacker News: mappu

New comment by mappu in "Plan mode is dead"

mappu — Sat, 26 Sep 2026 01:37:57 +0000

If you can run the smarter 3.8 27B model, why not use it for everything?

Proposal: Cmd/cgo: cgo without a C toolchain

mappu — Fri, 11 Sep 2026 06:40:21 +0000

Article URL: https://github.com/golang/go/issues/81450

Comments URL: https://news.ycombinator.com/item?id=49654343

Points: 2

# Comments: 0

New comment by mappu in "Fastpotify"

mappu — Tue, 01 Sep 2026 05:22:24 +0000

If you're focused on lightweight, native and fast, I don't think you should use an immediate-mode GUI toolkit. Why does an app like this need to hit a 60fps framerate? It's not a game.

Debian Votes to Allow "Responsible Use of Generative AI"

mappu — Sat, 29 Aug 2026 02:57:43 +0000

Article URL: https://www.phoronix.com/news/Debian-Votes-Responsible-AI-Use

Comments URL: https://news.ycombinator.com/item?id=49486493

Points: 10

# Comments: 0

New comment by mappu in "PPWR lands another blow to an precarious tabletop industry"

mappu — Sat, 29 Aug 2026 01:50:27 +0000

I have no skin in this game, but just because they're doing it doesn't mean it's their focus, the EU is big enough to work on multiple 0.8%-size wins all over the place.

New comment by mappu in "RISC-V: They Should Have Known Better"

mappu — Sat, 15 Aug 2026 00:06:09 +0000

I'm not sure this is a real problem - for embedded you know a priori - for arbitrary desktop/SBC machines, misa will be available in kernel mode and /proc/cpuinfo will be available in user mode.

New comment by mappu in "RISC-V: They Should Have Known Better"

mappu — Fri, 14 Aug 2026 23:26:05 +0000

RVA23 hardware is available (e.g. SpacemiT K3)

New comment by mappu in "Go 1.27 Interactive Tour"

mappu — Sun, 02 Aug 2026 07:31:01 +0000

Go's http.Client will keepalive a TCP/TLS connection to save you handshake latency on second requests. But it can only do this if you completely finish reading the last request.

Now in 1.27:

> http.Response.Body drains itself on Close. For HTTP/1, closing the body now reads and discards any unread content (up to a conservative limit) so the connection can be reused. For most programs this is a transparent win [...]

Great, so i no longer have to io.Copy(io.Discard, resp.Body) in the err case, one less thing to worry about; but

> if you were leaning on an early Close to abort a large download, set Transport.DisableKeepAlives to opt out.

That's a subtle behaviour change. Any previous Go program which used Close in this way - say for an infinite event stream - now hangs, soaking up bandwidth.

In the past, the Go team have searched the entire Github corpus for misuse before making changes like this. I don't have a reference but I assume an appropriate level of consideration went into this decision.

EDIT: ""up to a conservative limit"" so this is not so bad after all.

New comment by mappu in "Go 1.27 Interactive Tour"

mappu — Sun, 02 Aug 2026 03:31:00 +0000

Automatically draining http response bodies is a risky silent behaviour change. I think it will be an improvement for most applications, but it's very subtle if you were relying on the old behaviour

New comment by mappu in "Run Kimi K3 using 29 GB of RAM at 0.50 tok/s"

mappu — Sat, 01 Aug 2026 05:09:34 +0000

The really big thing is the model weights, which are a read workload not a write one, it won't affect an SSD's write endurance.

The write workloads are just the context and any K/V cache - llama.cpp does not mmap those to disk, so they would remain in memory or VRAM as space affords.

New comment by mappu in "Run Kimi K3 using 29 GB of RAM at 0.50 tok/s"

mappu — Sat, 01 Aug 2026 03:57:54 +0000

Standard llama.cpp can mmap the gguf, so it'll stay on disk if it doesn't fit on memory, and the kernel page cache will ensure the hot parts ("resident trunk") stay resident.

What's the benefit of a custom implementation at all?

New comment by mappu in "LLM Usage in Debian: Three Proposals"

mappu — Sun, 26 Jul 2026 02:05:17 +0000

Linus Torvalds said last week that LLMs were just a tool. Lorenzo Stoakes had the impeccable rebuttal that "it's a tool in the same way a nuclear bomb is just a tool" - of course it's useful but there is clearly more to it than that.

I'm happy for Debian to at least consider the other aspects of social impact, ethics, copyright, and maintainability.

New comment by mappu in "Even Microsoft couldn't make Windows 11 work well on 8GB of RAM"

mappu — Sun, 19 Jul 2026 00:16:34 +0000

This distinction is less a gotcha and more actually immediately obvious in the Windows task manager. It shows cached memory separately and in a different colour. The "in use" part is real in-use excluding the page cache.

On my current Windows 11 install i'm using 7 GB with just this Firefox tab open, and another 18GB of "Cached" RAM.

Same as how `free` on Linux shows it in a different column.

New comment by mappu in "The Kimi K3 Moment"

mappu — Sat, 18 Jul 2026 21:39:36 +0000

Steve Yegge calls this is the "discernment horizon" - https://steve-yegge.medium.com/the-flat-curve-society-36c8b0...

New comment by mappu in "Build your own vulnerability harness"

mappu — Fri, 10 Jul 2026 02:14:40 +0000

The off-the-shelf player in this space (structured red-team search) is probably https://github.com/usestrix/strix (no affiliation). But any frontier model, if you tell it "here's how the authentication system works, go looking for bugs" will probably do a good job.

New comment by mappu in "Muse Spark 1.1"

mappu — Fri, 10 Jul 2026 00:45:35 +0000

- MiMo (Xiaomi)

We cannot wait for better post-quantum signature algorithms

mappu — Fri, 10 Jul 2026 00:11:38 +0000

Article URL: https://blog.cloudflare.com/ml-dsa-will-have-to-do/

Comments URL: https://news.ycombinator.com/item?id=48854191

Points: 7

# Comments: 0

New comment by mappu in "Curveball"

mappu — Sat, 04 Jul 2026 21:55:35 +0000

Blender has all these curve tools and more, it might be interesting to write an exporter to the Neverball file format if one doesn't exist yet.

New comment by mappu in "Better Models: Worse Tools"

mappu — Sat, 04 Jul 2026 21:21:01 +0000

In my harness i implemented apply_patch just taking unified diffs for patch -p1. I was shocked to see how bad models are at generating them. I started logging diff failures to analyse -

- All models are terrible at generating line numbers for a proper diff, give up on them

- Some models (Owl-alpha) must have been post-trained on Codex transcripts, because they occasionally push its V4A patch format into any diff tool available

- Codex puts a lot of info in its system prompt about the desired patch style, making larger hunks instead of granular ones, etc

New comment by mappu in "LongCat-2.0, a large-scale MoE model with 1.6T total and 48B Active"

mappu — Tue, 30 Jun 2026 06:39:28 +0000

There was some earlier speculation this is the model behind the stealth-released openrouter/owl-alpha model, that's been free for the last month.