HN user

tough

3,152 karma

Software Engineer

tough [at] fastmail [dot] com

Posts30
Comments1,966
View on HN
ashallowlake.com 14d ago

Show HN: A shallow lake's report on itself

tough
3pts0
steuwer.info 11mo ago

Descend: A Safe GPU Systems ProgrammingLanguage [pdf]

tough
4pts0
github.com 12mo ago

Simulate Harsh User Review for Claude code

tough
5pts1
arxiv.org 1y ago

AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges

tough
3pts0
arxiv.org 1y ago

SEM-CTRL: Semantically Controlled Decoding

tough
3pts0
arxiv.org 1y ago

IterGen: Iterative Semantic-Aware Structured LLM Generation with Backtracking

tough
1pts0
arxiv.org 1y ago

ROCODE: Integrating Backtracking Mechanism and Program Analysis in LLMs for Code

tough
1pts0
arxiv.org 1y ago

SRLCG: Self-Rectified Large-Scale Code Generation, CoT, Dynamic Backtracking

tough
1pts0
arxiv.org 1y ago

CRANE: Reasoning with Constrained LLM Generation

tough
1pts0
arxiv.org 1y ago

Type-constrained code generation with language models

tough
257pts127
buntralino.github.io 1y ago

Buntralino Make better, faster cross-platform desktop apps with Bun

tough
1pts0
ssrc-static.s3.us-east-1.amazonaws.com 1y ago

Beyond Public Access in LLM Pre-Training Data: Book Content in OpenAI's Models [pdf]

tough
1pts0
headers.club 1y ago

Got a header that's too hot to handle? Share it

tough
1pts0
en.wikipedia.org 1y ago

Osteo-Odonto-Keratoprosthesis (OOKP)

tough
1pts1
news.ycombinator.com 3y ago

Tell HN: If you search for Barbie's Director on Google you get Barbie Google

tough
4pts2
news.ycombinator.com 3y ago

Ask HN: Archive Bot on HN Submissions

tough
2pts1
medusajs.com 3y ago

Medusa Recap (June)

tough
1pts1
github.com 3y ago

GPTyped: A small, but powerful NPM package that calls GPT in a type-safe way

tough
4pts2
github.com 3y ago

Show HN: Twitter Zero, Chrome extension to hide all numerical stats from Twitter

tough
2pts0
github.com 3y ago

Kinto.sh a nice GUI to get OS X like commands in windows and Linux

tough
3pts1
openai.com 3y ago

ChatGPT Plugins Waitlist

tough
3pts1
news.ycombinator.com 3y ago

Is Twitter.com Down?

tough
1pts2
news.ycombinator.com 3y ago

Twitter Is Down

tough
14pts4
news.ycombinator.com 3y ago

Ask HN: Gitcafe a Git federated hosting service

tough
1pts2
caravaggio.ramielcreations.com 3y ago

Caravaggio, A fast image manipulation service

tough
2pts0
www.genome.gov 3y ago

Human Genome Project

tough
2pts0
cal.com 3y ago

Cal.com v2. Formerly Calendso. COSS Scheduling Infrastructure

tough
4pts0
news.ycombinator.com 4y ago

Show HN: A little ETL framework I built in TypeScript

tough
5pts3
www.michaelbromley.co.uk 4y ago

But what about the bus factor?

tough
65pts85
news.ycombinator.com 4y ago

HN 15 Years ago

tough
558pts148

ratatui was still an excellent choice over a year and a half ago.

Nobody but a JS dev would choose react-on-terminal shenanigans over it

I could see CC getting a ratatui port eventually actually

Decoy Font 6 days ago

If you squint vs not squint it becomes way easier some how to differentiate i wonder why

Grok 4.5 14 days ago

well i do hope that wasnt grok 4.5

Don't trust, but verify. etc

To all the haters to not knowing what "pi" means in this context. Is mario zechner's ai agent / coding harness [1][2], similar in spirit albeit much less complex to claude code or openai's codex and others.

I wouldn't call it third tier. If anything alongside opencode, and codex, its one of the "first tier" and the only non-VC-backed (at least before mario joined earendil (idk about earendil raising or not)

anyways. Pi is good. I dont need a telegram client for it, but this precisely show why pi is great, because its really easy to extend pi building plugins or modifying the source (yay, open source)

peace

1. https://github.com/earendil-works/pi

2. https://pi.dev/

Grok 4.5 14 days ago

Ollama is just a local app wrapper/cloud service serving third party apis and models idk why it made it into this list tbh

GPT‑Live 14 days ago

the cybertruck seems designed for this KITT fantasy of yours

I think it's more about the bait and switch marketing optics of offering a "core" part of your codebase as "FOSS" to get the extra brownie points by doing so, while it's at the same time littered with caveat emptors, and imports from "enterprise licensed" code for the features that business care about.

Also all the "commercial open source" or COSS seems to lean into that, its a valid strategy to monetize open source, but should be done transparently and being very upfront about its bait-y nature to businesses or enterprises imho

so does anything change or will change for better-auth FOSS consumers? Or not really and that mostly depends on what Vercel imposes as new direction going forward? I guess vercel is a good home for foss projects, at worst they'll make first class adoption on their ecosystem without actively hampering others etc

There's been a lot of recent scandals going public against the social democratic party ruling on spain now (PSOE) and its previous dirigents. See Zapatero case. leaked by US agencies recently once Spain put some kind of friction to the Rota south spain bases getting involved on anything vs Iran.

The president P. Sanchez, has been clearly antagonizing Trump in these and other intl issues (even if only visible in spain, as he is not that relevant internationally, etc)

But anyways, this seems like deepstate fighting vs current US admin and current Spain admin, one can infer "Palantir" is basically a gag order away from giving the US govt anything it wants, so as an antagonist. to its current admin, it seems smart to avoid having them as critical providers.

why choose china? Makes no sense, but probably the only other big bro Spain can rely on if the US isn't it anymore

Interesting and related , saw on X earlier that CC was now adding some tracking etc, here a post related to how it does it and what its targeting (chinese users) https://thereallo.dev/blog/claude-code-prompt-steganography

one could also see the fable-5 getting pulled off, US govt-ant talks, etc as part of all this globally i think

which is may way to say maybe Anthropic knows this isnt true, but they still will say otherwise publicly to make this admin understand whatever they need regardless on potential security issues etc, idk im extrapolating toomuch probably