Hacker News Reader

There is a concrete expiry date listed for each reset in the codex desktop app.

O_H_Eon I used Claude Code to get a second opinion on my MRI

> Most people should have learned at a young age that absence of evidence is not evidence of absence.

That might be true, but it is definitely not the world we live in.

O_H_Eon DSpark: Speculative decoding accelerates LLM inference [pdf]

> Local-first session search, analytics, insights, and token use statistics for coding agents, supporting Claude Code, Codex, and more than 20 other agents.

solid piece of software

O_H_Eon Stopping bad guys from using my open source project (feedback wanted)

This is actually one creative idea, kudos. I encourage you to reach out to OP as he was asking for discussion in case he didn't see this comment. This is the only newer out of the box idea I saw in this thread

O_H_Eon Stopping bad guys from using my open source project (feedback wanted)

> There are already so many ways (and reasons) not to do free or open source software. People who find them convincing are using them.

To be honest, I don't think the space between GPL/MIT and commercial closed source is explored enough. I'm aware there's a few examples of things in between, but they are not common knowledge and they don't satisfy everyone. It is not a space that is easy to search online for established wisdom and comparisons in.

O_H_Eon Stopping bad guys from using my open source project (feedback wanted)

> Besides, why would "bad guys" be deterred by a license, anyway?

I imagine because we're talking about a subcategory of "bad guys" who still like to stay within the confines of the law (supposedly at least).

O_H_Eon We're losing our voice to LLMs

So in that case, would someone willing to publish LLM-speak under their name be similarly adopting that "voice".

Does that entail that LLMs are not in fact erasing our societal voices, only making it easier to adopt bland-corporate en-mass?

O_H_Eon We're losing our voice to LLMs

I don't find that current generation LLMs output such short sentences that would start with the same prefix such as "Your voice".

O_H_Eon We're losing our voice to LLMs

There was recently this link talking about AI slop articles on medium

https://rmoff.net/2025/11/25/ai-smells-on-medium/

He doesn't link many examples, but at the end he gives the example of an author pumping out +8 articles in a week across a variety of topics. https://medium.com/@ArkProtocol1

I don't spend time on medium so I don't personally know.

O_H_Eon Migrating the main Zig repository from GitHub to Codeberg

This is a valuable take tbh, I missed it.

Zig team ought to probably write about it in that manner.

O_H_Eon Show HN: Excel Custom Functions in Zig

welp...cool product (xllify) I gotta say. Didn't know custom excel functions (XLL/DLL) were a thing.

O_H_Eon Show HN: Optimizing LiteLLM with Rust – When Expectations Meet Reality

Commit history has 5 commits, 3 of them are 1day ago, and all of them add +1000 lines.

Definitely looks like it.

O_H_Eon How long can it take to become a US citizen?

Speaking of the US, how are TN visas nowadays? Are companies allergic to their paperwork like other visas that are harder to get?

O_H_Eon Ask HN: Who is hiring? (February 2025)

remotely only in france?

O_H_Eon Ask HN: Who wants to be hired? (December 2024)

Location: Ontario, Canada

Remote: Prefered, but open to in-person (depending on offer and location)

Technologies: Julia, Python, Rust, CUDA, C#, C, valgrind, gdb, SciML, low level performance profiling, scientific computing, Linux admin and scripting, Docker, ...

Resume: https://omar-elrefaei.github.io?source=HN (pdf upon request)

Email: omar02295+hn [at] gmail

I'm an experienced Julia developer with an appetite for performance challenges.

I'd be extra interested if the work involves any of the following:

  - Assessing different parallelization or threading strategies
  - Increasing testing coverage to help with maintainability and robustness
  - Cleaning up prototype code to prepare for production or publishing as a package
  - Contribute to open source dependencies

But I'd also be willing to hop on a codebase that's anywhere from an Elixir microservice to C firmware.

I'm a swift learner, if you get me excited about your product/project, I'm confident I can pick up whatever technologies you use quickly.

O_H_Eon Ask HN: Who is hiring? (December 2024)

Would you guys be interested in someone with pure julia experience, agility in digging into new fields, and vested interest in the topic? I'm having hard time picking one of the postings, but I think I could be a good fit.

O_H_Eon Ask HN: Freelancer? Seeking freelancer? (November 2024)

SEEKING WORK | Ontario, Canada | Remote I'm an experienced Julia developer with a enthusiasm for solving performance and optimization challenges.

I can help if you're looking for Julia expertise to help with any of the following:

  - Investigating long compilation or start-up times
  - Profiling and identifying performance bottlenecks
  - Low-level performance optimization for your simulation code
  - Assessing different parallelization or threading strategies
  - Increasing testing coverage to help with maintainability and robustness
  - Cleaning up prototype code to prepare for production or publishing as a package
  - Contribute to a dependency upstream to ease a pain point in your use case

I might also be able to help if you are: - Rewriting or redesigning from R/MATLAB to Julia/Python/Rust. - Having performance problems with your Python, Rust, MATLAB, or C code. - Want to investigate if leveraging GPU compute (CUDA) could be useful for your computations. I've worked across a wide range of the Julia ecosystem, from high-level tasks such as extracting data from non-compliant csv, and writing custom Plots.jl recipes, to low-level work such as extending the mul! matrix multiplication interface with LAPACK, and investigating type-instabilities and inference problems using SnoopCompile and Cthulhu. Think I might be able to help? Shoot me an email and let's find out.

Email: omar02295+hn@gmail.ca

O_H_Eon Ask HN: Freelancer? Seeking freelancer? (October 2024)

SEEKING WORK | Ontario, Canada | Remote

I'm an experienced Julia developer with a enthusiasm for solving performance and optimization challenges.

I can help if you're looking for Julia expertise to help with any of the following:

  - Investigating long compilation or start-up times
  - Profiling and identifying performance bottlenecks
  - Low-level performance optimization for your simulation code
  - Assessing different parallelization or threading strategies
  - Increasing testing coverage to help with maintainability and robustness
  - Cleaning up prototype code to prepare for production or publishing as a package
  - Contribute to a dependency upstream to ease a pain point in your use case

I might also be able to help if you are:

  - Rewriting or redesigning from R/MATLAB to Julia/Python/Rust.
  - Having performance problems with your Python, Rust, MATLAB, or C code.
  - Want to investigate if leveraging GPU compute (CUDA) could be useful for your computations.

I've worked across a wide range of the Julia ecosystem, from high-level tasks such as extracting data from non-compliant csv, and writing custom Plots.jl recipes, to low-level work such as extending the mul! matrix multiplication interface with LAPACK, and investigating type-instabilities and inference problems using SnoopCompile and Cthulhu.

Think I might be able to help? Shoot me an email and let's find out.

Email: omar02295+hn@gmail.ca

O_H_Eon Show HN: AST-grep, a grep/eslint/codemod hybrid tool, powered by tree-sitter.

Yay, nice! finally a simple cli for treesitter.

treesitter tech always seemed cool, but never felt I had access to it on my finger tips. one such tool is https://github.com/Wilfred/difftastic (semantic diff)

will definitely be trying this out.

O_H_Eon IPyflow: Reactive Python Notebooks in Jupyter(Lab)

Yeah, Pluto rules!

After getting used to it with Julia I found it really jarring to go back to plain Jupyter (when I need python) where I have to keep re-executing the cells.

This is going to make that much less painful.

O_H_Eon Zig and Rust

Yeah that made me wow too!

He's gonna have a wild career.

O_H_Eon Framework announces AMD, new Intel gen, 16“ laptop and more

Yeah website is down. He never mentioned a trackpoint unfortunately.

But I guess theoretically someone could design one (╯°□°）╯ ┻┻ they are releasing hardware and software interface docs on github.

O_H_Eon Framework announces AMD, new Intel gen, 16“ laptop and more

salvaging old laptop battery as a power bank made me actually lol in real life.

O_H_Eon Framework announces AMD, new Intel gen, 16“ laptop and more

Wow RPI-based QMK-enabled input modules. That is something else.

O_H_Eon On the left is GPT-3.5. On the right is GPT-4

I'm saddened to inform you that no scientific research being done these days holds up to these standards.

O_H_Eon Ask HN: Programs that saved you 100 hours? (2022 edition)

Actually just today I asked if for the density of silicon nitrite and to cite a source. It cave a citation that seemed correct (reference book on materials) but with completely made up authors.

O_H_Eon Ask HN: Programs that saved you 100 hours? (2022 edition)

Yup, works well enough in ST4

O_H_Eon Goodbye to the C++ Implementation of Zig

I'd imagine distributing tarballs is also an important use case.

O_H_Eon Tell HN: Giving ChatGPT access to a real terminal

> legal nonsense that allows Elon to call himself a co-founder of Tesla[0]

Wow, I had no idea.

O_H_Eon Memory Safe Languages in Android 13

just fyi, you can still write all the Java/Kotlin you want to stay relevant.

this post have nothing to do with writing apps for android. They are talking about developing the android system itself. They are writing new system code in Rust instead of C++. Userspace code is still enoraged to be written in Kotlin.