rrvsh
rrvsh on Hacker News
You are being downvoted but you are absolutely correct. I have never seen someone assert Rust is a 100% safe. Its touted purely as a _memory_ safe language. To claim otherwise is absurd and very telling of the writer's biases
Nowadays ELI5 makes it come up with an often nonsensical analogy about cars or kitchens or some such. My coworkers love it to death for some reason
Location: Singapore
Remote: Yes
Willing to relocate: Yes, preferably to San Francisco, but also open to New York and London. I am a Singapore citizen and eligible for the H-1B1 route for US roles.
Technologies: Python, Rust, TypeScript, PostgreSQL, Nix/NixOS, Linux, AWS, GCP, Docker, CI/CD, observability, backend and platform engineering
Résumé/CV: https://rrv.sh/assets/cv.pdf
Email: [email protected]
I'm a backend/platform engineer. I was previously a founding engineer at a B2B startup, where I worked across the backend, frontend, infrastructure, deployment, integrations, and internal workflows. I'm currently doing production reliability and cloud migration work at Autodesk.
I like working on messy problems where I can figure out what needs to exist, build it, and then deal with the boring work required to keep it running. Outside work I spend a lot of time building things with Rust and Nix, mostly around developer tooling, personal infrastructure, and AI agents. Some of that is at https://github.com/rrvsh, and I write about whatever I'm working on at https://rrv.sh.
I'm mostly looking for backend, platform, infrastructure, developer tooling, or high-ownership startup roles.
Surely you realize how rare the ability to do this is
Yes, I really don't like Astra - 5.6 models seemed to perform at literally the same level with less opaque prose; I guess Astra is great if you're working on insanely hard mathematical problems (or are fooled by its masked sycophancy) but for coding 5.6 seems to have better taste. I hope that they course correct or at least offer models that do better for coding, or even better that this oligopoly ends
I found that 6 Sol is dogshit; have you tried o5.5 vs. 5.6 sol? curious to hear if your experience is still the same in that regard
It doesn't - try it first
Its really not tiny; you can't compare Claude to GPT, they have honestly diverged enough that as the other reply said, 256k GPT is about equal to 1M Claude. The compaction is slightly annoying, and you can turn it up to 1M as you said if you truly need everything in context, but otherwise it's perfectly serviceable
Hard agree
I had to switch back to 5.6 Sol after trialling 6 Sol for like 3 days - I was getting insanely annoyed at how misaligned it is. Will try 6.1 but not very high hopes
I feel like everyone is focusing on the AI bottleneck hook rather than what the actual article is about. I enjoyed it a lot; my lowkey favourite thing is optimising CI beyond reasonable levels
Hire this guy he's cracked
Location: Singapore
Remote: Yes
Willing to relocate: Yes, preferably to San Francisco, but also open to New York and London. I am a Singapore citizen and eligible for the H-1B1 route for US roles.
Technologies: Python, Rust, TypeScript, PostgreSQL, Nix/NixOS, Linux, AWS, GCP, Docker, CI/CD, observability, backend and platform engineering
Résumé/CV: https://rrv.sh/assets/cv.pdf
Email: [email protected]
I'm a backend/platform engineer. I was previously a founding engineer at a B2B startup, where I worked across the backend, frontend, infrastructure, deployment, integrations, and internal workflows. I'm currently doing production reliability and cloud migration work at Autodesk.
I like working on messy problems where I can figure out what needs to exist, build it, and then deal with the boring work required to keep it running. Outside work I spend a lot of time building things with Rust and Nix, mostly around developer tooling, personal infrastructure, and AI agents. Some of that is at https://github.com/rrvsh, and I write about whatever I'm working on at https://rrv.sh.
I'm mostly looking for backend, platform, infrastructure, developer tooling, or high-ownership startup roles.
This is a bit silly, and has the same tinge as lolcow posts about asking "Chat" when "my son in 3rd grade is going to graduate college". There are a million fasting apps that show based on how long its been since your last meal what is happening in your body, say ketosis or autophagy. Off the top of my head for Android there's Easy Fast, Zero, MyFastingApp (OSS); it's public information that doesn't require inference to derive
Haven't they been migrating to Azure for a few years? How is it still only 58% done... Microslop needs to lay off the focus on AI features and get it done
What's so bad about this readme? It's concise and representative of the project - no LLM-isms other than the emojis at the top, and even those look more like a human put them in than an LLM
please enumerate the civil wars over the decades Nix has been around. It is insane to discourage someone from using a good piece of battle-tested software because you personally have opinions about their community.
What do you mean? Codex paid plan treats me quite well. I'm not going to complain that free users get more shit, especially when it doesn't affect me in any way.
Location: Singapore (Visa sponsorship not an issue)
Remote: Doesn't matter
Willing to relocate: Yes (San Francisco, New York, London)
Technologies: Nix/NixOS, Linux, Rust, Python, Bash, Lua, GCP, AWS, GitHub Actions, Terraform/OpenTofu, PostgreSQL
CV: https://rrv.sh/assets/cv.pdf
Website: https://rrv.sh
GitHub: https://github.com/rrvsh
Linkedin: https://linkedin.com/in/rrvsh
Email: [email protected]
Software engineer with a background in infrastructure, backend systems, and developer tooling. I enjoy building systems that are easy to operate, easy to change, and automate away repetitive work. Strongest areas are Nix/Linux, CI/CD, infrastructure, and Rust/Python.
Most recently I was a Founding Engineer at Alpha Story. I built features across the stack, led an authentication migration from Supabase Auth to an in-house system with no downtime, automated infrastructure with OpenTofu and GitHub Actions, cut CI times by more than half, and stepped in to lead engineering after the CTO left.
Before that I worked in production operations supporting more than 100 AWS-hosted government websites, where I wrote internal tools to automate operational workflows.
I contribute to nixpkgs, Home Manager, Hyprland, NVF, pullomatic, and other open-source projects. Personal projects are mostly around Nix, Linux, Rust, and developer tooling.
I'm looking for backend, platform, infrastructure, or developer tooling roles.
It's insane that after four years of these tools being ubiquitous people still defer to them for their own critical thinking
Why post AI-generated answers? Anyone reading your comment could do the same.
I feel like thats impossible - in the end the reason we have those useless roles is to have someone to shift the blame to
I unequivocally disagree as I once conducted a training session teaching them what Gmail labels were - was very fruitful. ;)
Didn't this slightly contradict itself? The conclusion I drew based on the problem statement was that the best way was to have a model router handle the root prompt and stick to a model for the rest of the session - engineers still have the discretion to handoff to a smaller or larger model when needs arise.
Nice! If you could plug in a controller and use a TUI I would 100% use this - Mixxx is great but sometimes I don't want to install another desktop app
If you want to optimise and have the money to spare, I think it's fine. There are going to be people in the world that are better than you at most given tasks, and if working on those tasks is not fun for you and delegation is not a net negative, why not? It's why personal trainers exist (though I don't like them for separate reasons)
Side note, why not just deploy onto a VPS and save the money from using hosted services, as well as protecting yourself from when they inevitably go down? The CBA clearly shows that while active development is going on its trivial to have hosted service subscriptions because the AI sub is way more, but what if that sub disappears? You're left paying (relatively) exorbitant amounts compared to $1-10 per month per VPS for life, or less if you have the privilege and wherewithal to self host. It's also dead easy with AI. Maybe this just reinforces how much a personal runtime solution is needed, as stated in YC's recent RFS.
Yeah, intuitively the request floor instead of up-down paradigm always seemed inefficient. However, if the issue is that it forces a single elevator and doesn't allow reoptimisation, I wonder if there's a middle ground where instead of up down there's a button panel for the floors, but you still just take whatever elevator's going up? It technically wouldn't require double work but kind of feels like it does psychologically - anyone seen prior work on this?
I went down a five hour rabbit hole learning about jwz as a result of this post. What an interesting guy - if I knew about the DNA lounge a week earlier, I would have visited when I was in the city
That's not true. AI was widely used to refer to the decision systems and state machines that produced NPC behaviour in video games, and I'm sure many other things than just science fiction