HNHacker News
TopNewBestAskShowJobs

stevefan1999

2,192 karma · joined July 1, 2017

submissionscomments
stevefan1999··on US Government directive to suspend access to Fable 5 and Mythos 5
So are you going to restrict access to Fable by another KYC scheme but this time prove that you are US citizen first amirite
stevefan1999··on Show HN: FablePool – pool money behind a prompt, and Fable builds it in public
sounds like FableBugBounty
stevefan1999··on Anthropic apologizes for invisible Claude Fable guardrails
Then reset the quotas as an atonement ;p

Seriously though, Fable was not that great facing a greenfield subject. It is excellent at oneshotting some math problems, but if you want it to do some cutting edge tech stuff, say like piecing together a new Crossplane XRD, by reading existing Helm chart and with application source code available. I still have to get a few pass for Fable to get it done right, and at this point I may consider making a skill for it. I even gave it the source code of the Crossplane itself and tell it to be careful about CRDs and data flow, but it is still pretty silly. Adaptiveness for Fable is still not great, and I think it is a well known problem for Anthropic, albeit all LLMs do suffer a lot from subjects they don't know and will hallucinate stuff very frequently.

stevefan1999··on Steve Jobs in Exile is a fine profile of Jobs' years at NeXT
Fun fact: the history of the (in)famous prefix NS in OSX -> macOS is a direct nod to the acronym of NextStep, hence NS. As NeXTSTEP was incorporated back into Apple, it was now re-appropriated as New System. Same thing goes to Cocoa and Cocoa Touch, as the legacy NeXTSTEP still lives on
stevefan1999··on Claude Desktop spawns 1.8 GB Hyper-V VM on every launch, even for chat-only use
> Ever hear of a large K8s migration going well

...well, AI infra stacks are more or less k8s/cloud native, especially on the inference side. Nvidia GPU operator plus KubeFlow makes deploying models easy, and I did that manually as well using just the ollama, llama.cpp containers and especially vllm operator. I'm not sure about what's the measuring notation of "large" is, but OpenAI had a blog post about how to manage 7500+ nodes Kubernetes (https://openai.com/index/scaling-kubernetes-to-7500-nodes/). That's 5 years ago and I speculate it will only be even more.

> completely re-implemented under the hood for anything halfway serious

...an example please? I'm building a platform to deploy hundreds of open source apps on Kubernetes, you know, the boring thing you can do with a VPS and maybe use Docker Compose to start -- with an ergonomic twist. I relied on so many features of what Kubernetes and its ecosystem provided, especially with persistent volume, volume snapshot, CRDs and cron job, and all of it is just composed from open source and cloud native software...I'm not sure if that sounds "half way serious" to you.

I know those boring monolithic apps can be done with a VPS, as I exactly came from that background, and I know what the shitty points of having just a VPS are. You don't have a clean control plane, you don't have HA, you don't have distributed storage, and you have to be super aware of the apps that you are deploying with Docker Compose.

What you would think "completely re-implemented under the hood for anything halfway serious", perhaps it means that you don't want to get in the fuzz to manage the complexity and prefer to trade for a far simpler but more primitive solution, and that's totally fine.

Pro tip: Codex seems to love generating great Helm charts so much for some reason unknown. I tried GPT-5.4 high in codex and it easily beats Opus 4.7 max on my own internal helm chart generation benchmark evaluation, which measures HA, deployment and app-specific probes. I've given source code to both Codex and Opus and research about the app config structure, and Codex did so well to just generate secret key-value and convert it to JSON and environ, while Opus insisted on using a dynamic generator at runtime.

> Nomad was cool though

Ahem. I used to deploy Nomad too, but I found it too underwhelming, especially with the networking side of thing.

Consul is perhaps one of the most hated thing in my entire career that I would call it bullshit. I know that later Nomad has its own Raft mode, but the Consul brain rot to me is already very sickening such that I don't want to touch Nomad no more. Vault is fine though, but I prefer OpenBao now.

> Dart is also complete shit: Speed of javascript with verbosity of Java

I don't get where the "speed of javascript" is from. Dart/Flutter is JIT on dev mode, and AOT compiled on production, is that you didn't get the compiler options right?

stevefan1999··on Claude Desktop spawns 1.8 GB Hyper-V VM on every launch, even for chat-only use
But Kubernetes, being a former Google project whilst donated to CNCF, is still one of the mainstream container orchestrator if not the biggest one (I believe we need to go over to supercomputers for that). Nomad is kind of dead and Apache Mesos is basically dead. Leaving Kubernetes as a natural moat, plus Google employees are still seemingly actively maintaining it.

And also Go. While I'm not a Go guy (speaking as a C# and Rust guy, I did wrote a good amount of Go before) it has a huge dominance in Cloud-Native application. For one, Zitadel, an alternative to Keycloak, is written in Go and only takes a fraction of what Keycloak needs.

Flutter/Dart is catching up, but the ecosystem is still relatively weak.

stevefan1999··on Ask HN: What was your "oh shit" moment with GenAI?
vibe coding a C compiler and doing a parser rewrite, all of a sudden it dropped all its code that was refactored to use a token stream, reverted back to using string symbols (that is a huge refactor since it involved parser combinator and turning it from string to logos and eat whitespaces is no easy feat). I almost shitted my pants and tell Opus to stop and fuck off, ask him to recover the code from the context window, but luckily it was git stahed. It scared the shit out of me.

During the process I learned a lot more about giving a chance for LLM to do commit after its work because it will sometimes just give up. Also learned a hard lesson on C preprocessor and why tokenization is necessary.

stevefan1999··on Moving beyond fork() + exec()
If fork and exec can exhibit persistent and algebraic behavior (beyond its CoW nature) that would not only be more useful but more interesting to use, for example using it for doing lazy evaluation
stevefan1999··on Python JIT project was asked to pause development
Honorable mention: https://github.com/tonybaloney/Pyjion
stevefan1999··on .NET (OK, C#) finally gets union types
Well, what's the difference between a (benefactory or not) dictator or an oligarch?
stevefan1999··on Gleam v1.17.0
Is there any AOT solution for running BEAM?
stevefan1999··on Multicore suppport for DOS is real – partly
Well, even so, multicore nowadays is supposedly just running the kernel over different CPUs by treating each CPU as its independent computer, and use a message box to deliver event messages, that's how x86 works with APIC and SIPI
stevefan1999··on Benchmarking SurrealDB 3.x vs. Postgres, Mongo, Neo4j and Redis (With Fsync)
Well, I'm still in a very early phase, but I'm indeed combining both Restate and SurrealDB together for a project that I'm building, where I persist the temporary state on Restate and permanent state on SurrealDB, and since both uses JSON as its lingua franca, it is pretty easy to serailize data between Restate and SurrealDB, very much so better than using MongoDB with BSON as many people would have naturally thought of what is supposed to be a better replacement than SurrealDB.

Oh, that's the reason the SurrealDB operator was here in the first place because I need the full K8S lifecycle to maintain the database state such as backing up, that is not really doable with Helm.

stevefan1999··on Doom Runs on Everything Except Neo Geo [video]
https://github.com/lantus/NGRayEx
stevefan1999··on Benchmarking SurrealDB 3.x vs. Postgres, Mongo, Neo4j and Redis (With Fsync)
Projects that's using already existing that is using Postgres already should keep it in Postgres.

It is worth a try for startups if you won't mind. Try to vibe code around it and give the data model a new look. I have a prototype project that combines both tree-sitter AST and converted it to JSON, then since SurrealDB accepts JSON as native input I now get free graph lookup on the control flow and easily did ancestry analysis and finding what functions potentially calls to this segement. All of it is in SurrealDB nested graph queries and the performance is alright, but is abysmal in Postgres JSONB since JSONB does not linearize the JSON data structure.

ps: I'm building a K8S operator for deploying SurrealDB with TiKV operator integration too.

stevefan1999··on Blue Origin's New Glenn blows up during static fire test
Let me quote a portion of lyrics from Still Alive:

> But there's no sense crying over every mistake

> You just keep on trying 'til you run out of cake

stevefan1999··on GitHub bans security researcher who posted zero-day Windows exploits
I mean he should sell those 0days to exploit.im market for a good money instead of going for "whitehat" if you want maximum damage
stevefan1999··on Anthropic is expanding to Colossus2. Will use GB200
Well, knowledge distillation requires a teacher model and a student model and the student model attempts to learn and extract and (preferrably) compress the information of the teacher model, so it is possible for model collapse due to high SNR in between [1].

What I suggested is to steal the (possibly intermediate) weight in between by sniffing the network communication bus, which means MITM for getting the exact values. Or unless it turns out OpenAI or Anthropic leveraged homomorphic encryption, or I'm not certain how is Anthropic would safely allow Mythos to run on AWS without their control.

[1]: https://en.wikipedia.org/wiki/Knowledge_distillation

stevefan1999··on Anthropic is expanding to Colossus2. Will use GB200
Aren't Anthropic afraid of Elon siphoning the model weights out from the network buses?
stevefan1999··on An OpenAI model has disproved a central conjecture in discrete geometry
That's proof by contradiction: https://en.wikipedia.org/wiki/Reductio_ad_absurdum
stevefan1999··on When can the C++ compiler devirtualize a call?
No. You don't have to add Box everywhere. You do need dynamic storage that is not guaranteed to be known in compile time. Box is just one of them, you can also use "thin box" or "stacked box" or a good ol' "&mut dyn T" would also do great. It basically means a construct of this object with this trait and solved on dynamic dispatch. In C++ that is virtual/pure virtual function.
stevefan1999··on When can the C++ compiler devirtualize a call?
I also wonder what about devirtualization of dyn traits in Rust. Sure, impl traits like foo<T: AsRef>(bar: T) or foo(bar: impl AsRef) is for sure devirtualized, but foo(bar: &dyn AsRef) or most likely foo(bar: Box<dyn AsRef>), far as I remember, isn't always devirtualized. Sometimes it does, sometimes it doesn't. I wonder if it is MIR that did it, or just completely handed off to LLVM for detection.
stevefan1999··on How to buy cheap Claude tokens in China
Well, Dario Amodei used to work in Baidu's SVAIL lab, and he certainly noticed the shady side of Chinese business practices and how he already has a prejudice that the Chinese are unfaithful. As a Chinese myself I don't really want to blame him because I know how's that working out first hand too.
stevefan1999··on Steve Jobs in Exile – New book on his years at NeXT Computer
Yeah, Trump went bankrupt 3 times, and he's still here
stevefan1999··on A new book on Steve Jobs at NeXT
So that's Intel few years later too. Looks good on the book, looks bad on the bone
stevefan1999··on Steve Jobs in Exile – New book on his years at NeXT Computer
Did people literally forgot that John Carmack's Quake was made on a NeXT workstation...
stevefan1999··on Int a = 5; a = a++ + ++a; a =? (2011)
As a guy who writes compiler the answer for this is it depends on the type of parser you are using.

LL and LR parser generates different derivation, and as such it is deterministically non-deterministic, hence UB.

stevefan1999··on Web Server on a Nintendo Wii
I wonder what kind of apps you can run. Matrix certainly won't but IRC server? Probably
stevefan1999··on If AI writes your code, why use Python?
I won't say it is just because of sexual orientation, but more because of the identity politics associated with it.

Not just like "what kind of gender people I like" this kind of oversimplification but it's more about your attitude towards gender stereotypes and roles, for that's what I saw in a more deep connotation.

stevefan1999··on If AI writes your code, why use Python?
Really controversial but my honest opinion: That's because programming languages, and its natural language counterpart, too, are nowadays increasing and more likely in becoming a political tool, rather than itself being a tech tool.

I observed this through observation of the attacks to Rust due to the huge presence of LGBT people.

Now while I'm pretty much straight myself, I don't reject LGBT people and don't want to partake in identity politics.

I just want things that works no matter what background you have, yet there are some people attacking Rust because of its inclusiveness nature.

And just like Linux is being perceived as nerdy and geeky and "gaming socks ready", the tokenization of things, and there attaching political meanings to it, are quickly coming to everything, so perhaps I'm too general here as well.

Let's say it is not political, but definitely adding more meanings to its technical origin and nature

← PreviousPage 5 of 31Next →