grantseltzer

I've played with bpf iterators and wrote a post about them [1]. The benefit of iterating over tasks instead of scanning procfs is a pretty astounding performance difference:

> I ran benchmarks on current code in the datadog-agent which reads the relevant data from procfs as described at the beginning of this post. I then implemented benchmarks for capturing the same data with bpf. The performance results were a major improvement.

> On a linux system with around 250 Procs it took the procfs implemention 5.45 ms vs 75.6 us for bpf (bpf is ~72x faster). On a linux system with around 10,000 Procs it took the procfs implemention ~296us vs 3ms for bpf (bpf is ~100x faster).

[1] https://www.grant.pizza/blog/bpf-iter/

'Watch' isn't the correct word.

Alive is a very high bar considering the alternative.

Writing it in C, compiling with clang, and loading with either C(libbpf), Go (cilium/ebpf), or Rust (Aya).

You can also write bpf in rust with Aya but i'm not sure how feature complete it is.

For very simple use cases you can just bpftrace.

bcc hasn't been relevant for years.

> it's the resulting content that matters, not how it's presented

What a wild thing to say. If you had a coworker who was brilliant and taught you many great things, but only screamed instead of talking, would you feel the same way?

> If a person anthropomorphizes an LLM in their mind (rather than just in their speech patterns), then they probably have pre-existing mental problems.

Correct, and that's why these tools should be built responsibly under the assumption that people with mental problems are going to use them. It's clear in the article I linked (and my wording linking to it) that it can exacerbate issues for people. Chatgpt told him that he's sane and his mom was trying to kill him. He didn't understand what an LLM actually was.

I'm not claiming the purpose of this prompt is to get better information. Yes, it's just a prompt.

You're asserting quite a lot of bias when you say "What most people want are useful results." Maybe in our circles of software engineers or lawyers, but many people are using AI for companionship. Even if they're not seeking companionship, unless you have a very clear understanding of how LLMs work, it's very easy to get caught up thinking that the chatbot you're talking to is "thinking" or "feeling". I feel companies that offer chatbots should be more responsible with this as it can be very dangerous.

Excellent analysis and deduction!

Can someone with actual fundamental understanding of LLMs explain to me why they think it's perfectly legal to train models on copyrighted material? I don't know enough about this. Please don't answer by asking chatgpt.

Now if only the rest of facebook would die

Contrary to popular belief, you can get a quality education at programs that don't put you $200,000 in debt.

If you want an actual project for this I use this one: https://github.com/knqyf263/pet

What overhead are you worried about?

I believe Cilium (ebpf based) is now default in kubernetes.

Well clear your schedule, it's most commonly used interface is through it's python bindings (a library called BCC):

https://github.com/iovisor/bcc

I don't see a reason to implement anything like this without using bpf.

I wrote a blog post on cgroups a couple years ago that's still accurate and goes further into depth and gives workable examples of using them both inside and outside of containers: https://www.grant.pizza/blog/understanding-cgroups/

I'm not so convinced how great of a job they did.Bank of America has by far the worst user experience I've ever seen. I had an experience recently where I changed my password, and after 2 days my new password stopped working and my old one did.

No shit Google has the final say, they're the maintainers. They pay developers millions of dollars to make the final call on what's best for the language. Even if someone were to start a community fork there would still have to be a central board of governance.

Suggesting that we should split the community for a feature that you like in other languages is a dumb and lazy argument.