Hacker Newsnew | past | comments | ask | show | jobs | submit | more FuckButtons's commentslogin

It’s good enough that I’ve been having codex automate itself out of a job by delegating more and more to it.

Very excited for the 122b version as the throughput is significantly better for that vs the dense 27b on my m4.


Hopefully just kill off the javascript for everything mindset to be honest.


Exactly, the Venn diagram of programmers using c/c++ and programmers who can use a garbage collector for their workload is two circles.


Definitely not true. I've been using Boehm GC with my C/C++ programs for decades — since the 90s, at least.


Does this also hold true when you look at codebases that others also worked on, rather than just you?


A lot of C++ programmers use C++ and garbage collection daily because their C++ compiler uses a tracing garbage collector.

https://gcc.gnu.org/onlinedocs/gccint/Type-Information.html


Which only appears relevant if you disregard critical differences like this:

The GCC garbage collector GGC is only invoked explicitly. In contrast with many other garbage collectors, it is not implicitly invoked by allocation routines when a lot of memory has been consumed. [1]

[1] https://gcc.gnu.org/onlinedocs/gccint/Invoking-the-garbage-c...


Except for:

- Me. I'm a C++ programmer.

- Any C++ programmer who has added a GC to their C++ program. (Like the programmers who used the web browser you're using right now.)

- Folks who are already using Fil-C.


I’m also a C++ programmer, I can’t even use half of the C++ stdlib for real time thread work, I certainly can’t use a GC.


There are many C++ programmers and we are not the same!

My original foray into GCs was making real time ones, and the Fil-C GC is based on that work. I haven’t fully made it real time friendly (the few locks it has aren’t RT-friendly) but if I had more time I could make it give you hard guarantees.

It’s already full concurrent and on the fly, so it won’t pause you


∃ ≠ ∀


A few important, production C++ codebases do use tracing mark/sweep GC.

Most famously: Chrome does (Oilpan), GCC does (or did), Unreal does (for core game state heaps), I think WebKit also does.


Except if they happen to work with .NET, Unreal, V8, COM/WinRT,...


if I understand you correctly, this is essentially what vllm does with their paged cache, if I’ve misunderstood I apologize.


Paged Attention is more of a low-level building block, aimed initially at avoiding duplication of shared KV-cache prefixes in large-batch inference. But you're right that it's quite related. The llama.cpp folks are still thinking about it, per a recent discussion from that project: https://github.com/ggml-org/llama.cpp/discussions/21961


Because he should know better? Because it’s obviously a shit show but he keeps on being very vocal about his shit show? Because it’s annoying to have to see yet another delusional vibe coded project being hyped up instead of this forum being used to discuss actually industry relevant information?


> Because it’s annoying to have to see yet another delusional vibe coded project being hyped up instead of this forum being used to discuss actually industry relevant information?

It's industry-relevant. This is what the industry is now. All in two short years.


he's doing it in the open. Its instructive for us all either way.

> he keeps on being very vocal about his shit show?

I'm not really sure what this complaint is. You want someone doing something to not.... write a blog about it?

> Because it’s annoying to have to see yet another delusional vibe coded project being hyped up instead of this forum being used to discuss actually industry relevant information?

I think I've seen around 2 posts, one the original gastown one and then the gascity one. Is two posts in like a year too much or do I miss a midday rush where the front page is all Yegge?


What is the point of a cli if you aren’t allowed to script it? Nonsense.


It seems pretty transparent that they are heavily resource constrained, (training run for Claude 5.x, higher usage / growth than anticipated). I don’t disagree that their long play is monopolistic pricing, but what we’re observing seems better explained by the fact they have a very tight compute budget they are trying to optimize over to put as much as they can into next gen experiments / training to make sure they stay competitive over the next 6-months / year.


No, when you’re the richest most powerful men in the world you don’t get compassion.


It only feels like that if you’re just catching up. The logical consequences you are just realizing are the reason OpenAI was founded.


The limitation is efficiency and efficacy. If you have to add an additional layer of inference to any request you’re negatively impacting your bottom line so the companies, which are compute bound, have a strong incentive to squeeze everything into a single forward pass. It’s also not clear that a separate model that is smaller than the main model will perform better than just training the main model to detect prompt injection. They are both probabilistic models that have no structural way of distinguishing user input from malicious instructions.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: