Hacker Newsnew | past | comments | ask | show | jobs | submit | deagle50's commentslogin

You kinda proved the point...

How?

If you're that easily biased then why trust your assessment?

Where did I say I was biased?

the hypothetical you presented above

It was a hypothetical. How does presenting a hypothetical equate to proving anyone's point here?

you implied that not being given early access could bias you in the other direction. Which in my opinion would demonstrate that you are easily biased. Which would then call into question any opinion you share about the subject.

Someone accused me of being biased in favor of model providers who give me early access, after I praised Fable's performance.

I said "Anthropic didn't give me early access to this model, shouldn't that bias me against it?"

I was explicitly pointing out that their failure to give me early access had not, in this case, lead to me reviewing their model poorly.

I try very hard not to let things like early access affect my reviews of models. I was hoping this particular situation could help illustrate that.


Don't feed the trolls Simon.

Or it could be insane to pay the cloud memory costs when you have tools that can write rust for you.


What "cloud memory costs"? Most Rust code is an informally-specified version of the old Python reference-counting GC. That's how you're supposed to write it, with clone() everywhere, and then dropping down to optimize. You can do the same thing with Go in the other direction by writing an allocator.

People believe a lot of weird things about these languages.


Why would I roll custom allocation strategies in Go (and then be accountable for supporting them) that affect multiple teams and services when I can have an LLM port to Rust and get dozens of additional benefits?


Depends on whether you'll ever need to use a mutable shared tree structure in your project, I guess.

But do remember that the logic you're using depends on nobody on your team reading the LLM code. If you're close-reading LLM outputs, all the complexity of Rust's memory management model is back on the table.


Looks promising, is OpenAI subscription support planned?


step function


Because they don't want you to improve.


codex cli. I switched, no regrets. Also, $20 for top model vs being limited to sonnet.


Plus (the $20 plan) is still stuck on 5.2 right now..


5.3 codex xhigh works for me


Honestly even medium is quite good.


> the compiler can now allocate the backing store for slices on the stack in more situations

Any details on when this happens?



For the same reason we don't commute on horses anymore.

And safety.


ZGC? It should be on par or better than Go.


Java’s collectors vastly outperform Go’s. Look at the Debian binary tree benchmarks [0]. Go just uses less memory because it’s AOT compiled from the start and Java’s strategy up until recently is to never return memory to the OS. Java programs are typically on servers where it’s the only application running.

[0] https://benchmarksgame-team.pages.debian.net/benchmarksgame/...



IIRC the native image GC is still the serial GC by default. Which would probably perform the worst out of all the available GCs.

I know on HotSpot they’re planning to make G1 the default for every situation. Even where it would previously choose the serial GC.


I have the same feeling, and my best guess is that it's the intentional (and imo arbitrary) friction that has been sprinkled into the language. And camelCase.


Lol. Funny enough I actually like camel case because I've spent so much time in Java.

Yeah, I can see that. It's also difficult to put my finger on, but for a language that claims to be simple it seems to make a lot of things needlessly complicated. I'm also not loving how everything is deeply nested structs so I have to do struct.doThefirstThing.doTheSecondThing.doTheThirdThing().etc() all the time.


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

Search: