risho

is there any understanding of how big of an improvment av2 will be over av1?

in what way does this restrict how you are able to use the model?

remixing ideas that already exist is a major part of where innovation and breakthroughs come from. if you look at bitcoin as an example, hashes (and hashcash) and digital signatures existed for decades before bitcoin was invented. the cypherpunks also spent decades trying to create a decentralized digital currency to the point where many of them gave up and moved on. eventually one person just took all of the pieces that already existed and put them together in the correct way. i dont see any reason why a sufficiently capable llm couldn't do this kind of innovation.

their software is not great but they literally make the best hardware on the planet right now. you don't get to being a 4 trillion dollar market cap by being trash. they must be doing something right.

4k fsr uses less vram than native 4k but more vram than native 1440p

there is almost no one who has multigigabit internet and even for people that do, you spend significantly less than 1 percent of your time on that device downloading. its a complete non issue. this device is a midrange at best pc, so having a gigabit connection is exactly where it should be. if you want to have the best of the best build a pc.

the 8gb vram is very concerning to me. it claims to be 4k ready and 8gb of vram is nowhere near enough for 4k gaming natively. they say that this is offset by using fsr upscaling, which is fine, but then you need whatever amount of vram that is necessary for running the game at 1440p or 1080p and then additional vram for the fsr. this will be fine for casual games or even AA games, but I can't imagine AAA gaming on this thing being anything less than a disaster. hopefully i'm proven wrong.

i dont think they could have possible chosen a word to make me want to use it less than gamified.

when this project was first announced I was incredibly skeptical it would ever become something useful. then sometime last year they actually put out something that worked way better than i ever imagined and i became incredibly optimistic and hopeful. then hector, lina and alyssa all left and this project appears to be on life support.

i really tried to use cursor and really wanted to like it but i constantly ran into this problem where the agent wasnt able to see what was happening in the terminal.

yes you can

>There won't be an M4 Ultra as the chip was not designed with a fusion connector.

which is the same thing that people said about the m3

codex has improved DRASTICALLY over the last 2 weeks. your claims about it were true in the past but far less true today. its still missing a little bit of polish compared to claude code, but i suspect it is much closer today than you realize. either way the lack of features of codex even in the past was never caused by hubris of openai knows better than you, it just hadn't implemented it yet. it is a brand new project that gets commits to the project every single day.

i really hate the fact that every single model has its own cli tool. the ux for claude code is really great, but being stuck using only anthropic models makes me not want to use it no matter how good it is.

which was expensive compared to it's competitors

omarchy has brought in thousands of new linux users that previously had no interest in desktop linux. its one of the best things that has happened to desktop linux in recent memory. most everythign else in linux is incestuous self referential stuff for people who already use linux. that is why it is different.

i will note that their source appears to be moore's law is dead which is a speculative youtube channel that has a long history of being wrong about the death of arc. dude has been predicting the imminent death of arc since the first one released years ago. it wouldn't surprise me if this did lead to the death of arc, but it certainly isn't because this moron predicted it.

this is a psychotic question but have you actually tried doing that? like using a macbook as a vessel for running linux under parallels as a primary use?

as it turns out raid 5 and 6 being broken is kind of a big deal for people. its also far from ideal that the filesystem has random landmines that you can accidentally step on if you don't happen to read hacker news every day.

These companies need to be destroyed by antitrust violations. I am so tired of these tech companies abusing their market position. I want the FTC to stop being toothless and useless and just absolutely crush these companies. The amount of disdain I have for these companies can't even be properly expressed.

the linux developer community is not a monolith and drew devault is an extremist activist gatekeeper.

i've tried codex, cursor, and a few other agentic tools and nothing compares to claude code when it comes to UX. The other service's models are quickly catching up to claude, but the claude code ux is just magical to me. i havent used it with rust personally. like you suggested would be the average user, i've mostly stuck with js and python.

I've been trying it out with openai codex over the last day and a half and I have been incredibly impressed. It has been working quite well. I also had it look over some code that claude produced for me and it said that it would be better to approach it another way and it completely rewrote it in a way that actually was significantly better. The UX for codex is quite a bit worse than Claude Code, but the model has been good enough to justify the switch for now. I'm hopeful that cursor cli will eventually have a good enough ux such that I can switch to it and have access to all of the models rather than needing to use disparate tools for everything. I would strongly suggest you check out gpt 5 for agentic stuff if you are interested.

I think the complete opposite. I love the ux for claude code, but it would be better if it wasnt locked to a single vendor's model. It seems pretty clear to me that a vendor neutral product with a UX as good as Claude Code would be the clear winner.

is there a way to get it to display more information? its stuck not doing anything and i cant tell if that's because it timed out or it is running a script or it is thinking or what is even happening. sometimes it just does things without even giving any feedback at all. i dont know what it is thinking or what it is trying to do and i cant really see the output of the terminal commands it is running. it just pauses every once in a while and asks to run a command.

is there a way to make it more verbose?

over the last week or so I have put probably close to 70 hours into playing around with cursor and claude code and a few other tools (its become my new obsession). I've been blown away by how good and reliable it is now. That said the reality is in my experience the only models that actually work in any sort of reliable way are claude models. I dont care what any benchmark says because the only thing that actually matters is actual use. I'm really hoping that this new gpt model actually works for this usecase because competition is great and the price is also great.

Quantization is a massive efficiency gain for near negligible drop in quality. If the tradeoff is quantization for an 80 percent price drop I would take that any day of the week.

seems like rosetta 2 will be around for a long time, especially considering they are still putting dev effort into game porting toolkit which is heavily dependent on rosetta 2.

What does the native wsl kernel not offer that you need?

I don't understand. Docker/podman/distrobox/lxc all allow you to do the exact same thing without the virtual machine overhead. I think the real win of WSL is that its a best of all worlds. You get to use Windows with access to every game ever made plus all of the proprietary apps everyone needs to use, with all of the upside of having a full and complete linux command line experience.