You assume that exploitation and material improvement can not coexist. You can be exploited just as well, by that I mean you're not getting a fair share for what you contribute to the system.
That's amazing, over time I got a few memory related crashes w/ bun but have deep respect for the performance work put in. Hopefully Rust's compiler will help even more.
Off: I'm wondering if now when more JS finds place on our machines and bundle size is 2nd place for most, would a revival of prepack or projects in the same vein would be worth it, especially with agents.
I feel this is the wrong way to go about things and I agree that it rude. Why not start by engaging with the warp project and see if some of this work could be upstreamed and if you like warp, target longevity?
Congratz to the team. I really like zed and started using it quite early, loved the text threads and was using them a lot as I don't think llms fit in a box of only agents, they were a nice way to manage conversations, work through them, edit responses to lead the agent better, copy-paste full text, sad to see them go (text threads).
I'm trying right now the ACP with my own agent and I'm of mixed opinions but that's maybe because I care how my agent works. I believe that for the agent view a plain buffer with small ui elements would be the best ui for an agent conversation but I may have been spoiled by their text threads. I may spin a personal fork but the thought of tens of mins of compile time isn't that attractive.
Edit: I realized I started moving to terminal based editors like helix due to agents: claude -> codex -> custom pi, with the open sourcing of warp I was considering making a native integration for warp + pi but now I'm thinking zed's text threads (~17k lines) + pi might be a better way, any thoughts or ideas?
I estimated that even with heavy usage it would cost your around 30-70$ depending on caching at around 40M tokens. That would give you around double the usage compared to gpt-5.5 on the 200$ sub
Claude seems so frustrating lately to the point where I avoid and completely ignore it. I can't identify a single cause but I believe it's mostly the self-righteousness and leadership that drive all the decisions that make me distrust and disengage with it.
I've forked it locally, to be honest I haven't merged upstream in a while as I haven't seen any commits that I found relevant and would improve my usage, they seem to work on the web and desktop version which I don't use.
The changes I've made locally are:
- Added a discuss mode with almost on tools except read file, ask tool, web search only based no heuristics + being able to switch from discuss to plan mode.
Experiments:
- hashline: it doesn't bring that much benefit over the default with gpt-5.4.
- tried scribe [0]: It seems worth it as it saves context space but in worst case scenarios it fails by reading the whole file, probably worth it but I would need to experiment more with it and probably rewrite some parts.
The nice thing about opencode is that it uses sqlite and you can do experiments and then go through past conversation through code, replay and compare.
The only thing I'm wondering is if they have eval frameworks (for lack of a better word). Their prompts don't seem to have changed for a while and I find greater success after testing and writing my own system prompts + modification to the harness to have the smallest most concise system prompt + dynamic prompt snippets per project.
I feel that if you want to build a coding agent / harness the first thing you should do is to build an evaluation framework to track performance for coding by having your internal metrics and task performance, instead I see most coding agents just fiddle with adding features that don't improve the core ability of a coding agent.
I think it should matter. When you ask the AI something you are in a frame of mind, you have a specific context, the question also holds value and context that might completely change the parsing of the answer or at least the difficulty of it.
What I'm asking and the response from AI through an intermediary lose some context (the prompt), it's like the telephone game where the data becomes more and more distorted, that's why people don't have an issue with their own AI generated answers.
Another issue is that when I'm talking with someone and parsing through what they've said I'm considering them, as a person, taking all available context (some of this might happen unconsciously).
In any case I don't think there is an easy solution to the problem.
I don't disagree with the concept of AI being another abstraction layer (maybe) but I feel that's an insult to a CNC machine which is a very precise and accurate tool.
I really don't understand this. My line of thinking is that if someone is technical enough to root his phone he understands the risks. Why would they force banking apps to detect and not work on rooted phones? Why would the government care so much?
I also meant more advanced optimizations beyond what svelte does, like: inlining, loop unrolling, partial evaluation that would trickle down to the frameworks as well. I am aware that some of these and others are very hard to do on javascript as prepack shows.
I’d actually love to see something that goes in the opposite direction, highly optimized and compiled, where the result is as small, fast, and efficient as possible. I get that a lot of people dislike compilation, but once I have the CI set upI never found build steps to be a problem for me.
Some time ago while I was experimenting with writing Debian benchmarks[0], I found that by completely avoiding strings, using Uint8Arrays, and manually managing bounds/memory, I could squeeze out performance that almost made you forget you were writing JavaScript. I never ended up submitting a PR, but it was pretty eye-opening.
At one point I went into a rabbit hole and tried to build something similar on my own, but it got complicated very quickly given my limited compiler knowledge. That’s why I always thought Prepack[1] was such a cool idea.
From what I know it can't be "rotated losslessly" in all cases, only if the dimensions of the images are multiples of the MCU which are block of pixels whose size is determined by the chroma subsampling. Ex.: With the common "4:2:0" subsampling the MCU is 16x16 the image's height and width must be exactly multiples of 16, otherwise I think it's just visually lossless and uses some tricks that I'm still not sure how they work.
Lindsay Graham looks more like he pissed himself. So after he saw that bullying and the way the president conducted himself he couldn't be more proud? I've met some ppl, some crazies included, but none that distort reality the way trump and his supporters do. Is there like a crisis in the american education system, what is happening there?
The model is expensive, it almost reaches what I charge per hour. If used right it can be a productivity increase otherwise if you trust it, it WILL introduce silent bugs. So if I have to go over the code line by line I'd prefer to use the cheapest viable model: deepseek, gemini any other free self-hosted models.