@davidjoeldyck I get that but for me I've managed to make it work. I've been a lot more productive. I thought macOS 27 fixed that issue but I guess I was wrong
I'm getting so sick and tired of having to fight macOS, especially the permission system. Apps just randomly lose permissions. Then I have to go to settings disable permissions, wait a bit and then enable again. I think it's time to sell some devices
I forked it as an experiment, but the things I added for me was:
- notification sound
- tts for when I'm to lazy to read
- issue field in commit modal that adds a fixes footer to the generated msg
- orchestration between models, no matter provider
- setting to see thinking stream from opencode
@MiaAI_lab@arena Haven’t tried it for webdev yet. I did have it edit a video, and for that I prefer GLM-5.3 Flash. Slower than Qwen, but it used way fewer tokens, so it finished faster.
Fable 5.1 got a little worse for my use. It misses the obvious solutions and tries to over-engineer.
Ran the same lob app prompt with a detailed spec with fable 5.1, sol, grok and glm. To my surprise, Fable 5.1 and Sol both agreed Grok and GLM-5.3 Flash did a better job.
I trust AI more when I can watch it think.
Grok still makes that easy. Local does too, which is not exactly a flex. It's on your hardware. Of course you can see it.
@NetworkChuck Im not part of the haters but i love coffee. what coffee are we talking? i love a dark roast that is not to bitter. Keep up the good work
It's been a while since I have enjoyed to use my os as much as I have the last week. @dhh did a fenomenal job with Omarchy. I don't like everything but here is the beauty, just change what you don't like and make it truly yours, you don't need to fight the os