I believe (although I won't claim to know details) that Rosetta is unable to handle the code our self-hosted backend emit. I don't know whether it's related to the machine code or the Mach-O object, nor do I know in what way it breaks. Feel free to try it out of course!-- my understanding could be wrong.
Next time you try Rosetta and it fails can you open an issue with the error messages? I think this task would be perfect for o3 to work on, you basically have to browse hundreds of issues on GitHub and find other people that got into the same problem
Why didn't they go with `defer` which is the more ergonomic and established pattern?
`defer` also doesn't need to change the API of thousands of objects to use it, instead now you have to add a method any resource like object, or for things that are not objects, you can't even use this feature.
Do they actually improve the model you can get without cursor? Or in reality all the development goes to cursor's autocomplete without making this available to supermaven subscribers? It is hard to make sure of that from their website and lack of info online.
Running models locally is very expensive in terms of memory and scheduling requirements, maybe instead they should host their model on the Cloudflare AI network which is distributed all around the world and can have lower latency
reply