deleted by creator

deleted by creator

as you might have guessed i haven’t really tried it, but i have been reading about it. that said i have used “drop in replacement” tools like this (we use pnpm at work), and a drop in replacement is not without quirks. they wouldn’t have made a different tool altogether if it was really a 1:1 replacement. just because the commands are the same doesn’t mean it behaves the same. i.e. i doubt one person on the team could be using uv while everyone else sticks to pip

definitely not the real reason for a project like this to exist. Python package management can be nightmarish at times depending on what you’re doing. between barebones requirements.txt, Poetry, and the different condas there’s a ton of fragmentation, and none of them do everything you’d want in an ideal way. above and beyond speed, i think uv is another attempt at it. but it could just be another classic xkcd moment where now there’s just another standard to deal with

i mean, you’re right. i’m just saying it’s a little silly to ship a Python interpreter when there are easier, better supported ways to do the same thing.

looks like tesseract provides C bindings which are probably being utilized in those apps.

no need for Python. there’s a Google SDK, ML Kit, that will do the heavy lifting on this. if that’s not acceptable, TensorFlow, PyTorch, and ONNX support Android, albeit not as nicely integrated.

your image processing pipeline will be imageSource -> RGB encoding -> OCR -> profit. your OCR just needs an RGB encoded image. doesn’t matter if that’s a JPEG or YUV video feed at the source.

as for if there’s an app that fits OP’s exact use case, dunno.

i guess the rare thing is the public commitment, but Apple has generally had a good track record for updates compared to its Android counterparts, who have previously failed to meet their goals or set laughable goals like 2 years.

counting the days at this point

i used fish as a login shell for years before i switched to nushell. it’s not super risky. bash is just as capable of having a system breaking config. worst case login as root with a minimal bash config and fix it up from there. if that’s fucked you need a rescue ISO.

you’d be surprised how fast a model can be if you narrow the scope, quantize, and target specific hardware, like the AI hardware features they’re announcing.

not a 1-1, but a quantized Mistral 7B runs at ~35 tokens/sec on my M2. that’s not even as optimized as it could be. it can write simple scripts and do some decent writing prompts.

they could get really narrow in scope (super simple RAG, limited responses, etc), quantize down to even something like 4 bit, and run it on custom accelerated hardware. it doesn’t have to reproduce Shakespeare, but i can imagine a PoC that runs circles around Siri in semantic understanding and generated responses. being able to reach out on Slack to the engineers that built the NPU stack ain’t bad neither.