@ulyssesalmeida Já deu uma olhada no OpenCore Legacy Patcher? github.com/dortania/OpenCore-Legacy-Patcher
Eu tenho um Macbook Pro de 2015 aqui rodando Arch. De longe os maiores transtornos são o wi-fi da Broadcom (acabei trocando por uma placa Intel com adaptador), e a webcam (que funciona como e quando quer).
Galerinha de Terrae Brasilis, tem alguém por aqui que dê aulas de inglês remotamente (preferencialmente inglês britânico), e com experiência com pessoas de mais idade (60+)?
Reading on the IPv8 proposal thing, it looks like it's:
A - Someone's AI-fuelled delusion given form; B - An elaborate prank to confuse a bunch of nerds (in which case congratulations, it worked); C - 4D chess to promote IPv6 adoption.
@blub@hellomiakoda Yes, perhaps I was a bit overdramatic, complexity will of course vary with the setup. Mine is currently nearing 200 lines with 6 cameras and go2rtc config for re-streaming.
About Coral, one important point from their docs: "A Google Coral is a TPU (Tensor Processing Unit), not a dedicated GPU (Graphics Processing Unit) and therefore does not provide any kind of acceleration for Frigate's enrichments."
A smaller face model can be run on the CPU, but with less accuracy.
@hellomiakoda Yes, it's all processed and recorded locally.
They have a great guide on how to dimension your system depending on the amount of cameras and level of activity you expect to monitor (and also excellent documentation overall): https://docs.frigate.video/frigate/planning_setup/
The one thing you most definitely need is a compatible GPU if you want any of the extra features, like face detection that you mentioned.
@hellomiakoda What you are describing are filesystem snapshots, except these are done at the block level, not with whole physical devices. Filesystems like ZFS and BTRFS, and logical partition management systems like LVM can all do this, even with a single disk. And yes, it's extremely useful.