Using the @XREAL_Global eyes and beam gives near real time performance in a small package.
Next step is to get this running in a container @MatthewCroughan suggested Nix + Waydroid
Enjoyed learning more about the #scion network protocol and integrating it deeper into #NixOS at the #scionhackathon 2025 in #zurich, thanks @Adrian_Perrig for making something so interesting that I can't stop trying to figure it out.
#weeknotes
@whitegavriel can you please play at https://t.co/NTNohCjNP1?
Been thinking of making the Windows 96 Appreciation Society at my local hackspace, Birkenhack. You should get on the Fediverse too.
https://t.co/YIgIQrQunN 2025 is on
27th August - 3rd of September @ https://t.co/14Kf1XHw5g & https://t.co/hCKuM0T9dA. An opportunity to hack on things before flying out to #nixcon!
Read more @ https://t.co/YIgIQrQunN
#nixcamp#nixos#nix#reproducibility
@ProgrammerDude@jonringer117 Feels like the reason the cache costs so much is because we keep all paths online because of prestige and not because we need them. Hard truth. Grace Hopper points out this contention in her talk.
https://t.co/DTTEvADKVh
@cguida6 It means that you compile linux, and you make sure the dr_mode is set to host, and not to peripheral or otg. I'm not sure why you think you can change this on the running system. The device-tree is something you load when launching the kernel. Haven't seen that work differently.
@cguida6 Can't help with postmarketOS. If you have a device tree with dr_mode set to host, it works fine. I'm not sure what you're talking about with echoing into files. Build Linux with the DTS with dr_mode set to host.
@joseph_h_garvin latest is just a name. It does not mean it will go and fetch some random shit each time. It means it will look at the "inputs.nixpkgs.url" and flake.lock for a specific ver, or <nixpkgs>. Then it will get the latest that was pinned to. It will never change.