Unless you have code and specific, defined ways of vote manipulation, just shut this shit down now.

Need a breakdown first before everyone starts freaking out. This sounds like a UUID leak.

While I love the idea, I just don’t see this moving forward unless any of these projects can focus on splitting up these types of projects into a solid base, driver layer, and then UI layer. Instead they are all spending a ton of engineering resources building something from scratch.

So many projects have similarly started the same way and failed instead of working towards a base that replaces AOSP first, then spinning their own UI on top. The big device manufacturers figured out a decade ago this is the right way to go, and these small projects buck that and fail instead of just focusing on the thing they ultimately intend to focus on.

Get a good base that is removed from Google, THEN do this project.

Funny, because it’s indefensible.

Loldafuq

The face of a Denisovan

And then proceeds to describe a face. A skull is not a face.

Pictures would be nice since this article specifically is talking about the features of the skull.

The other issue is we need a rather large sample to find out if this is a one-off, or a standard of a defined species.

Interesting either way.

So confusing…

I’m trying to understand why you would need this when metrics should be exported to a centralized system which you then view with a WebUI if you need such a thing.

OHHHH NOOOOOOOOO

Nothing made with purely HTML would be capable of this.

Wow, they can’t even fake texts properly. What a fucking fail lolzzz

Downvote this idiot propaganda

Basic stuff is fine, but anything video or gaming related is a struggle. I’ve also noticed the power saving benefits aren’t there.

TPM2+Tang+Clevis. You’ll need to get really familiar with the TPM registers to properly set something up that will meet specific criteria, but you can make a pretty safe system with Tang that you can enable/disable remotely.

M1 running Asahi is mostly fine. You’ll be missing some major features though, most notably the GPU hardware acceleration for certain things, and port features like Thunderbolt: https://asahilinux.org/docs/platform/feature-support/m1/#table-of-contents

Also, the leader of the project quit awhile back, and development seems somewhat stalled/slow now.

All data and benchmarks would disagree with you. If you find something showing that SMB isn’t slower than the others mentioned, I’d love to see it.

Yeah, one or the other works well depending on how your network is deployed. Example: Tailscale gets whacky when dNAT issues are present, but ZT blasts through.

Sounds like OP is having the opposite issue as you.

I think you’re asking about the Samba share? If it works, there’s no real downside except the speed and general wonkiness of the SMB protocol.

As I would rank the different options:

1. Rsync+SSH would be fastest if you’re sending entire directories and not packaging first. A bit slower if sending huge files.
2. NFS would be fastest if sending prepackaged large files, but overall slower if sending a bunch of smaller files
3. SMB will be slowest in any scenario, but may be easier for you

Powershell?

Maybe give Zerotier a shot. Similar premise as Tailscale, but a simplified NAT and routing implementation.

My overall question though is…why??? If you have access to a VPN, why would you connect to another location to use it when you can just use it from anywhere?

https://cylab.be/blog/92/measure-ambient-temperature-with-temper-and-linux