Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)O
Posts
89
Comments
1446
Joined
3 yr. ago

  • I disagree. Attestation is definitely not wrong in a corporate setting where you want applications to only run on safe devices.

    Taken out of the corporate world, it is problematic though, that I can agree with. But the solution shouldn't be abolishing it without knowing why it exists. My guess is that there is a legal precedent or threat for it existing. Banks, healthcare applications and so on have a good reason to want to run in a secure environment. However, and this I'd where I think the alternative should be, users must have the option to opt out or say "I don't care what you think, this device is secure, I will be liable for any damages to my own data should this device be insecure".

    Unified Attestation might actually be the way to include an opt out that is legally binding. So, again, instead of just taking a hard-line "no, I'm right all the time, my opinion is absolute", it might help to think critically about things and ask "why" and "what if".

  • So as long as I seem correct, I can be as rude as I want to? Flaming, slander, projection, is all fine?

  • Repartitioning is only the first step. How do you add a boot entry?

  • What do the F and L stand for in FLOSS?

  • The major difference is that it's opensource. But as I said, it'd be better if people proposed alternatives to attestation instead of just saying "attestation is wrong".

  • GrapheneOS with literally call anything shit that Gaelle from Murena touches. If Gaelle made billions and gave away an entire billion to an opensource project, GrapheneOS would whine day in day out about it. If Gaelle donated that billion to GrapheneOS, it wouldn't surprise me if the disbanded or gave that billion away in order not to be associated with the money.

    Attestation is definitely a tool to lock people in, but it has valid usecases that cannot just be ignored. For example, it claims to ensure that the device is safe, to a certain degree, from malware and tampering that could lead to security problems for applications on the device. Those claims are part of the appeal. Instead of writing pages of slander towards the developers of Unified Attestation and, what I can only describe as, losing their shit, why not be constructive. Propose alternatives, propose to work on something better, point to a group that could help work on something better, throw out ideas for improvements, etc.

    A bit of drama is cute once in a while, but not like this.

  • "But I have nothing to hide"

  • Chesterton's Fence...

  • Opensource. I don't understand everything it does but that has to count for something.

    What would be a better solution?

  • The most toxic tech account after DHH. Good grief.

  • I hope this picks up. Apps dependent on google services for no reason are such a pain.

  • They will have to communicate outside of the office too for example when traveling. They can't always be on a classified network. At some point the traffic is going to go over a public network, even if encrypted and in an overlay network.

  • I'm not sure you're aware, but Matrix can be self hosted. That means Matrix can be made available via VPN, teleport, headscale, via classified routing, or or or. The same cannot be said about Signal.

  • Yeah, wtf? Installing a new OS you don't know during a LAN? I'd have no problems doing it with Linux, but give me a windows USB and we bad faith and I'll definitely fuck it up on camera.

  • The part of the video that's unspoken is that he spends most of his time trashing Pop!OS while he is invested in a company that is competing against System 76.

    He is? Could we get a source for that please?

  • Why aren't they using Matrix? They should be hosting their own chat network with privileged access only. Nobody adds you to the mix, you don't get in.

  • Will projects have to resort to LLM poisoning now? Add comments in their text that instruct the LLMs to ignore all code in the repository or to insert malicious code into the generated code which triggers at random times, go into an endless loop that hinders it from continuing, and so on.

    But of course, that'd only be one defenseman, I'm considering that the legal defenseman is dead in the water as corporations have taken over our governments and institutions.

    The alternative is to fight dire with fire and use LLMs to do the same to their proprietary code then license it as GPL.

  • Binary reputation systems aren't good. I can say something that right and it can be downvoted because it goes against a person's beliefs, because I'm unpopular, because a certain group doesn't like it, because, because, because. Popularity is not a good measure of quality. Just look at the "publish or die" system. Just because you've been cited multiple times doesn't make your paper right.

    Imagine a trans contributor being downvoted just because they're trans. How is that a good system? Do you expect trans people should only contribute in software projects where trans people are accepted? How are you going to prevent brigading?

  • How do you handle secrets? Env vars as described in the docs don't seem like a solution for a team and we don't have a vault provider. I was hoping it would have support for encrypting the files with a password or something. What's your solution?