• sem@lemmy.blahaj.zone
    link
    fedilink
    English
    arrow-up
    2
    ·
    1 day ago

    I would subscribe to a YouTube channel where a smart person explains other smart people’s blog posts

    • Truscape@lemmy.blahaj.zone
      link
      fedilink
      English
      arrow-up
      4
      ·
      24 hours ago

      Unfortunately a lot of those are just slop voiceover channels now (and before they were just text to speech monotone narration of whatever was provocative at the time).

  • Danitos@reddthat.com
    link
    fedilink
    English
    arrow-up
    16
    ·
    2 days ago

    Excellent post, like usual with Dhole, but I feel the defense of NIST/NSA push on non-hybrid Post-Quantum Cryptography (PQC) is weak. I think the author’s arguments boil down to:

    1. NSA is pushing to use it themselves, so why would they use a scheme they know it’s unsafe.
    2. There’s no mathematical space to leave a NOBUS backdoor.
    3. Peter Schwabe, main author of ML-KEM, used to work with Bernstein and was his Ph.D student.

    My responses are:

    1. They very link they used for 2., says: “[NSA is] using double encryption already.”, so they really wouldn’t be using pure non-hybrid PQC, as opposed to how they intend the rest of us to use.
    2. The same could have been said about SIKE (another non-hybrid PQC proposal), yet it was found to be insecure. Furthermore, author of the cited blog with the mathematical argument for 2. works for Google, so she has a conflict of interest in the topic. However, I’m not an expert on the topic, so the mathematical arguments are beyond me.
    3. I get it that Bernstein has a toxic attitude, and thus a toxic community that follows him, so I guess this was added to shut up a lot of incoming toxic responses that simply link to Bernstein’s blog. But yet, I can only see this as such an irrelevant tangent to dismiss criticism.
  • PierceTheBubble@lemmy.ml
    link
    fedilink
    English
    arrow-up
    13
    ·
    2 days ago

    So use post-quantum (PQ) atop classical encryption algorithms, instead of PQ alone? Because using just algorithms that aren’t nearly as battle-tested, purely on the basis of a hypothetical threat (the eternal, decade-away, modern-encryption-standard-breaking quantum computer), seems rather irrational to me; especially in context with recent efforts from the NSA to push a TLS standard which does exactly that.

    And how exactly does one seek to independently verify quantum resilience? It’s not like any researcher has a true, cutting-edge quantum computer at their disposal. And among those that do have the resources to do so? Right… ₍ₜₕₑ ₙₛₐ₎. But the NSA said they use the algorithm internally, so it must be secure, right? Which isn’t to say the entire cryptography community is composed of NSA collaborators, but some (key figures) definitely could be (irrespective of their residence); or share the same objective.

    But this comment could also be part of the NSA PSYOP, or simply the deranged thoughts of yet another tin-foil-hatted conspiracy theorist.

    • Securetron
      link
      fedilink
      English
      arrow-up
      4
      ·
      21 hours ago

      I think most people misunderstand that PQC migration should be just in-case. Unfortunately lots of our peers are selling fudd to scare people into buying their platform. In contrast it should be PQC readiness, crypto discovery, and validation that you can migrate to safer ciphers if the existing ciphers are prone to failures.

      • PierceTheBubble@lemmy.ml
        link
        fedilink
        English
        arrow-up
        1
        ·
        6 hours ago

        I agree, it should be considered a redundancy, rather than a substitute; at least at this stage. Considering ‘store now decrypt later’, I believe a sense of urgency is appropriate when it comes implementing PQC, but it shouldn’t be redirected to force a change of (robust) cryptographic standards. If the quantum computer threat finally materializes, you’re taken steps against it, and if forever decade-away, you’re not risking compromise for hypotheticals. It’s the best of both worlds, with a, to my knowledge, computationally inexpensive trade-off; not at all worth risking the potential risks (primarily a nation-state actor SIGINT threat, since the basement hacker probably doesn’t have access to the resources required).

    • Ŝan • 𐑖ƨɤ@piefed.zip
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      4
      ·
      1 day ago

      You don’t validate a crypto algoriþm by trying to break it wiþ brute force and failing. You do it wiþ maþ proofs. You don’t need a quantum computer to verify quantum resiliance.

      • willington@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        4
        arrow-down
        1
        ·
        1 day ago

        What stops someone from simply claiming that they have a solid math proof?

        That’s why science and engineering are two different disciplines.

        A qualified scientist can just trust the math. A qualified engineer will want millions of tests no matter how sweet the maths look or what kinds of songs the scientific community is singing. To err is human. Engineers know this best.

        Basically engineers are not in the habit of trusting a thing without a track record.

        • fruitycoder@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          3
          ·
          15 hours ago

          If you consider all the variables your formula can’t be wrong. The things that is impossible because of unknowns

  • felsiq@piefed.zip
    link
    fedilink
    English
    arrow-up
    16
    ·
    2 days ago

    i’ve run into this person’s blog a couple times now and each time i’ve been impressed with both the writeups and the cute art. this post is no exception, but there’s nearly half as many dislikes as there are likes - is there a valid reason for this or is it just regular old intolerance of furries?

      • Truscape@lemmy.blahaj.zone
        link
        fedilink
        English
        arrow-up
        1
        ·
        24 hours ago

        What components do you take issue with? I thought his articles were pretty fair about the accessibility vs. customization argument.

        • banana_pie@piefed.zip
          link
          fedilink
          English
          arrow-up
          4
          ·
          21 hours ago

          I take issue with a common argument about security folks. And that is a narrow view about the threat actors. Google’s 24/7 location tracking of every Android user is perfectly ok in that view. And when Signal pushes you to that ecosystem is perfectly valid and ok as well.

          Those arguments just don’t sit well with me.

          • Truscape@lemmy.blahaj.zone
            link
            fedilink
            English
            arrow-up
            3
            ·
            20 hours ago

            I’m currently using Molly (unofficial signal fork) that I acquired via Accrescent on GrapheneOS, and I haven’t had telemetry issues.

            There’s also a desktop client that doesn’t interface with Google whatsoever.

            I do agree that the “default installation” method through Google Play is problematic, but I don’t think advocating the Signal project as a gateway communications app is a bad thing on the whole (and even default Signal is far superior to SMS and WhatsApp with on-par setup in my opinion).

            Matrix/XMPP is not a bad project by any means, but I don’t think I’d be able to successfully convince my family to set it up (while I can just have both available and use signal with them as needed)

    • Truscape@lemmy.blahaj.zone
      link
      fedilink
      English
      arrow-up
      7
      ·
      2 days ago

      Personally I think the second scenario is more likely, since most people like and dislike without reading the article.

      • Darkenfolk@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        7
        arrow-down
        1
        ·
        2 days ago

        Furry haters:

        "You really expect me to be able to read when there is no blood in my brains and I am at half mast because of all the degenerate furry pictures?

        Yes I know that they are not spicy pictures, no I am not a closet furry, stop projecting."