• ayyy@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    20
    ·
    4 hours ago

    Dafuq is a character.ai and why is it in the same list as google and cloudflare? Like yea my local grocery store had some issues should that be on the list too?

  • HugeNerd@lemmy.ca
    link
    fedilink
    English
    arrow-up
    11
    arrow-down
    2
    ·
    10 hours ago

    Uh huh, just before a major democracy attacked an evil, evil country. Must be coincidence.

  • arararagi@ani.social
    link
    fedilink
    English
    arrow-up
    1
    ·
    8 hours ago

    Strangely enough, twitch streams and chat were working fine but everything else was broken, had an empty following list.

    • pinball_wizard@lemmy.zip
      link
      fedilink
      English
      arrow-up
      2
      ·
      4 hours ago

      It’s seen tech talks by Twitch’s engineering team. Some of those folks are scary smart.

      Not that it takes that much cleverness to avoid using CloudFlare, of course.

      But might be related. Twitch had some clever fallbacks and work-arounds for slow Internet, in that tech talk.

      • FaceDeer@fedia.io
        link
        fedilink
        arrow-up
        95
        arrow-down
        1
        ·
        1 day ago

        If you’re not running your waifu on a local machine she’s just a digital prostitute.

      • tal@lemmy.today
        link
        fedilink
        English
        arrow-up
        30
        ·
        edit-2
        1 day ago

        When you stop using a search engine and use AI-generated responses for everything, I guess having your LLM backends go down could qualify as having a disruptive functional impact.

    • ZeroCool@lemmy.ca
      link
      fedilink
      English
      arrow-up
      50
      ·
      1 day ago

      Well, the actual headline is:

      Massive internet outage: Google services, Cloudflare, Spotify all down, users report

      Character.ai is only mentioned in the lengthy list of services that are down.

            • blargle@sh.itjust.works
              link
              fedilink
              English
              arrow-up
              8
              ·
              20 hours ago

              That’s how I read it too at first- and as bad as ‘erotic role play’ with a LLM chatbot sounds, the other one is so much horribly worse.

              • seralth@lemmy.world
                link
                fedilink
                English
                arrow-up
                2
                ·
                edit-2
                9 hours ago

                Actually RP and by proxy ERP is basically a perfect fit for what LLM are good at.

                Good RP is basically nothing but improv and rolling with a functional scenario no matter how weird it gets. Which LLM are fantastic at doing.

                You arnt worried about accuracy or memory after all. Just that it can string together coherent English and play pretend with you. Literally the ideal use case for a LLM.

            • toynbee@lemmy.world
              link
              fedilink
              English
              arrow-up
              6
              ·
              edit-2
              15 hours ago

              I had guessed the nature of the site and “roleplay.” It is a tad silly on my part to not have guessed “erotic”; thanks for getting me there (and not in an erotic way).

              edit: Had left out the word “there.”

  • Vanilla_PuddinFudge@infosec.pub
    link
    fedilink
    English
    arrow-up
    86
    arrow-down
    4
    ·
    22 hours ago

    I remember the day I went why are we all selfhosting and still relying on a site like Cloudflare?.

    I’m glad I agreed with me on it.

    • Lka1988@lemmy.dbzer0.com
      link
      fedilink
      English
      arrow-up
      1
      ·
      3 hours ago

      I used Cloudflare tunneling for a while, then started to have similar thoughts. I’m off Cloudflare now.

    • talentedkiwi@sh.itjust.works
      cake
      link
      fedilink
      English
      arrow-up
      16
      ·
      21 hours ago

      I’ve migrated about half of my services off of cloudflare DNS proxy. Guess which half still worked this afternoon. (Self hosting pangolin with CrowdSec as replacement). I wasn’t even using the cloudflare tunnel. Just their proxy for some bot mitigation.

      • SidewaysHighways@lemmy.world
        cake
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        9 hours ago

        been really enjoying pangolin so far! i got scared and ended up not installing crowdsec.

        am i missing out? how hard would that be to add after the fact?

        • talentedkiwi@sh.itjust.works
          cake
          link
          fedilink
          English
          arrow-up
          2
          ·
          9 hours ago

          Can’t speak to how difficult it would be to do after the fact. My guess is just adding in the lines in the docker compose. As for CrowdSec, it seems to run in the background and I haven’t looked at it much. I’ve triggered it a few times and locked myself out. So I’ve added my home IP to the whitelist (it’s running on a VPS). It also locked out my uptime Kuma so I whitelisted that too.

    • GreenKnight23@lemmy.world
      link
      fedilink
      English
      arrow-up
      5
      ·
      21 hours ago

      still hard. most projects still CDN online instead of packaging into a single application.

      it’s one of the most infuriating things about self-hosting. so I made my own self-hosted CDN and have rewrite rules that replace most public cdb domains and reroute them to my own local.

      it would be great if I could just have something in the middle that would download once and cache locally on request, but it’s effort that would be taken from my time afk.

  • Mwa@thelemmy.club
    link
    fedilink
    English
    arrow-up
    127
    ·
    edit-2
    1 day ago

    Ohh that’s why everything is slow and this is my sanity check.

    But its scary how if Cloudflare went down most popular websites will go down with it

      • Xylight@lemdro.id
        link
        fedilink
        English
        arrow-up
        18
        arrow-down
        1
        ·
        17 hours ago

        Most people who dislike cloudflare do so for their centralization of the internet, rightly so.

        Me? I HATE THE STUPID “VERIFYING YOUR REQUEST” PROMPTS. THEY TAKE WAYYY TOO LONG AND DONT WORK WITH MY USER AGENT SWITCHER. STUPID WEBSITES WITH 3 MONTHLY USERS ARE TURNING ON ADVANCED “DDOS PROTECTION” JUST TO MAKE EVERY FUNCTION OF THEIR APP TAKE 18 MONTHS

        • nyan@lemmy.cafe
          link
          fedilink
          English
          arrow-up
          34
          arrow-down
          1
          ·
          22 hours ago

          It blocks anyone not using one of its preferred browsers, among other things. It’s become the gatekeeper for a large fraction of the Internet.

            • nyan@lemmy.cafe
              link
              fedilink
              English
              arrow-up
              5
              ·
              8 hours ago

              Minority browsers. Since I daily drive Pale Moon, I’m among the people affected. It’s suspected that they test only the 3-4 most popular browsers, and whether anything else works with their code is up to luck.

              You may think browsers with tiny market shares aren’t important, but all new browsers start out that way. I fear for Ladybird if it ever makes it past the alpha stage, for instance.

        • melroy@kbin.melroy.org
          link
          fedilink
          arrow-up
          21
          ·
          21 hours ago

          He is not wrong. Look the whole internet is basically one centralized cloudflare if we continue this path.

          • Ernest@lemmy.zip
            link
            fedilink
            English
            arrow-up
            6
            arrow-down
            1
            ·
            21 hours ago

            great article, and I had no idea that happened to Brian Krebs, of all people! o.O

            I do think the EFF makes a good point though, and I think personally I tend to be biased towards content neutrality over moderation (at least, more strongly the larger the platform is, and Cloudflare is very large). Not to the point of Xitter, obviously, but I think there’s at least a reasonable argument for Cloudflare in this case.


            that said, after some searching, I did find the following two articles, and I find their arguments against Cloudflare very compelling:

            Fortunately I’m already using end-to-end SSL certs via Caddy, but now I’m considering just moving off Cloudflare entirely and instead providing regular backups to Internet Archive–most of the stuff I host is entirely static and very lightweight.

    • buddascrayon@lemmy.world
      link
      fedilink
      English
      arrow-up
      6
      ·
      10 hours ago

      I’ve been looking everywhere for somebody to explain it. Like, there are dozens of articles outlining exactly what happened. Which I don’t need because I was there. I know what happened. I just have been wondering why and finally you have provided a link that actually gives me a reason.

      So, thank you kind stranger.

    • tal@lemmy.today
      link
      fedilink
      English
      arrow-up
      37
      ·
      edit-2
      1 day ago

      I think that at least some instances use Cloudflare for various things, so depending upon what and how much stuff at Cloudflare is broken, some lemmy instances may be impacted.

      • prole@lemmy.blahaj.zone
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        7 hours ago

        As someone who hasn’t worked anywhere with a profit-motive in over a decade, I don’t know how you guys do it. Truly…

      • The Octonaut@mander.xyz
        link
        fedilink
        English
        arrow-up
        10
        arrow-down
        2
        ·
        edit-2
        23 hours ago

        Your code can compile while you’re in a retro, retrospectively looking back at the action items raised in a previous retro but not assigned any sprint time in the current sprint so you just raised a “what went wrong” about it not being planned and now you have more non-sprint work to do because every problem should be presented with a solution and if you could have that for the next retro they’ll review this retro’s action items and see if you have a proposal for fixing retro action items

        • wise_pancake@lemmy.ca
          link
          fedilink
          English
          arrow-up
          11
          ·
          23 hours ago

          This is why I call scrum masters scrumlords.

          Thankfully my boss is hands off so I took a walk and then worked on an internal tooling side project that really only I will appreciate.

    • corsicanguppy@lemmy.ca
      link
      fedilink
      English
      arrow-up
      13
      arrow-down
      1
      ·
      1 day ago

      You’re not wrong. The gangly stacks of orchestrated black-box containers at a pubcloud beholden to shareholder whim is going to be a huge factor. Sorry grandma can’t get her pills, but - phew - the number will still go up.

    • grysbok@lemmy.sdf.org
      link
      fedilink
      English
      arrow-up
      6
      ·
      22 hours ago

      Same. Feeling pretty good about using Anubis instead of Cloudflare for our dinky systems.