• 0 Posts
  • 494 Comments
Joined 2 years ago
cake
Cake day: July 7th, 2023

help-circle

  • PM_Your_Nudes_Please@lemmy.worldtoSelfhosted@lemmy.worldWhat is Docker?
    link
    fedilink
    English
    arrow-up
    5
    arrow-down
    1
    ·
    6 months ago

    It can be, yes. One of the largest complaints with Docker is that you often end up running the same dependencies a dozen times, because each of your dozen containers uses them. But the trade-off is that you can run a dozen different versions of those dependencies, because each image shipped with the specific version they needed.

    Of course, the big issue with running a dozen different versions of dependencies is that it makes security a nightmare. You’re not just tracking exploits for the most recent version of what you have installed. Many images end up shipping with out-of-date dependencies, which can absolutely be a security risk under certain circumstances. In most cases the risk is mitigated by the fact that the services are isolated and don’t really interact with the rest of the computer. But it’s at least something to keep in mind.




  • Yeah, this can be an unpopular opinion on Lemmy, because there’s a giant Linux circlejerk. But the unfortunate reality is that changing to Linux does have some major stumbling blocks. The “switching is so easy, just do it” crowd totally glosses over it, but that’s kind of rhetoric doesn’t help long term adoption. Because if some new user has only heard “switching is so easy” and immediately runs into issues, they’ll be more likely to go “well if it’s super easy and I can’t figure it out, I guess it’s just not for me” and abandon things.

    There’s also a very vocal (and toxic) part of the Linux community that basically just screams “RTFM” at every newbie question. New users shouldn’t be expected to dig into a 350 page technical document just to learn the basics of their new OS.



  • The fact that you don’t need to actually supply any real CSAM to the training material is the reasoning being offered for supporting AI CSAM. It’s gross, but it’s also hard to argue with.

    Yeah, this is basically the crux of the issue. When you get into the weeds and start looking at more than just surface-level “but it needs CSAM to make CSAM” misconception, arguments against it basically boil down to “but it’s icky.” Which… Yeah. It is. But should something being icky automatically make it illegal, even if there are no victims?

    I hate to make the comparison (for a variety of reasons) but until fairly recently homosexuality was psychologically classed as a form of destructive/dangerous kink. Largely because straight people had the same “but it’s icky” response whenever it got brought up. And we have tried to move away from that as time has passed, because we have recognized that being gay is not just a kink, it’s not just a choice, and it’s not inherently dangerous or harmful.

    To contrast that, pedophilia has remained stigmatized. Because even if it passed the first two “it’s not just a kink/choice” tests, it still failed the “it’s not harmful” test. Consuming CSAM was inherently harmful, and always had a victim. There was no ethical way to view CSAM. But now with AI, it can actually begin passing that third test as well.

    I don’t know how I feel about it, myself. The idea of “ethically-sourced” CSAM doesn’t exactly sit right with me, but if it’s possible to make it in a truly victimless manner, then I find it hard to argue outright banning something just because I don’t like it.

    This is really the biggest hurdle. To be clear, I’m not arguing that being an active pedo should be decriminalized. But it is worth examining whether we’re basing criminality purely off of the instinctual “but it’s icky” response that the public has when it gets discussed. And is that response enough of a justification for making/keeping it illegal? And if your answer to that was “yes”, what if it could help pedos avoid consuming real CSAM, and therefore reduce the number of future victims? If it could legitimately help reduce the number of victims but you still want to criminalize it, then you are not actually focused on reducing harm; You’re focused on feeling righteous instead. The biggest issue right now is that harm reduction is very hard to study, because it is such a taboo topic. Even finding subjects to self-report is difficult or impossible. So we’ll have no idea what kinds of impacts on CSAM consumption (positive or negative) AI will realistically have until after it is widely available.






  • Quite the opposite. I read best in the corner of a busy bar, or with music in the background. I guess that’s just the AuDHD talking though.

    Similarly with audiobooks, I prefer them when doing menial tasks like driving. Something that I don’t need to actively think about, but which keeps my hands busy. If I’m just listening to the audiobook without doing anything else, I’ll find myself understimulated, and I’ll inevitably reach for my phone. And then at that point I’ll stop paying attention to the audiobook entirely, which defeats the purpose. I need tasks which hit that “Minecraft parkour brain rot” sweet spot to keep me busy but not distracted.





  • Http: Your employer can see every webpage you visit on the work WiFi. If you visit PornHub, they can see which specific videos you watched. If you were logged into your account, they can (depending on how the site is set up) likely even see account details if you visited your account page.

    Https: Your employer can see the base URLs you visited, but not specific content. They can see you visited PornHub, but can’t see which specific videos.

    VPN: You create an encrypted connection with a VPN server, so all of your traffic passes through that. Now your employer only sees the encrypted traffic to and from the VPN. If you visited PornHub, all the employer would see is the encrypted VPN traffic. The same rules about http and https still apply to the VPN server, (for instance, on https the VPN provider can see you visited PornHub, but can’t see which specific videos,) but your employer basically only sees encrypted white noise.

    Tor: VPN servers connected together in a chain, with an entry node, secondary node, and exit node. Since the VPN server can still follow the same rules about http and https, Tor takes it a step farther by obfuscating which user is connecting to which site. You connect to the entry node and establish an encrypted connection. It sees your traffic to/from an encrypted connection, and passes it to the secondary node. The secondary node only sees the encrypted traffic, which it passes to the exit node. The exit node establishes an encrypted connection with the site, which it uses to pass the site data to/from you. So no single node sees you, the unencrypted traffic, and the site. So (without owning at least the entry and exit nodes and performing a rather technically complicated timing attack) nobody has any way of figuring out which site you’re visiting. If you visited PornHub, the entry node would only see you, the secondary node would only see encrypted traffic, and the exit node would only see PornHub.




  • RetroArch is the go-to for most people, because it can emulate just about anything. But first time setup can be kind of a pain if you’ve never done it before; The UI for settings can be unintuitive, you need to dig for what you want, and it’s easy to forget to save your settings because the save option is in an entirely separate page. It also suffers from some software bloat, because it has so many features that it can get bogged down when emulating more intensive systems.

    For Nintendo I tend to use Delta. It’s simple, has cloud saves via Google Drive, and runs everything flawlessly. For PSX, I tend to use Gamma. Again, it has a simple interface and syncs via Google Drive.