• 8 Posts
  • 88 Comments
Joined 2 years ago
cake
Cake day: April 17th, 2024

help-circle


  • At least to me, Overwolf is the third or fourth iteration, following acquisitions, buyouts, restructurings, etc. The original FTB launcher worked perfectly fine. It’s mostly just obnoxious now and I make sure not to have it running in the background. No direct rent-seeking behavior just yet, I don’t have an account on there and it’s not a problem.

    Right now I have it on my computer just use it to update packs that are only available there and then yoink them straight into MultiMC.

    AFAIK it is owned by Curse and I guess those guys make most of their money from those godawful wikis and their ads.


    I thought I’d check this before posting it, and it turns out it’s the other way around. Overwolf bought Curse. Worse, Overwolf is a company based out of Tal Abib… that’s two discoveries in one day. I was looking into getting a CaribouMini until I learned where that comes from. Less than two hundred kilometers away from me as the missile crow flies (and sadly, has flown). Great. Fucking great.

    The shitty thing is that a lot of cool pack creators only publish through Overwolf, so I don’t want to delete it just yet, but I don’t like this at all. At best a minor security risk, at worst I don’t even want to know. I just thought it was just some shitty ad company’s Curse buyout as a billboard for more ads. For all the issues I might have with Nexus Mods, I don’t think they’re quite this bad. Concerning that this is the de facto standard repository of MC mods.



  • I think I should have been more clear, this is exactly what I’m asking about. I’m somewhat surprised by the reaction this post got, this seems like a very normal thing to want to host.

    Doesn’t help that some people here are replying as if I was asking to locally host the “trick” that is feeding a chatbot text and asking it whether it’s machine-generated. Ideally the software I think I’m looking for would be something that has a bank of LLM models and can kind of do some sort of statistical magic to see how likely a block of tokens is to be generated by them. Would probably need to have quantized models just to make it run at a reasonable speed. So it would, for example, feed the first x tokens in, take stock of how the probability table looks for the next token, compare it to the actual next token in the block, and so on.

    Maybe this is already a thing and I just don’t know the jargon for it. I’m pretty sure I’m more informed about how these transformer algorithms work than the average user of them, but only just.


  • Oh that’s a good point. I have only ever encountered these on Lemmy or similar places where you are clearly clicking a link that starts with “xn——————“ and then seeing how it ties together on my phone’s browser.

    Maybe we shouldn’t be using these. I did find myself looking at domains with emojis in them, weirdly enough for someone who doesn’t use or really like them. But the fact that this extends to basically any Unicode character is an absolute security black hole.

    Unless the standard is extended to have more guardrails/to make it impossible to resolve domains with the most egregious fake characters. Or better, to make characters interchangeable the same way domains aren’t case-sensitive.

    The learning curve for understanding the actual web and its protocols looks more and more insurmountable to me every day lol


  • Welcome to today’s 10,000. Today’s episode is about Punycode. It’s basically a standardized way of putting unusual characters in a domain name.

    The way the link is shown in your interface/client, it’s giving you the encoded version that looks nonsensical. But if you click on it, the link in your browser’s address bar will more likely render properly.

    I’ve seen this done with URLs that contain emojis, this one contains katakana (?) characters.




  • Thing is, even with how bafflingly evil Google is, the one corporate service I could see myself paying for is the YouTube subscription. I use the phone app a lot, it would make sense for me.

    The problem is that they’re notoriously ban-happy with paying VPN users, due to some of them using their exit countries to pay less for a service. Thing is, if I tried to pay for premium from a country I’d exit from, I’d be paying more compared to where I am. I’m perfectly content overpaying slightly for a few things online with this situation, I don’t buy much, I’m fine. I also don’t know where the line is. If I pay for my account with a card from my IRL location, using the pricing for said location, will I get my account suspended after I jump back on the VPN? It’s not like they’ll publicly announce a clear breakdown of what is and isn’t okay.

    Google knowing I use a public VPN on Google services is not an issue for me. I don’t do anything sketchy, I really just want an uncensored internet out of the eye of my ISP.









  • So, there are other peers, some of which gave me all of the data that I have so far, a small number of them. They can connect just fine to me, so their ports are open.

    There are other peers with only a small percentage of the total torrent that are permanently connected, which might mean that they only wanted to download some of the files, but looking at the file list availability when they were the only connected peers, they had scattered chunks of data and not one continuous folder.

    Now the convoluted part that I feel needs its own post:

    In my quest to get the data, I’ve done some digging on BTdig, and added a few torrents with what seemed to be almost identical content. Most of these were completely inactive, but it seems like my client figured out that there’s overlap with this torrent, somehow. It now shows one of these inactive torrents as having a small % completion, despite also showing that 0 bytes were downloaded. I suspect it was able to match some of the data.

    I wish I better understood that part of torrenting, a lot of what I want is relatively obscure and I love nothing more than seeding files that took months to complete. Being able to stitch together torrents and make rarities a bit less rare is exactly what I want to do.




  • Unfortunately, just referring to Covid still acts like an activation phrase for some people, sending them into a locust frenzy and shutting off their critical thinking or empathy.

    You can’t get a word in edgewise that one of the reasons Covid is so bad for you is that it fucks with your immune system in a way that wasn’t expected of similar diseases and can have horrible combined effects with other pre-existing diseases. No, you must be insisting that someone got a coma from what basically felt like the common cold to them, and you must be a gullible moron.

    Lemmy is growing, and like any large enough user base, you will get these people.

    Also, good on you for posting this, I didn’t know there were any updates over the past year.




  • Everything I learn about this project is so cool. I can’t go through the docs right now, but I’m assuming it can prioritize things like emergency communication over sensor data.

    There’s no public nodes in a 200+ km radius around me on that site someone linked, so something tells me I’ll have to do a lot of guerilla solar panel installation if I want to anonymously set up something.

    I’ve thought about it on and off over the past two years, more of a private network for family and friends than anything, for emergencies and so on. The real, big problem is that I could be accused of espionage and thrown into jail forever if I do this. So I don’t think I’ll see anyone putting any nodes up for the foreseeable future. At least not public nodes.