Skip Navigation

Posts
136
Comments
3509
Joined
3 yr. ago

  • It's a known reoccurring issue. The fix applied in .423 didn't work for it, unfortunately. .423 is still likely the current beta, so I am not quite sure if we are all seeing the exact same problem. (That is confusing, but I am probably seeing a half-patched issue and if you aren't on beta, you are still seeing the original iteration of the bug.)

  • That poor bastard would've hated me. The first thing I do is saturate the logs and test any shitty monitoring that might be in place.

    Honeypots are fragile places, usually. That is why I gave up using simulators and just drop a vanilla image with some lightly masked monitor script or something. Drop your rootkit and botnet keys and GTFO. Fuck key logging. It's interesting, for sure. Some kind of basic post mortem FIM is usually good enough.

  • I am staunchly anti-hype and a firm "whatevers" on the rest of AI. Like you said, it has its place. The biggest issue I have is that it is being pushed as another dopamine fix. I ain't gonna lie, turning an idea into code in 5 mins is really fucking cool. Unfortunately, its really fucking addictive so I give myself multi-day breaks after a day or two of coding and fucking around with my homelab.

    (Tinfoil hat time) If anyone hasn't noticed, there are some clear distinctions between enterprise LLM tools and regular consumer LLM tools. The consumer-grade plans are more prone to random mistakes and forgetfulness. My theory is that more "mistakes" and resulting fixes not only burn more tokens, but also push dopamine levels higher and lower. Aside from coding blatantly obvious security issues, enterprise LLMs are much more reliable.

    LLMs are a tools. You solve problems. If an LLM is solving all of your problems, you are the tool.

  • Typically I would go into vast detail about issues, bit all I can say about search is that it's completely broken when searching across some instances that aren't Lemmy. Zero results for communities that are known to exist.

    I think the issue I quickly saw with sorting is simply the lack of posts in the Connect community. I see mine, and then from 11 days ago. Refreshing the community will sometimes show different results per refresh. Like I said, it's probably not a bug, BUT similar issues have been tied to post refresh issues in the past.

    Also, the text select issue makes it an absolute pain to type out long replies like this or make extensive and detailed bug posts. Once that is fixed, I'll try to resume the detailed bug posts I usually like to make. Simply put, I am not going to fight this issue testing and documenting. That can involve flipping back and forth between testing the steps to replicate the issue and updating a draft post.

    But in a nutshell, there are many small issues. I would like to document them all, but I can't without getting seriously frustrated.

  • It's a reoccurring issue with connect: Move the cursor to edit text somewhere and text selection starts. (Like you would use for cut or copy.) It's random, but when it happens, you can't actually stop the text selection without selecting a bit of text and hitting delete (backspace).

    Yeah, I didn't dig into it too much for the sorting by post so I chalked that more up to being a me problem.

  • Moab, I think.

  • Totally. :)

  • Might be the other way 'round with who is the grand master.

    There had to be a ton of development with the pawns at the start of the game. That usually leads to every pawn getting jammed in the middle of the board which allows the players to move out their back rank.

    Badger went full Leroy Jenkins with his rook, causing the collapse of the pawn defenses. (A pawn taking a rook might cascade the pawn defense destruction since rabbit was willing to sacrifice one or more pawns for that rook. I certainly would.)

    An experienced player probably couldn't resist walking the pawn down, because 1. He could and 2. YOLO. Plus, it's a good lesson for rabbit, regardless. Pawns actually do mean something and creating chaos on the back rank with a single pawn is a lesson I wouldn't ever forget.

  • Connect for Lemmy App @lemmy.ca

    Lots of small bugs.

  • I am just puzzled about how that pawn got to where it is at. Badger moved the pawn into double-check against the knights. Or, the pawn moved into check against one knight and rabbit moved the second knight after. But it's chess, so a thousand other things could of happened, but still... That pawn was hauling ass and there probably wasn't time for anything more complex.

  • It's always been broken, disjointed and tribal. You can tell everyone, but many have already known this. Hell, most of humanity is like this naturally.

    Almost every large organization is this way, really. Most of it is just covered up by goverment or corporate propaganda or some weird sense of duty people have to jobs or organizations.

    This ain't anything new, is my point. It's new and shocking to you, sure. Welcome to the tribe of the disillusioned. It was always better in the past and new people are always going to make it "like it was" and "better". (Quite literally the selling point myth of MAGA, to be honest.)

  • I seem to remember that phenomenon quite a bit in the late 80's/early 90's. Many of my memories from that time are mostly lost to time, I do remember specifically just buying singles for Top 40's as I knew everything else was going to be crap. (MC Hammer comes to mind, but there were many more around that period.)

  • Gross. A Facebook link. This is one of the few times I would have considered another layer of jpg a good thing.

  • Rule

    Jump
  • Genetically-optimal? How in the fuck do they determine that? What happens if their spawn is not genetically-optimal? Recycling bin?

    Geezus fuck people are dumb sometimes.

  • Yeah, its a hell of an experience many people should have. (Many people probably also shouldn't.) At the core of it all, I believe that being able to view problems through a very different lense is a big part of how psychedelics work when used for deep therapy. In many cases, I could see and interact with my emotions and feelings like they were an independent thing. I could almost visualize and touch my own emotions. Being able to see through my problems and get closure for issues that were supposed to be long in my past was a very beautiful thing. Trippy stuff, quite literally.

    Also, (and this is really for others that are reading this) I am not really joking with my personification of a mushroom. I used to think that was just some crazy burned-out hippy talk, but there is so much more to it than that. Yes. A mushroom talking is absolutely a hallucination. That isn't what that literally is though......

    It's more of a very primal, internal dialogue. It's like the voice that we choose not to listen to when we have a "gut feeling" about something and can't vocalize the concern. It's the voice in your head that always knows the right decisions to make even if we brush it off through a normal day. That is the mushroom talking and it's got a really powerful voice if you ever choose to follow Alice down that rabbit hole far enough.

  • I attribute mushrooms to finally breaking my years long journey as a fairly committed alcoholic.

    The decisions or realizations people can have during an intense trip tend to be really sticky for a very long time regardless if it's a good trip or a bad one. It's the nature of the beast.

    But mushrooms be like you described sometimes. I won't go near the dosages I was taking when I was kicking booze. 1-2 grams every once in a while is just fine for me.

    After my last power trip (+5 grams) I saw what I needed to see and probably will never go in that range again. It was a life changing trip and thankfully not a bad one. However, when the mushrooms speak to you like that, you listen. They told me I was done and I was ready to heal on my own.

  • LPT, keep a stopwatch going for psychedelics. I'll start a timer on my phone when I drop and it helps snap me out of any kind of time related disorientation as I peak. For your average time dilation stuff, it's awesome. If I can't see my phone, then time probably is the last thing on my mind.

  • Neural networks just do what they do and they are fine. It's all the idiots that think ANNs are smart and connect them to anything and everything assuming they are intelligent, is the actual problem.

  • I am quite literally an expert in security and know the community quite well. Of course, there were some raised eyebrows to this, but that was about it. It's the big company execs that are calling this the next big thing, because money. (The article basically reflects my opinion. There were some reputable and people quoted, but then there was Jeetu Patel (Cisco) going all weird with this. Frickin idiot.)

    TBH, I have written off all the AI shills I knew in the industry. Sure, make a buck where you need to but goddamn, don't turn full fucking evangelist.

    Disclaimer: I am paid to be an expert in security for a day job, but I still think I can be an idiot with this stuff. Meh. It's paid the bills for the last 20 years.

  • Most of this is just marketing crap from Anthropic.

    Finding vulnerabilities in code and generating complex, multistep exploits with publicly available models is possible now. This biggest hurdles now is setting correct context and actually knowing what to look for. Any "guardrails" for this behavior are easily bypassed by framing the detection and exploit generation as a valid dev style question in the most difficult of situations.

    They likely just trained a model without guardrails in this case.

    What they are doing here is over-hyping a problem and framing it like they are the only ones with a solution. LLM security issues are more in-focus now that companies have dumped a ton of resources into building AI systems they don't really understand.

  • Connect for Lemmy App @lemmy.ca

    NSFW Setting is not persistent.

  • Connect for Lemmy App @lemmy.ca

    Cache file duplicates (maybe)

  • Connect for Lemmy App @lemmy.ca

    Scrolling lag, part deux

  • Connect for Lemmy App @lemmy.ca

    Clicking on thumbnail for external link post shows invalid image data exception

  • Connect for Lemmy App @lemmy.ca

    Lemmy feed gets laggy over 400mb memory usage

  • Web Development @programming.dev

    I am looking for a basic web dev framework that is lightweight and focused around modularity.

  • Connect for Lemmy App @lemmy.ca

    Finally found the source of a hung feed!

  • Connect for Lemmy App @lemmy.ca

    Assorted comment bugs (new or edit)

  • Connect for Lemmy App @lemmy.ca

    Post alignment(?) may move up past top of screen and get hidden by post menu

  • Connect for Lemmy App @lemmy.ca

    "Post Title Above Images" description typo

  • Connect for Lemmy App @lemmy.ca

    Incorrect community shown for post edit; Edited post text not updating

  • Connect for Lemmy App @lemmy.ca

    Unusually low number of comments in c/all

  • Connect for Lemmy App @lemmy.ca

    Comments not always shown; May need to reload post to show comment

  • Connect for Lemmy App @lemmy.ca

    Not all posts will not open when clicked. (press-and-hold image previews still work, so that is nice.)

  • Connect for Lemmy App @lemmy.ca

    I just do what I'm told when I open Connect.

  • Connect for Lemmy App @lemmy.ca

    Right side menu refresh not refreshing for c/all

  • Connect for Lemmy App @lemmy.ca

    Bug test post (Please ignore)

  • Connect for Lemmy App @lemmy.ca

    Feed will randomly stop loading.

  • Connect for Lemmy App @lemmy.ca

    Tap and hold image preview not working