• 4 Posts
  • 31 Comments
Joined 3 years ago
cake
Cake day: June 4th, 2023

help-circle
  • I love Debian, I use it as a desktop and it’s installed on all my machines, but it has some problems/quirks that make it hard to recommend it for desktop usage.

    First one is that it doesn’t have sensible defaults, the OS expects a user who is an admin, one who will configure things to their liking. I would consider this to be a feature, if it weren’t for the second problem, the documentation. The documentation is wayy too dated, lots of pages in the wiki haven’t been updated since Debian Buster days!

    A better recommendation would be probably a Debian derivative like MX Linux. They are more opinionated, but as far as I know, they come with better defaults and offer newer guides, documentation, etc.




  • I went into hyperbole there, my bad. I agree wholeheartedly with your reply.

    Admittedly, I have a bias here, since using water for food production feels more justified. Same with some industries. But I concede and agree that meat is not the most efficient way nor the most ethical source of food.

    I think data centers rub me the wrong way because, for me, they are the straw that broke the camel’s back. We see how lots of countries pull out of the fight against climate change (not that they were really fighting it, but at least they seemed to care a little) and now they waste more resources on things that are, from my point of view, not productive nor useful to most people. (Ethics aside, my view on Gen AI is that it is a tech that is not mature enough, and that the use of brute force to make it better is a waste of resources)

    I will try to be more mindful about my comments.



  • I should have seen this coming, even then, I am extremely sad.

    This was of the few uses for machine learning that I think it was genuily useful. Mature enough to make an impact without destroying the enviroment.

    It was also my main argument for people that are full hate AI. Now gone.

    I hope others companies continue this research. It’s in this type of problems where machine learning truly shines.

    GenAi is still not mature enough, throwing so much compute power at it, is mostly wasting resources.

    Imagine what could be done if all the computer power and researchers wasted on chats went to this type of project.



  • Unpopular takes incoming.

    Signal.

    Way too many red flags.

    • Why ask for mandatory phone numbers? You could at least make it opt in.
    • Why we can’t inspect the latest server code?
    • Why not make it easy for people to run their own servers?

    Do you truly believe that a company that wants to preserve your privacy would take this direction?

    And i don’t care how secure the protocol is, how well the code is audited. They can still map your social graph.

    Anyways, because of my threat model, i still use Signal. But if i were an activist i wouldnt touch it.

    More unpopular takes:

    Tor and Mullvad probably compromised too. If a service gets too mainstream, I dont believe for a second that they would let it run without care. They would take it down, or control it.

    Now, these services are still usefull. For example mt threat model is to deny my shit to the big tech. So they are useful if you want to escape data collection for adversiment purposes.

    I don’t think they would burn the reputation of these services for low hanging fruit like selling data for ads.


  • Some of the few i played the last year:

    Tome 4: One of the best traditional rogue-like out there, with lots of classes and items. I won’t lie, the graphics are dated but they grow on you. The greatest thing going for it, it’s the number of races, classes, items that you can combine. You can play it for free on their page or get the paid steam version. It has some paid dlc. Sadly the developer hasn’t appear in a while, i think they had a few real life problems, i hope they get better. Still the game is still top notch and with lots of content. If you are into traditional rogue-likes this one is a must play.

    Cogmind: A traditional rogue like game but with a twist on the genre. Instead of exp, you get parts for your robot and part of the progression is learning about specific mechanics (info-war, stealth, hacking, etc). It has a really modern immersive ui, and really nice ASCII animations. It’s tagged as early access but i consider it complete. It has dozens of different endings and ten years in development. But take my heed, It’s pretty hard and may not be for everyone, some runs can be a bit frustrating. Check some videos or reviews before buying.

    Caves of Qud: Lets keep rolling with traditional rogue-likes. This one is fresh out of the oven, I think they finished it last year and it was a fantastic ride. Unique setting in a post-apocalyptic world with mutants psychics and cyborgs. It has lots of customization, you can be a gunslinger mutant with 5 arms, each with a gun, four legs, two heads and that would be a normal run.
    Just so you get around the level or craziness the game has, you can have your face chop up by a conscious tree, pick it up and wear it over your already mangled face just to assert dominance (Really, the chopped face gives +1 EGO when equipped). Check it out, fantastic game.

    Bonus track:

    Song of Syx: A civilization builder, that plays like a mix of colony management and paradox grand strategy games. Great game with a great developer.

    Dyson Sphere Program: Ok, this one may not be an indie game, not sure, it has way more developers than the other games. But they are great and the game is gorgeous. If you like Factorio like games this one should be your jam.

    Non-Steam Game:

    While not in steam if you have experience running things with wine and proton, I suggest giving it a chance, you won’t regret it.

    Star Sector: Space game where you explore the galaxy with an ever growing fleet. It seems to take inspiration on Star Control. The battles are fantastic, the effect when a ship goes booms it’s glorious. It has a great community with lots of mods. It’s still in development but i consider it already finished. It has enough content for hundred of hours and that’s no even counting mods. There is a big update coming, I think at the end of the year, may be worth waiting for it.



  • Fully agree, while people losing jobs to ai is sad, i have zero sympathies for meta employees. They knew pretty well how insidious that company is, and as you said, they are smart enough to find a good enough job in another company.

    A bit off topic, but all this ai stuff replacing so many jobs is a bit baffling to me. At least for certain jobs. It’s not like in the industrial revolution where you could see tangible results in production and efficiency. Even if it come at a terrible human cost you could see the cold logic.
    Here in change, they are firing good talent only to get a subpar costlier replacement. Sure there are some places where it works, but i don’t believe it’s yet to the scale where you would fire so many people.

    I suppose the wet dreams of a future with no employees is too good to pass.


  • There are plenty, as others have said, it has been able to help a lot in science. It has been used for finding new drugs, search for composite materials.

    Sadly i don’t think the pros will outweigh the cons, the society as is right now won’t let it happen. I truly think this tool, even without any fancy AGI, if used in a responsible manner could help us solve some of the most pressing problems the world has.

    But it won’t happen, look at the irresponsible manner the capitalist society has rolled out this tech, which is still in its infancy. It has wasted tons and tons of energy on models that are not yet fully mature. Brute force training of shit models just to get the best monthly benchmark over competitors. It’s so sad to see all the promised goals in terms of saving energy and water that went to garbage the moment the AI hype exploded.

    And what we got for that? A bunch of models causing problems because they are not ready. Executives cutting workforce, in part to justify the use the new shiny tooling, only to get shittier services. And worse of all, shortage of water and energy to pump more compute.

    Even if we get to a future where AI is cheap, powerful and capable, it will only be used to widen the gap between the wealthy and the common people.

    Stephen Hawking said it way more eloquently already in 2016

    The potential benefits of creating intelligence are huge. […]. Perhaps with the tools of this new technological revolution, we will be able to undo some of the damage done to the natural world by the last one — industrialization. And surely we will aim to finally eradicate disease and poverty. […] In short, success in creating AI, could be the biggest event in the history of our civilization.

    But it could also be the last, unless we learn how to avoid the risks. Alongside the benefits, AI will also bring dangers, like powerful autonomous weapons, or new ways for the few to oppress the many. It will bring great disruption to our economy.

    Link Article

    Sorry for the doomer rant, its a conflicted topic for me, feel free to skip.



  • (From South America)

    About phones, same advice as monovergent, i use a separate phone for things that require google and another one for work. It’s not ideal since they probably can correlate them, but it’s the best compromise i found.

    About ID, you are required to show it for most of the things you ask, which I think is reasonable. We don’t have face id thankfully yet





  • I did, but i don’t know if changing things there will cause conflicts.

    Since i have no experience using luci i don’t know what behaviour is normal or expected, and to what extent the custom software glinet uses changes it. Do you think that tweeking the preconfigured networks would cause conflicts in the long run?

    After sleeping over it, im leaning towards going full vanilla. It should be easier to find tutorials to solve and isolate problems.

    Anyways thanks for your comment.





  • Fair enough, now i feel a bit ashamed since you are way way more knowledgeable than me. I have only been a Debian user for a year and half.

    I made the reply because i remember that when i was looking to enter into Linux, Debian testing was recommended as a great compromise between stable and unstable.

    My surprise when i went to the Debian wiki and said, pretty ambiguously at that, that i shouldn’t use it! Reason being that it wasn’t as updated in security patches as stable. No one told me that bit when i was asking. Since i didn’t know the risks involved, i took the safe option and went with stable, in the end loved it.

    I have to admit that for your case it makes sense to use it. You know the risks, know where to patch it up, and it helps to contribute to it by testing it and submitting bugs. Thank you!

    I do still think that testing shouldn’t be recommended, but i see and agree that it has it’s niche where it works and can be great for some people.

    Anyways, i hope i didn’t came too hostile in my first reply! Cheers


  • I disagree, the strong points of debian are the stability (long periods of testing, without new changes) and security (by applying security updates quickly).

    Using testing or sid means to forego the strong points. At that point you are better served by other distros which focus on having newer packages.

    Also i would be cautious about using Debian testing (forky).As far as i know its the worst in terms of security. Stable has security update priority over testing. And some people say even sid it’s better on that front by having even newer packages.

    Disclaimer: I daily drive debian stable and game on it without hiccups. Rock solid. BUT i have 7 year old amd rig and the games are not demanding.