• 0 Posts
  • 14 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle

  • You pull up. Get out. Put the nozzle in. Then you go inside. There, you wait in line for 5 minutes, because the dick from another pump decided to buy a fucking coffee and a sandwich, and the only employee is busy making those for him, instead of operating the pumps. Then you actually pay and get the gas flowing. By the time you’re back at the car, it’s already finished pumping.

    So, there can be a time gap of several minutes with multiple actions and distractions during it. Is it really that surprising people forget to pull the thing out, occasionally?


  • You’re linking a post… From 2010. AMD replaced radeon with their open source drivers (AMDgpu) in 2015. That’s what pretty much any AMD GPU that came out in the last 10 years uses now.

    Furthermore, the AMDgpu drivers are in-tree drivers, and AMD actively collaborate with the kernel maintainers and developers of other graphics related projects.

    As for Nvidia: their kernel modules are better than nothing, but they don’t contain a whole lot in terms of actual implementation. If before we had a solid black box, now, with those modules, we know that this black box has around 900 holes and what comes in and out of those.

    Furthermore, if you look at the page you’ve linked, you’ll see that “the GitHub repository will function mostly as a snapshot of each driver release”. While the possibility of contributing is mentioned… Well, it’s Nvidia. It took them several years to finally give up trying to force EGLStreams and implement GBM, which was already adopted as the de-facto standard by literally everybody else.

    The modules are not useless. Nvidia tend to not publish any documentation whatsoever, so it’s probably better than nothing and probably of some use for the nouveau driver developers… But it’s not like Nvidea came out and offered to work on nouveau to make up to par and comparable to their proprietary drivers.


  • k, so for the least used hardware, linux works fine.

    Yeah, basically. Which raises a question: how companies with much smaller market share can justify providing support, but Nvidia, a company that dominates the GPU market, can’t?

    The popular distros are what counts.

    Debian supports several DEs with only Gnome defaulting to Wayland. Everything else uses X11 by default.

    Some other popular distros that ship with Gnome or KDE still default to X11 too. Pop!_OS, for example. Zorin. SteamOS too, technically. EndeavorOS and Manjaro are similar to Debian, since they support several DEs.

    Either way, none of those are Wayland exclusive and changing to X11 takes exactly 2 clicks on the login screen. Which isn’t necessary for anyone using AMD or Intel, and wouldn’t be necessary for Nvidia users, if Nvidia actually bothered to support their hardware properly. But I digress.

    Worked well enough for me to run into the dozen of other issues that Linux has

    Oh, it’s no way perfect. Never claimed it is.

    I like most people want a usable environment. Linux doesn’t provide that out of the box.

    This both depends on the disto you use and on what you consider a “usable environment”.

    If you extensively use Office 365, OneDrive, need ActiveDirectory, have portable storage encrypted with BitLocker, etc. then, sure, you won’t have a good experience with any distro out there. Or even if you don’t, but you grab a geek oriented distro (e.g. Arch or Gentoo) or a barebones one (e.g. Debian) you, again, won’t have the best experience.

    A lot of people, however, don’t really do a whole lot on their devices. The most widely used OS in the world, at this point in time, is Android, of all things.

    If all you need to do is use the web and, maybe, edit some documents or pictures now and then, Linux is perfectly capable of that.

    Real life example: I’ve switched my parents onto Linux. They’re very much not computer savvy and Gnome with it’s minimalistic mobile device-like UI and very visual app-store-like program manager is significantly easier for them to grasp. The number of issues they ask me to deal with has dropped by… A lot. Actually, every single issue this year was the printer failing to connect to the Wifi, so, I don’t suppose that counts as a technical issue with the computer, does it?

    wacom tablets

    I use Gnome (Wayland) with an AMD GPU. My tablet is plug and play… Unlike on Windows. Go figure.







  • “AI” models are, essentially, solvers for mathematical system that we, humans, cannot describe and create solvers for ourselves.

    For example, a calculator for pure numbers is a pretty simple device all the logic of which can be designed by a human directly. A language, thought? Or an image classifier? That is not possible to create by hand.

    With “AI” instead of designing all the logic manually, we create a system which can end up in a number of finite, yet still near infinite states, each of which defines behavior different from the other. By slowly tuning the model using existing data and checking its performance we (ideally) end up with a solver for some incredibly complex system.

    If we were to try to make a regular calculator that way and all we were giving the model was “2+2=4” it would memorize the equation without understanding it. That’s called “overfitting” and that’s something people being AI are trying their best to prevent from happening. It happens if the training data contains too many repeats of the same thing.

    However, if there is no repetition in the training set, the model is forced to actually learn the patterns in the data, instead of data itself.

    Essentially: if you’re training a model on single copyrighted work, you’re making a copy of that work via overfitting. If you’re using terabytes of diverse data, overfitting is minimized. Instead, the resulting model has actual understanding of the system you’re training it on.



  • Do you expect to find a company that sells a calendar-only subscription? “Calendar - 49c/month”?

    I’ve been looking at lot at all kinds of services and most start their pricing at around 5 USD/month. Regardless of how much actual features they actually provide.

    I’d say your best bet is NextCloud. You can rent some, self host or use a free instance (there’s a couple around).

    Personally, I’m self-hosting stuff on a VPS. For whopping 5USD/month I’m getting things I’d be paying 50, if not mere, if they were offered as separate products by your average service-providing companies.