• 0 Posts
  • 15 Comments
Joined 2 years ago
cake
Cake day: June 25th, 2023

help-circle
  • not sure i agree with that. I mean ok, i recently had three interviews for a company where each interviewer asked me almost the same questions. That was clearly a waste.

    At my place, we do a 30min introductory call with the boss first, to quickly weed out unfit candidates and not waste employee and interviewee time with interviews. if that’s ok, then there’s three interviews of 45-60 minutes, one with the product owner that focuses on soft skills and team fit, one with the team your applying to and one with the other team (like frontend or backend) with more technical things, and also just if you’d like to work with this person.

    no amount of interviewing will ever guarantee that things work out and unfit people can slip through cracks. And i hate wasting time in tons of interviews. But i’d also not want to work at a place where i know my coworkers were hired after just 1 hour quick chatting. That so little time to get an idea of a person, to spot any red flags. Heck, the ‘tell me a bit about yourself’ section of an interview is already 15 minutes and not usually very helpful.




  • You misunderstand, the first two commands are just one time setup to install a specific python version and then to create an env using that version. After that all you need is `pyenv activate myenv´ to drop you into that env, which will use the correct python version and make sure everything is isolated from other environments you might have.

    You can also just create an env with the system python version, but the question was specifically about managing multiple versions of python side by side and this makes that super easy.

    You could also combine it with direnv to automatically drop you into the correct environment based on the folder you are in, so you don’t have to type anything after the initial setup.





  • ‘Programming from the ground up’ the main idea of this one is to teach programming in a bottom up way, so very low level.

    it’s mostly about teaching (linux) assembly to beginners, so in a way it is just learning a new language. But it’s mainly about understanding low level how a computer works, like registers, kernel calls, how function calls are handled, all for beginners. It’s really easy to pick up.

    Knowing those fundamentals can go a long way in understanding other computing concepts.

    Others that come to mind are :

    • Designing Data-Intensive Applications: The Big Ideas Behind Reliable, Scalable, and Maintainable Systems
    • A Philosophy of Software Design
    • Software Architecture: The Hard Parts"





  • there’s a lot of stuff you can do, and you can end up with something usable, though not great, at least not in my experience. NVidia’s drivers are to blame, they don’t really work well with opengl and have lots of issues (and also regressions).

    The 550 beta driver is ok-ish, steam flickers but I can play games. Drivers before 535 also somewhat worked, though it really depends on your GPU.

    But I don’t think you will have it working acceptably without some work.

    Here’s some pointers on stuff to try:

    • check protondb for how other people got games to work, you can filter by your GPU.
    • try running through gamescope or gamemoderun
    • try the modeset=1 (and maybe fbdev) kernel parameters for nvidia drm
    • and there’s tons of env vars and other things that can help, I couldn’t summarize them all here, but as a pointer: XWAYLAND_NO_GLAMOR=1, WLR_RENDERER=vulkan, LIBVA_DRIVER_NAME=nvidia, GBM_BACKEND=nvidia-drm (for the drm above), __GLX_VENDOR_LIBRARY_NAME=nvidia
    • try the beta drivers, if those are available somehow (I’m on arch so they were easy to install), or just different driver versions in general.

    The above is meant more as hints than something to copy paste, so use at your own risk. You can of course always just install a second DE with X11 and log into that for gaming and use your regular DE for everything else


  • that is a terrible idea. Even ignoring the energy use and that market cap is a bad metric for value, especially in a market with as much wash trading and painting the tape and just plain fraud as the bitcoin ecosystem.

    The code base is under the control of around 5-10 people, the mining is under the control of 2-5 mining pools, with the largest two being the result of the largest pool splitting because it was too big and that’s not good optics, the majority of those pols and mining is in China. It has not reliably relayed transactions, being severely congested for weeks on end several times in its history. It does not have a “stable fiscal policy”, it has no fiscal policy. It is limited to 7-10 transactions/second (and no, lightning does not solve this, as you still need a regular transaction to settle, onboarding something the size of the EU would take on the order of 50 years until everyone has their channel, and then another 50 years if everyone were to settle their channels, along with many others fundamental problems with lightning), completely laughable if you want to use it as the backbone of the international financial system.

    You want there to be a proper fiscal policy with knobs that can be turned, so you can deal with crisis/extraordinary events. For instance having your country’s currency tied to the dollar is horrible for managing that country’s economics and only done if there is no other alternative, using bitcoin as the global reserve would be that, just on crack and there’s no reason to do it unless you’re incredibly desperate, in which case the dollar is still a safer bet. You’d hand over control of your financial system to a shadow group of unelected people, so you lose even more autonomy. You want there to be checks and things like sanctions, to prevent fraud, theft and, in the case of sanctions, to have a political tool that is harsh but less so than an actual war. These things are features, not bugs. You can debate whether it’s good that the US is in control of a lot of these tools, but proposing to get rid of these tools altogether (which moving to bitcoin would do) would be even worse. There is no country that has gotten itself out of debt using bitcoin, so saying it is a solution is disingenuous at best.

    These are just the same old bitcoin talking points that make it sound like it could actually work but do not hold up to the tiniest amount of scrutiny. And the end goal is always to just make the price go up so the gambling pays off, not any use case or making the world better or anything.