• 20 Posts
  • 1.5K Comments
Joined 5 years ago
cake
Cake day: May 31st, 2020

help-circle
  • It is similar to Bluesky, yes. They both got a lot of inspiration from Twitter (before Musk turned it to shit/X).

    And I would say that the discussions are more shallow than on Lemmy. Even though Mastodon has a higher character limit than Twitter and many Mastodon instances effectively remove the character limit, it’s still fundamentally a platform for shortform interactions. Infodumping is rarely seen, because you need to create a silly number of chained messages.

    On the flipside, though, you get to know people. I do appreciate the time I spent on Mastodon, because of that. It’s a very different perspective as not everything is about discussing cold hard facts, but rather also people’s hobbies and struggles and whatnot.


  • Man pages are displayed in less (which acts as the so-called “pager” here), so you can search man pages interactively like you search in less. And you do that by pressing /, then typing your search term and pressing Enter. Then you can jump between results with n and Shift+n. This is also how search works in vim, by the way.

    Perhaps another tip in this regard, to search in your command history with Bash (for re-running a command you’ve previously used), you can press Ctrl+R, then start typing your search term. Pressing Enter will run the displayed command. To skip to older search results, press Ctrl+R again. If you want to edit a command before running it, press or Ctrl+F instead of Enter.
    This UI is a bit fiddly in Bash, but worth figuring out.

    As for Fish, it’s great for new users, because:

    • it has a much more intuitive Ctrl+R UI, displaying all the search results interactively and not behaving weirdly in certain situations.
    • it automatically sifts through your command history as you type and suggests the most recent command which starts with the prefix you typed. You can fill in its suggestion with or Ctrl+F, or only use the next word from it via Alt+F. You can skip to older matches with , which is then a proper search like Ctrl+R in Bash, so not just prefix-matching. And yeah, overall just really useful, because it’ll both make it quicker to run frequently-used commands, and sometimes suggest a complex command which I didn’t even remember that I once ran.
    • its tab-completion shows short descriptions of what most (sub-)commands or arguments do.

    But:

    • don’t set it as your system-wide default shell or there’s some chance of shell scripts not executing correctly. What you should do instead, is to set it as the startup command to run in your terminal emulator.
    • the syntax of Fish is somewhat different to that of Bash, which can be confusing when you’re still learning the Bash syntax. It’s not the worst thing in the world, as it basically only affects scripting and more complex command chains.
      Scripting is not a problem, because you can throw a shebang into the first line to use Bash syntax (#!/bin/sh or #!/bin/bash). You should add a shebang to your scripts anyways.
      And running more complex commands isn’t too big of a deal either, because you can run bash in your terminal to launch Bash, then paste the command into there to run it, and then quit back to Fish with exit or Ctrl+D. Typically you’ll know to run bash, because Fish’s syntax highlighting turns red after you’ve pasted a complex command.




  • Yeah, I don’t like when corporations put stuff like that into their ToS, but at the same time, I 100% understand why every open-source license under the sun has it. You’re giving it away for free, so you don’t want people to sue for more than you’re providing for free.

    Mastodon.social is currently very much in the latter camp of giving things away for free. I also understand that a service is yet another beast than a piece of software, since they hold your personal data and may leak/sell it. But yeah, at this point in time, I wouldn’t want someone to be able to sue Mastodon.social out of existence. I guess, it depends a lot on how it’s formulated in the end…


  • Ephera@lemmy.mltoLinux@lemmy.mlKDE Plasma 6.4 released
    link
    fedilink
    English
    arrow-up
    6
    ·
    6 days ago

    Those Spectacle changes look good. The old UI made some amount of sense, if the primary use-case was taking complete screenshots, but even for that, there’s probably a single shortcut to do that directly.
    And I do find, I generally want a smaller cutout these days, because you can just fit more stuff onto modern displays, some of which is going to irrelevant.


  • Ephera@lemmy.mltoGames@lemmy.worldMarathon is delayed
    link
    fedilink
    English
    arrow-up
    2
    ·
    7 days ago

    I could imagine that they didn’t want to do something called “Destiny 3”, because people would expect that to be better than Destiny 2, which is virtually impossible, if you’re gonna start over from scratch, with how many years of development have gone into Destiny 2 by now…




  • Yeah, the wording is confusing. A long time ago, there was no paid software, there was only software where you got the source code and other software where e.g. it was pre-installed on some hardware and the manufacturer didn’t want to give the source code.

    In that time, a whole movement started fighting for software freedom, so they called their software “free”.


  • Well, it didn’t feel like I’m tweaking to my needs (that came afterwards on top), it rather felt like I’m just undoing design decisions that someone made to cater to their specific needs.

    And I named the time mainly to give an idea of how much there was to tweak. My main problems were:

    • That I could not undo some of those unusual design decisions.
    • That it doesn’t exactly make the system more robust when you need lots of non-default settings.

  • Well, that was just kind of one example to illustrate that it isn’t just a static screenshot, you actually see what’s going on in real-time. It’s also useful when you’re running a longer operation, like OS updates or encoding a video, and want to see when it’s done or that it hasn’t failed. You can just tell when the command output has stopped moving or a popup has appeared…

    But thanks for the recommendation anyways!



  • I tried it a few years ago. I was really impressed by how lightweight and gorgeous it is. In particular, I found it really cool and actually useful that you got a live view of your other workspaces on your panel. You could even fullscreen a video on your other workspace and then watch (a very small version of) it in your panel.

    But yeah, even though I came back to it multiple times, I never ended up sticking around. It would crash regularly (not the worst thing, since recovery was generally seamless, but still meh), but in particular, it had some peculiar design decisions.

    For example, if you double-click a window titlebar in virtually any window manager, it will maximize. In Enlightenment, I believe it got shaded (i.e. the contents of the window got hidden and only the titlebar was still visible).

    Another prominent one was that its applet for connecting to WiFi and such didn’t support NetworkManager, but rather only ConnMan. If you’ve never heard of ConnMan, yeah, I only know it from Enlightenment, too. Similarly, my distro (openSUSE) didn’t package it either (and openSUSE was said to offer a relatively good Enlightenment experience). That’s something which should just work, because you can’t expect people to look up how they can connect to WiFi while they can’t reach the internet.

    And yeah, these are just the big ones that stuck in my head. There were lots of smaller usability issues, too. Many things you could fix by changing the configuration, but we’re talking many in an absolute sense, too, i.e. you might spend an hour or more just tweaking things so that they behaved like you might expect.






  • You can make external tools available to the LLM and then provide it with instructions for when/how to use them.
    So, for example, you’d describe to it that if someone asks it about math or chess, then it should generate JSON text according to a given schema and generate the command text to parametrize a script with it. The script can then e.g. make an API call to Wolfram Alpha or call into Stockfish or whatever.

    This isn’t going to be 100% reliable. For example, there’s a decent chance of the LLM fucking up when generating the relatively big JSON you need for describing the entire state of the chessboard, especially with general-purpose LLMs which are configured to introduce some amount of randomness in their output.

    But well, in particular, ChatGPT just won’t have the instructions built-in for calling a chess API/program, so for this particular case, it is likely as dumb as auto-complete. It will likely have a math API hooked up, though, so it should be able to calculate a logarithm through such an external tool. Of course, it might still not understand when to use a logarithm, for example.