Honestly, why enable this kind of behavior in any way? Any user is free to make an informed choice by installing it themselves.
We all know how this goes. Once a critical mass is reached, enshittification begins to milk everything dry. By making it an installer option, you’re legitimizing it and supporting a worse future for the Linux desktop.
Yeah, Flatpak is far better. The most glaring issue: Canonical hosts the only Snap backend, you can’t host it yourself. Flatpak on the other hand is fully open.
Don’t introduce proprietary crap just so companies can profit off of it.
Nice code, good job!
They also removed all previous versions except a very old one with known issues, thus exposing people to more danger than necessary in any way.
Debian is amazing, but you’re right that they are far from noob-friendly. I recently switched to Fedora due to the fast availability of new packages (e.g. KDE Plasma 6.1 with fixed Nvidia drivers), and even the arguably easiest option - Ublue images - had some issues I wouldn’t have been able to fix without deep Linux experience.
But there definitely has been a lot of progress over the last couple of years, and I’m sure that will continue. We just have to be mindful of not participating in creating the next Microsoft. Ubuntu is already seen as the default Linux distribution - the further it gets entrenched, the worse for all of us.
But why move people from Microsoft to another company that is implementing more and more user-hostile “features”, when there are alternatives like Mint? If all the new Linux users are herded towards Canonical, it’s just giving them even more power to extract profits in the future.
It’s far easier to have them start with a community-led project on the same basis. Imagine Ubuntu being enshittified and forked - how should they decide which fork to use, and how can they know it will still exist in a couple of years?
No, it’s not just about stalkers, it’s about harassment in general. But even if it were, even stalkers are still people and don’t work fundamentally different.
Feel free to show any research proving me wrong, but unless you find any, the reasonable position is “humans work the same on this topic as on others”.
I know, but it still didn’t fully remove it.
Sure, but it doesn’t have to be fully removed to have an effect.
The thing is that there really is no price, nor was there ever one. Your suggestion that you think there is demonstrates that the way blocking worked gave people dangerously wrong ideas.
Sorry, but you don’t get to redefine how humans work. There is a price, because friction reduces the likelihood of people following through. Removing that friction increases the likelihood of people following through. You might not want to believe this to be the case, but please read studies on the topic - it’s just how humans work. You don’t get to dismiss negative effects because you don’t believe in them.
Twitter massively reduced visibility for logged-out users, so just logging out doesn’t help, you have to log into a different account. This additional fraction reduces the amount of harassment a lot. Not sure that being “more honest” is worth the price, especially when an info box could achieve the same without making harassment easier.
You seem to be fundamentally misunderstanding my point, as I didn’t mention the average person’s intelligence in any way. All I’m saying is that minimizing the effort required to really try multiple distributions is a terrible way of introducing people to Linux. It will only lead to frustration and rejection. Choosing your bread doesn’t require investing dozens of hours.
No, it absolutely is hard, and those are bad comparisons. Growing up you interact with bread and cars, and you build a preference based on what you’re taught and what you experience. If I go into a new store and see a dozen types of bread I’ve never eaten, I can still make inferences about their taste, texture etc. This is not the case with Linux distributions - if I’ve never used Linux before, I literally don’t know what the hell I’m doing.
And it’s absolutely unrealistic to expect your average person to try a few out. They won’t be able to decide on technical grounds, and they’ll have to use the distribution for some time to build enough experience for a preference. Going back to your car example, it’s like suggesting people buy a few cars and decide which one they like (since they don’t have the experience to make judgements based on short test drives) - you’re asking them to invest a lot of time for something they don’t really need or want.
People learn how to do that while growing up. The same doesn’t apply to software, people usually choose what they know.
It’s 2024 and this guy still can’t read.
Okay, but why do you tell me that I’m wrong and keep going on about unrelated points? I don’t care if the user-facing name is different from the binary name. I have no position on the topic.
I corrected a wrong statement (who is responsible for the .desktop file of an application). You tried to counter-correct me, but did so on an unrelated point (who displays the application name? I’m still not sure). Positions on whether .desktop files defining separate names is good aren’t relevant.
Your Mint/Xed example doesn’t show what you think it does. Mint doesn’t just ship with .desktop entries for a bunch of applications, they are still managed by the respective developers and part of the packages themselves. Mint is also the developer of Xed, so the repository is in their organization, but the .desktop file is still part of the package. If you install Xed on any other distribution, you’ll still get the same .desktop entry, because it’s part of the package.
That is all I’ve been talking about. I’m not sure how your reply relates to that, but it would help me if you tell me what you’re arguing against.
No, your Desktop Environment doesn’t have a huge list of package names to app names. It has a list for all your installed packages, but the list entries are part of the packages.
If your system doesn’t have gnome-system-monitor installed, you won’t have the corresponding .desktop file, because it’s part of the package. It would be incredibly wasteful and unnecessarily complex for your system to get shipped out with .desktop files for all possible applications.
Thanks! Sorry for coming on so aggressively.
Do you think DEs just have a huge list of package names to app names, or how do you imagine this would work?
In reality, it’s of course fully on Gnome, as it’s part of their code. Nobody except for Gnome has anything to do with the name that’s being shown.
I love it. Finally, I remember what these pictures felt like before I knew they were bullshit. It’s a certain mysticism that was missing.
My guy. There is no open backend for Snap. If Ubuntu enshittifies Snap, nobody can host an alternate backend for them. How does the client being open source help you?