Because they something to lock you in to Ubuntu. They want Ubuntu to be the only thing that uses snaps. They want to get snaps to be an Ubuntu exclusive feature, and once they can start convincing some random closed source devs to ship in only the snap format they have a hook to keep you on Ubuntu. And they want those random random closed source devs to be focused on more of the corporate world so they can sell some support licenses.
because they won’t need to maintain it, they won’t even need to maintain the dependencies, some guy online will maintain the package and it’s dependency for them, whether it’s updated or not, it’s going to launch, that’s the whole point of those style of packaging
Because maintaining snaps is a lot less work for whoever maintains the package, upstream developers, volunteers, or Canonical. If I’m shipping software for Ubuntu and I can use snap, I sure as hell will use it instead of deb.
Flatpaks are so much better than snaps. There’s nothing that Snaps can do that Flatpaks can’t do better, aside from CLI tools. But CLI tools should just be in Docker anyways.
Flatpak is mainly for packaging desktop apps, whilst snap can update the entire distro (kernel, mesa, system apps, cli). Snap does things Fedora needs rpm-ostree for.
In my opinion docker isn’t as useful for cli tools. I need easy access to many little tools, and this results in me having one container with everything. But that doesn’t work well with network capture etc. In the end being able to install packages system wide quickly is really useful.
You could do it of course but it’s far from ideal for the purpose of a general case CLI. Unless your script yourself a wrapper, the invocation is fairly verbose. It also leaves containers behind unless you specifically pass –rm which isn’t default. Then there’s the intricacies of the different ways of passing data to it. Oh and let’s not forget that unless you setup rootless Docker, or you do something dangerous like adding yourself to the docker group, you have to always invoke with sudo. Therefore I wouldn’t say that Docker is an appropriate tool for CLI apps in general. For dependency-bundled CLI apps, Snap doesn’t have these gotchas and is therefore much closer to ideal CLI UX.
It had modern kernel and software. It was the reason I use it. To support newer hardware. Thus was a year ago and I’ve been happy since. Maybe not Manjaro not master branch untested up to date, but as good as you’ll ever need.
Generally I agree. Many of the largest and most popular distros are run by corporate entities: Canonical (Ubuntu and its various flavors), Red Hat (RHEL, Fedora), SUSE (SLE, OpenSUSE), and so on. Many more of the popular distros are community developed but are based on, or draw heavily from, corporate distros. Most of the more “beginner friendly” distros just so happen to be these corporate distros or ones based on them. It would be foolish to think Linux would be where it’s at today without the contributions of these companies and others such as Valve, who has almost singlehandedly made Linux gaming commercially viable. It’s still up to the community, however, to keep these companies honest when it comes to staying true to FOSS principles and compliance with the FOSS licenses they work under. That includes things like telemetry and a respect for privacy and security, allowing for freedom as to when an end user wishes to update their software, and retaining the open source nature of code and companies’ contributions to it. Corporations have the freedom to use and contribute to open source software, and they even have the freedom to make profit from it. But they have no more or less freedom than anyone else has to do so as well, and that’s where we have to keep an eye on them.
I think the most interesting thing out of the Red Hat/CentOS/downstream thing was that Red Hat used the absolute classic argument against FOSS - “they’re getting value out of this without contributing back”. The argument that Red Hat themselves spent so long fighting against and building their company around proving that argument wrong.
I think it shows a shift in mind-set, perhaps born from the IBM purchase, perhaps as they start to feel the squeeze, and that they no longer fully believe in FOSS.
But it’s early days, only time will tell - certainly there seems to be a fair few shifts going on at the moment though!
If you like to brag that you use Arch or Gentoo, or you like a rolling update to occasionally break your system Manjaro style do not use OpenSuse Tumbleweed. Otherwise, go ahead and enjoy. I’ve used for over a year without issue. It’s fantastic.
I think all the flexibility and distributed nature of open source is simultaneously it’s greatest strength and greatest weakness. It allows us to do so much, to tailor it to our specific needs, to remix and share, and to grow communities around common goals. But at the same time, those communities so rarely come together to agree on standards, we reinvent the wheel over and over, and so we can flounder vs big corporations with more clearly defined leadership. Flexibility and options seems to lead to an inability to compromise.
But also I think open source and standards have become a battleground for Big Tech, with different mega-corps looking to capitalise on their ideas and hinder those of their competitors. Microsoft trying to push TypeScript into the ECMCA Script standard, Google trying to force AMP down our throats, Apple saying fuck-off to web standards/applications, the whole Snaps/Flatpak/Appimage thing, WebAssembley not having access to the DOM, etc.
I think one of the great things that open source does is that it effectively puts the code in people’s hands and it’s up to them to get value out of that however they can. But so often now it’s these mega-corps that can garner the most value out of them - they can best market their offers, collect the most data to drive the software, bring to bare the most compute power, buy up and kill any threats to their business, and ultimately tip the balance very firmly in their favour.
Open source software needs contributors, without them it’s nothing - sure you can fork the codebase, but can you fork the team?
Most people do the work because they love it - it’s not even because they particularly want to use the software they create, it’s the act of creating it that is fun and engaging for them. But I wonder if perhaps we’re starting to cross a threshold where more restrictive licenses could start to gain more popularity - to bring back some semblance of balance between the relationship of community contributors and mega-corps.
I was using it for over a year, and it was very stable. My only issue was that the opi method to install aur like packages was not as good as using aur itself, so I have recently moved back to arch.
The file browser looks like windows. But the top bar is very Mac like, maybe if you put a start menu it would help. But to get a windows feel, you’ll need the start menu, application window bars, and the status info all at the bottom (default windows).
Switched to Manjaro after running vanilla Arch for several years and haven’t looked back. I appreciate the slightly less bleeding edge updates and extra added stability around it.
Easy installs are probably less of a big deal nowadays after Arch overhauled their installation process.
(I mentioned most of this in another thread recently.)
I couldn’t fully get on with TW when I tried it because of very large updates appearing at random, which is due to its rolling nature. The problem I had is I would see a massive load of new packages that I didn’t have time to install, and I would wonder “can I leave this for now or is one of those a critical patch?”. On Fedora, my distro of choice, it’s a no-brainer, I just do the upgrade every night and the big version update twice a year when I’m ready. On TW you have more of an entanglement of major feature updates and regular essential patches which are hard to separate out. I prefer the predictability of Fedora, and the release cycle gets the balance right of not having to wait too much to get the latest stuff.
Just to be clear though, overall I think OpenSUSE is a great distro, I’m being critical but it still probably would be my second choice.
I used it for about a year and it was good - I had some issues with some bits of my laptop hardware working out of the box, and I sometimes got into an error when doing an update due to mirror synchronization. If you use see an error similar to this, just give it a while before updating again:
<pre style="background-color:#ffffff;">
<span style="color:#323232;">Downloaded data exceeded the expected filesize
</span>
In the end I moved away to match my server environment. Initially to Fedora (CentOS server) but then to Ubuntu (I was mid upgrading from C7 to C8 when Red Hat cut the C8 SLA and discontinued it, so I jumped ship). Both Fedora and Ubuntu are really solid and support absolutely every feature of all my hardware out of the box - I’ve come to really appreciate their stability now.
linux
Active
This magazine is from a federated server and may be incomplete. Browse more on the original instance.