I’m not particularly militant about Linux distros, but Alpine is one distro I disapprove of in particular. The reason is that it isn’t GNU/Linux – it strips out (copyleft) GNU libc and coreutils and replaces them with permissively-licensed alternatives. I think that (whether intentional or not) it caters too much to corporate interests that exploit “open source” without truly respecting the users’ freedom, and therefore its popularity is potentially harmful to the Free Software movement in the long run.
Considered in and of themselves, permissive licenses are “fine.” They confer all four of the freedoms the FSF lists here, so there’s nothing wrong with them from the perspective of the person receiving the code as an end-user.
The problem is that, unlike copyleft, they fail to bind that recipient to the same conditions and guarantee those freedoms will be maintained for all downstream users who receive the code in the future. They are thus exploitable by those who would take without giving back in return. This makes permissively-licensed code popular with the exploiters, but is bad for the users in the long run.
See, for example, MacOS and iOS: in theory, they’re just BSDs with fancy proprietary UIs, but in practice they can be made so locked-down and user-hostile there’s an entire movement devoted to creating new laws to force Apple to stop bricking people’s property because they needed to replace a bad hardware component. Those four freedoms I referenced earlier are definitely no longer being upheld by Apple, even though Apple itself benefited from them to make the software in the first place.
There’s a reason why copyleft-licensed Linux is so much more popular than permissively-licensed BSD, and resistance to selfish bad actors (even as flawed as it is, what with the “tivoization” exploit of the GPLv2 and all) fragmenting the community with proprietary features is undoubtedly part of it.
There are some opinions mascaraing as fact here and some not very evidence driven at that.
Linux is a beneficiary of great timing. The pre-cursor to FreeBSD, BSD 386, already existed and was much more mature when Linux appeared. The reason that Linux became popular was primarily that AT&T launched a lawsuit against BSD which made its legal status questionable during a critical few years. This was at the dawn of the Internet and the distribution and collaboration that enabled. By the time the lawsuit was resolved, Linux was massively more popular and BSD was left behind. Ironically, early Linux never faced early legal trouble as it was not taken seriously by UNIX players. The Linux lawsuits came later but, by then, Linux had major corporate backers ( see SCO vs IBM with IBM being the on the Linux side ).
Hell, Linus himself has said that he would never even have created Linux is Minix had been free ( meaning explicitly free as in beer, not as in freedom at the time ). In fact, Linus did not want to adopt the GPL at first because it allowed charging for the software.
One reason that Linux was able to advance so quickly ( or exist at all ) was the existence of GNU and especially GCC. I hate the amount of credit GNU tries to take for moderns Linux distros but there is no denying its importance in making Linux viable early on.
Today, Linux succeeds over BSD primarily because of the greater corporate interest. Apple does not really use the BSD kernel either.
These days, the most popular license used in typical Linux installs is MIT and permissively licensed software is more common than GPL. Some MIT communities, like the X Window Project, are decades old and represent strong trends away from corporate dominance and exploration over time. The vibrancy of all the Open Source communities cannot be explained in terms of the world-view expressed in the comment above. I do not have the numbers in front of me to support this but it is my own impression that permissively licensed software generally succeeds more often at creating sustainable communities. Or maybe it is just the FSF. While there are many successful GPL programs, fewer than 500 of them are GNU and there are almost as many abandoned GNU projects as there are active ones.
In my view, the most important GNU program by far is GCC. That evil Apple company you cite created LLVM / Clang and licensed it permissively. They did by far the most work on it and yet have it away. Today, other evil companies like Microsoft contribute to Clang / LLVM as well. LLVM is of course the basis for the Rust language, another corporate contribution. The lack of GPL here does not seem to have prevented any of this innovation, the massive contributions to the community, or collaboration between these giant corporate interests. This is just one example.
I tried PopOS finally after many glowing reviews… and it was beautiful, snappy and had lots of unique features. But while it was very friendly, I had trouble finding my way around. I think still aimed at linux users who are a little more knowledgable. (Not me.)
Technically speaking, OpenRC doesn’t really have any benefits in the real world, some people may claim faster boot times, but that’s debatable on modern hardware. In fact objectively, it’s inferior to systemd in many ways.
The real advantage though is that it’s pretty simple and easy to use, understand and maintain. It follows the Unix philosophy of “do one thing, and do it right”. People who like to have full understanding and fine control over their systems would prefer using OpenRC or similar init systems (with a mix-and-match of other utilities and daemons as per their need), instead of relying on a giant monolothic package like systemd which keeps getting bloated with more and more “unnecessary” features with each release.
Basically, you can say that it’s a difference of ideology.
Systemd is not monolithical, it just happens that it’s made of elements optimized to work well together, instead of running everything through different text parsers when “mixing and matching” random daemons.
FWIW, my (non tech-savvy) mum and dad have been running Zorin for years (and Xubuntu prior to that), without any issues. The only times I’ve had to intervene is for doing an OS upgrade, which was a manual process, but Zorin now includes a GUI upgrader which should make things even more easier.
Folks who claim Linux is too unstable or complicated for home users, and think you need to use the commandline for every small thing, should check out Zorin (or talk to my mum and dad!).
Open source Windows is an interesting premise, but Windows-focused Linux is already a thing going on. Not only has Microsoft basically adopted Ubuntu, but most of their recent projects have been open source. They are actually one of the most numerous contributors to the Linux kernel and it’s mostly to make Ubuntu run better on Azure hardware and to make Windows Subsystem for Linux more effective.
Microsoft has historically not been friendly to anyone else. Until they prove otherwise, this is going to be my assumption. It’s some form of embrace/extend/extinguish.
I love that Linux is everything that Microsoft is not. I love that I have full control of my hardware. I have control over processes. I have control over packages. And user control is the default.
I can already join a Linux PC to a domain and run VPN. I can easily transfer files. I’m good.
Microsoft has historically not been friendly to anyone else. Until they prove otherwise, this is going to be my assumption. It’s some form of embrace/extend/extinguish.
That’s a good point. I feel like there would be a lot of suspicion or skepticism behind it
My guess, based on Proprietary Codebases I Have Seen and the apparent general philosophy of Windows development, is that people would react to the now-open-sourced Windows code with either hilarity or horror (or both at once). There would be critical articles in the tech press. Then a small select group would mine it for low-level interoperability, but none of the code itself would be retained. Emulation layers such as WINE would end up being able to reproduce Windows’ quirks more thoroughly, but finding the important bits in a mess of Someone Else’s Code would slow down development as much as having an exemplar would speed it up. This all assumes that the code was released with an acceptable license.
On the Linux side, mostly a wash beyond some small interoperability gains, in other words. What would happen on the Windows side, I wouldn’t venture a guess on.
I make a bash script for the whole process and document everything in Zim Wiki. I would only make an alias if I want to supply options to an existing command. Just how I do it.
Yeah I have actual notes and processes in my personal wiki too. I don’t like having to look up that one command that I only ever rarely use, like the I need to know if this webpage is serving a 200 or not… what did I do last time??
Zim is a desktop Wiki so no serving issues. The other thing I do is just list my script directory or grep it. If I know I have a script I can often find it that way.
I want to like Joplin and considered using it, but I don’t like the electron base it is. And then converting my existing KB to a new format yet again… ugh.
I’ve found Joplin to be “acceptable” to share notes from mobile. When sharing through a shared drive, it saves each note as a separate markdown file, so it technically is greppable plain text, my only gripe is it puts all the files in a single dir. I think it can also import them, or you could add notes directly to the dir.
Over time, I’ve migrated most of my KB to some sort of markdown: Zim Wiki, DokuWiki, Joplin.
I still have some locked up in MediaWiki dumps, probably should spin one up and migrate it all.
It’s a mix of some aliases from across several distros, some from as far back as 4DOS and a part for Cygwin, a somewhat complex prompt colorizer that highlights remote/local and root/user shells, and some other stuff that’s piled up over the years.
I’m using Fish, rather than Bash, and it has type-ahead suggestions, which help a lot.
So, I’ll type rsync and then it’ll show inline that I typed rsync -ah --info=progress2 a long time ago. And then I’ll be like, oh wow, this past-me-guy was very smart, I’ll be having the same.
Obviously, this is an imperfect system. If you run another rsync command without these flags, it won’t directly show these flags next time, because it’s not the most recent entry in history.
But it’s rare that I know I’ll want to run a command again in a few months, so it’s still really helpful.
And of course, there is nothing stopping me from creating aliases and scripts as well.
I use nushell, same thing with the suggestions. With nushell, you can also press up/down to traverse through the command history for commands starting with what you typed. For example, you could type ls and press up a bunch to go through ls | where size > 2kb, ls | where type == ‘directory’, etc (if you’ve executed those before).
Same with fish. It actually shows anything if you just put in one part of the command, so you don’t need to specify the exact starting command (in case you might not remember).
Example (which I use regularly):
install, then up (and up and up, and so on), and I see everything I’ve ever used that has install somewhere in it.
So I’ll get results ranging from sudo apt install foo to sudo nala install foo to flatpak install foo.
operating_systems
Oldest
This magazine is from a federated server and may be incomplete. Browse more on the original instance.