hagar

It's a thing!

I think you have a point there, but the reasons why Mint does not ship a streamlined version may be simply because the maintainers don't want to bother with a whole different context to build, document and support.

I do think there would be value in a less "batteries included" Mint. I disagree with people in this thread who claim the "whole purpose" of Mint is all the stuff it packs, because it goes far beyond the essentials. Mint develops a lot of GUIs for the user to be able to configure the system. I think just these plus the in-house Mint core apps would make for a sweet, lightweight and less bloated system that would have real appeal, but that would also mean more work for the Linux Mint team and perhaps it wouldn't really mean much for their audience.

Your mileage may vary for performance. It really depends what OS and what hardware. In my experience saying all BSDs are slower at rendering would be too broad a statement.

If you've done Arch and Debian server installs, you'll be fine installing a major BSD. Just answer prompts and you are done, particularly if you are using the default disk partitioning scheme. Consider NetBSD. It's known for its wide hardware compatibility. X is pre-installed, just "startx".

Consider antiX. It's very lightweight, supports 32 bit and you'll have access to the Debian Repos.

Recently did something similar and yeah it seems Mint, specially LMDE in my case, is a great fit for such cases. It's on that sweet spot between being too bare and too bloated.

I'm not the biggest fan of VBox either, it's just very popular and full of sequential "wizards" to guide the user along the process of creating VMs, so it might be one way to get started. I'd much rather work with QEMU though.

That might be fun then.

QEMU can be as simple as this:

qemu-img create -f qcow2 mydisk.qcow2 20G

Here you are first creating a disk image with the format qcow2 and maximum 20G capacity. This is a QEMU disk image format that will take up very little space and grow as you use up the VM disk.

qemu-system-x86_64 -m 256M -cdrom alpine.iso mydisk.qcow2

This will start a VM with 256MB of RAM, the alpine.iso image in its virtual CD/DVD slot, and the disk image you just created as a virtual drive. This will come with networking enabled by default, so you'll have internet access from within the VM.

It should now drop you into the Alpine installation. Alpine is very lightweight so it's great for experimenting, but you could do virtually the exact same for most other flavors of Linux and BSD images out there.

Once you are done installing, you can power off the VM and then start it with this:

qemu-system-x86_64 -m 2G mydisk.qcow2

That's basically the same without the -cdrom argument, this time with 2GB of RAM. I find QEMU a delight to play with because it has sane defaults like that. Hope you have fun too!

I think you might like DIstroSea. If you'd like to persist your experiments, then likely learning how to emulate systems with QEMU or VirtualBox (the latter if you'd like a friendlier GUI-led experience, the former if you want to go full-CLI virtualization). QEMU is great in how lightweight and easy to create and discard self-contained VM disk images can be.

What I mean is that no one will stop you. When you ascertain your own right to do it, it doesn't mean much that I don't believe you are entitled to it. It's pretty much common practice. That is more a semantic matter at this point, but yes I stand by that being messed up for a project the size of Nix.

I don’t think that being a dictator for your project is necessarily “toxic”

Yeah, it is not necessarily toxic. It is at a lot more risk of being, though. Even a collectively managed project will mess up and upset the community, but then there is a sense of shared responsibility and more deliberation on what to do. With a BDFL, it's just whatever. After your project reaches a certain size, that risk keeps increasing... exponentially.

I have projects that take contributions and I work on others that do not

Precisely. You see, if we take this into the context of a smaller project, specially one managed by a single person as you seem to be coming back to, that is a very different context. I don't think an OSS maintainer should be laboring physically and emotionally to meet the demands of users. That is a well-known problem there. If this person doesn't even want to have contact with the community and just ship code once an year, fine. They are just sharing things with the world at no cost. In this context, "suck it up and just fork it" is indeed the way to go.

When you take something as big as NixOS though, that can really be inverted. Now you have a very large number of people who are laboring physically and emotionally to sustain a very large project, and the original creator shifts to a very different place to. It's another discussion entirely.

If pursuing your own vision is the sole purpose intended, it would not be limited at all.

I guess you can, yeah.

My point is not that you can't. You clearly can. And many do. The thing is, when you create your foundation that "you fill with whoever you see fit", when you faithfully believe that the BDFL will "stop them doing stupid things", or that you get to choose your board members arbitrarily and tell everyone it's not a democracy like you are proud of running it as a dictatorship, that's just a incredibly narrow and toxic culture you have set up. It's not impossible. The ethic you are posing is actually quite widespread in the world I live in, anyone arguing for it will get many around to agree, it's very assertive and rightful. Still, a shitty choice the way I see it. And from this bleak outset of things, I suppose forking is indeed the only option you have.

I guess it can be simple like that when you are the maintainer. It is definetly not as simple when there are many of them. Of course you can run it like that and many do, but the whole mentality is pretty limited.

My statement is not that you have to do whatever anyone asks in your project that you maintain. My statement is that a community that contributes towards a project has a say in it. You might want to ignore it, handle it BDFL-style, politely and cynically decline, whatever.

Not really about what is the absolute correct answer. Our values are clearly different. More like what I believe works best in the long term.