Slackware 15 Beta Process Begins

[jacob]

Still exactly the same:
- It is the most unix like linux distro, where everything is as simple as possible and with less layers. Only on this 15 release slackware will finally add pam support, not because it really needs it, but more software requires it. Gentoo and debian have many layers to allow then to be flexible, but that also makes everything more complex
- it just works! either server or desktop, everything included works and was well tested. If i need to compile anything, everything i need is already there. Gentoo is too complex and with too many software combinations that makes things fail. Debian also works well, but if you want to compile something, you will need to hunt for many packages to be able to get everything ready to compile things
- small , fast and clean. As no extra layers are added and do not aim to have all packages, slackware installs are small and simply enough to understand
- KISS! keep it simple, stupid! Neither debian or even worse, gentoo are simple. Try to understand how everything works on those systems and you will have a hard time
- excellent disto to learn about linux as you can follow how things are connected and very few "black magic" is needed, unlike gentoo and debian. Even packages are simple tar.gz. No dependency check, if you fail to install a dependency, you will learn that package X needs package Y too. Not to say that learning is simple, but on each failure you will learn lot more than using other distros. For learning, the only distro that surprass slackware is LFS (linux from scratch)
- while 14.2 is old already due to the age, 15 have up to date packages... that work!. No need to choose a stable but outdated software distro (debian) or a unstable bleeding edge distro (gentoo)

in recent years you also get this one:
- systemd free! slackware uses still simple BSD scripts to boot and session management and likes use elogind
- pure arch or multilib. officially slackware is either 32bit or 64bit, but you can make it multilib by replacing the glibc... so you choose if you want a pure system or a multilib one
- while many people want other software not included in the distro, external slackbuilds allow one to quickly compile most of those software, similar with gentoo... but you also have external repos, namely the biggest one, Alienbob, with most of those packages already build. Yes, steam and games work perfectly in slackware

I don't deny that Slackware has its pros and its place but I'm not sure about the learning part. If learning how stuff works is someone's goal, then with Slackware they will learn how Slackware works, not how a modern mainstream distro works (one that has PAM, systemd, cgroups, advanced package management etc...). To tinker with that, Arch might be a better choice.

[Nille_kungen]

Similar story here... Was running 14.2 until late last year, I finally got the point where I couldn't wait for 15.0 anymore which seemed to be eternally "coming soon". Even now I'll be surprised if it's really as "coming soon" as we're talking about here. I do hope I'm wrong though and that it ends up being a great release (ideally with another one coming in < 5 years) because Slackware is a joy to use in all its simplicity

It's Slackware so don't expect a release to soon, it was 6month between beta and release for 14.2
The one thing we know about Slackware is that it is released when it's ready.

[lucrus]

It's dead JIm.

[swagg_boi]

If learning how stuff works is someone's goal, then with Slackware they will learn how Slackware works, not how a modern mainstream distro works (one that has PAM, systemd, cgroups, advanced package management etc...).

Disagree 100%. For example that you gave, "advanced package management" is more difficult if you don't understand the basics (SlackBuilds build scripts, creating your own slackpkg+ mirror to deploy them, GPG keys for signing the packages, etc.) systemd can be a lot to digest if you have no concept of /sbin/init; you may not get why the service supervision is a big deal if you've never dealt with basic init scripts. And 15.0 will include PAM! (finally...)

[jacob]

Disagree 100%. For example that you gave, "advanced package management" is more difficult if you don't understand the basics (SlackBuilds build scripts, creating your own slackpkg+ mirror to deploy them, GPG keys for signing the packages, etc.) systemd can be a lot to digest if you have no concept of /sbin/init; you may not get why the service supervision is a big deal if you've never dealt with basic init scripts. And 15.0 will include PAM! (finally...)

I guess we'll just agree to disagree about that. I really don't see the point of learning about old-style /bin/init on a systemd distro, it's a little bit like saying that you should be learning cmd.exe before you start developing with powershell. I really don't want to get into another pro/against systemd discussion, but I think everyone can agree that systemd is not an extension of, or front-end to init but something different, and knowledge relating to init doesn't apply to it (and vice versa). For the packages, the hard (and important) part is dependencies, build-dependencies, multiarch, reproducible builds, packaging policies etc. An aspiring package maintainer doesn't get exposed to these concepts on Slackware.

Once again, that's not diminishing the merits of Slackware. Its users are obviously happy with it and I can see use-cases in which it can be very useful. But the way I see it, it's no more suitable as an entry into the Linux ecosystem than MS-DOS is into modern-day Windows or K&R C is to Android.

[swagg_boi]

systemd has been around since only 2014 my friend... The other "gaps" you're pointing out (K&R C to Android??) are much more significant I think. Fine to disagree, just don't want folks out here thinking I'm suggesting to learn COBOL before trying to understand Rust or something

[andyprough]

I don't view release versions of Slackware to be all that important for my uses. I'd stick with current anyway, which is a rolling release.

[fulalas]

I don't view release versions of Slackware to be all that important for my uses. I'd stick with current anyway, which is a rolling release.

Exactly! Also, all this excitement with the latest version of everything is kind of pointless.

[higuita]

I don't deny that Slackware has its pros and its place but I'm not sure about the learning part. If learning how stuff works is someone's goal, then with Slackware they will learn how Slackware works, not how a modern mainstream distro works (one that has PAM, systemd, cgroups, advanced package management etc...). To tinker with that, Arch might be a better choice.

Notice that i do not claim you will learn ALL the linux, of course if you just use one distro (whatever it is) you will miss many features used in other distros. Slackware is good to learn more low level linux than most of the other distros. I interview many people and it is amazing how many linux people do not even know how to manually configure a network interface or even what ss command is. Why? because all that is taken care by higher level features and many people do not even touch those.

[jacob]

Notice that i do not claim you will learn ALL the linux, of course if you just use one distro (whatever it is) you will miss many features used in other distros. Slackware is good to learn more low level linux than most of the other distros. I interview many people and it is amazing how many linux people do not even know how to manually configure a network interface or even what ss command is. Why? because all that is taken care by higher level features and many people do not even touch those.

The point I was trying to make is that most people don't actually need to manually configure network interfaces or use ss or they can worry about it on the one occasion when they actually need it. Learning GNU coreutils is not a goal by itself, people's goal is to get productive stuff done. And, where to draw the line? If you say that people should start by learning to use the "ip" command manually (or even the old "ifconfig"...) then why stay at the shell level? Why not start with learning the syscall, or rather the "sysenter" assembly instruction etc...?

The fact is that for better or worse, the linux distros' mainstream is systemd, and it's not something sitting on top of the classic scripts/tools, but in their stead. For most people coming from Windows and especially MacOS, systemd will feel a lot more familiar than the classic *nix environment. Besides, from a learning prospective, when dealing with prod systems (talking server here, because that's where you may have to do that), people don't won't be using Slackware but rather with RHEL, SLES, Ubuntu Server etc. There is not much point in pointing them towards Slackware only to tell them later "ok now forget it, because here it's all different". That's just pointless confusion, so Slackware is not IMHO a good way to get started in "Linux".

But by the same token I think it's worth recommending Slackware to those who are actually interested in unix. It has several advantages over the ***BSDs as a system to learn on, not least because it has better hardware support and a greater software library - someone posted here that you can even run Steam on Slackware, although I'm not sure about snap, flatpak and similar things.