this thread idea comes from here and here.
The problem (to me):
The problem (to me, better explained):Too much information could be like no information at all. And too many choices (and the resulting indecision) could be worse than a single path.
- (Trying to) learn about Linux (no matter the details) about 25 years ago was frustrating: little "non-guru" documentation, hard to find the right contacts, long time for answers, misunderstandings and flames, expensive import books (not covering all of your problems), too much technical writing and sysadmins repeatedly saying like RTFM...
- (Trying to) learn about Linux today is... really different! A lot of quick and flashing input (for everything!) coming from everybody and everywhere... and "the way of doing" is constantly changing...
20 years ago somebody "told" me about GNOME and KDE; luckily, they're still here (not the "then-promising" and sponsorized Unity!), and, althought quite different nowadays, having leadned how to use them at the time really helps today (so it loooks like it was a good investment...).
But... I'd also like to "learn" a TWM. Which could be a good choice? (general thought, no need for actual answer now). In the last three years I've heard (mostly) about DWM, XMonad and LeftWM. Yet I'm not able to detect if they're a come-and-go "fashion" of this moment or if they're "here to stay" and so "invest" time/resource in some on them (also because it looks like tiling is now being "forced" also in "classic" DE(s), now)...
So below I describe the "only" rolling "ongoing" studying path(s), involving some "parallel roads", that currently seems to "work well" with me as of now (As a background, I use Debian SID via CLI only at work everyday, but unfortunately at work you don't always have the chance to deepen the knowledge of the topics you're interested in...)To me, unfortunately, "make your own choices and do what you want" is an approach that doesn't work (at least, without a "little" pre-knlowledge of the matter).
I've performed several installations by hand/CLI following the handbook and everything works very well; once only partitiong NVMe failed through archinstall, but I solved it by manually partitioning "The Hand Way".Everyday use and best distro: Arch
Create (from scratch, with pacman "and friends" tools) a custom ISO live XFCE (vanilla appearance), with *fdisk/, Vim, rsync, arch-chroot, GParted and some other recovery tools, maybe also afterwards-installable (through pacstrap), although not updated (maybe "refreshed" from scratch once per year...).Personal interest study/fun project (not ongoing, no time limit): Archiso
Secondary goal: live recovery and/or local installation with no Internet at all, through a previous created offline local repository. For the same reason, it would also be nice to include custom config and scripts inside the ISO, instead of git-cloning them (maybe a little bit updated, but better than nothing...).
(this would be a toy/experiment: of course for "serious" solutions, I'd go after e.g. Knoppix, Rescatux, Gparted, SystemRescue and so on...)
Arch, to me, really "wins" over Gentoo given the (almost) "no time at all" time required to compile of the first over the second. Gentoo, anyway, has a wonderful wiki, useful when you don't find something on the Arch one.Personal interest study/fun for greater knowledge, about kernel and compiling (no time limit at all): LFS (wonderful) and Gentoo (worse).
LFS (of course, not suited for everyday use), in particular, provides/allows for very greater knowledge about the compilation process (not only of the kernel). It takes forever, but you do it when you can, not expecting anything (hoping for the best but expecting the worst), "backfalling" on the reliable Arch for the everyday use.
As suggested by Erik, you "start" with a DE (on a secondary PC and/or VM), apply your custom configuration, then "isolate them" as (maybe online) personal package(s) that you can later deploy on the other "everyday-use" Arch system.Personal interest study/fun (less important project than the others above) for config/ricing: Arco
This is my conclusion (as of now): the "main killer application" of Arco is, to me, starting with an "already setup" DE/WM of your interest, in order to study and try it, ready out-of-the-box and (maybe), if interested, rice it. In a "create and destroy" ongoing approach.
Also, I deeply thank Erik not only for sharing its knowledge, but especially for sharing his "way of doing" (Meld-Sublime-Git...) repeated with energy for every task in every video... you have to watch a lot of videos to get acquainted with this interesting "approach"... maybe the "approach" itself is Arco's "secondary killer appication".
Unfortunately, I still find Arco documentation/information really sparse and decentralized: this vast knowledge expands in every direction (but maybe that's a problem of my way of learning things!), still don't understand it very well about nemesis, skel, etc., and, in my opinion, being a "one man only" project (despite really notable!) doesn't help its cause...
So these are my thoughts as of now... if somebody has different visions about the aforementioned distros and projects and/or have other suggestions about the "learning and doing approaches" (and/or more "proper" distros for specific needs), please propose and share ideas!