I imagine the "update from another system" path runs in troubles with more complex gentoo installs than just the base system. For a full update from the live disk it will have to include lots and lots of (often exotic) tools that might be used in the building process (document generators like doxgen, lexer, testing frameworks, several build systems and make-likes. programming languages...) in addition to being able to build against the already installed updates for packages while not accidental building against packages that are not updated yet.

Or you go the simpler way and only do a base update from the live-system...only update the base build system and package management of the gentoo system and afterwards boot in a "broken" system in which only the basics works and rebuild it from there.

For be both those options sound less desirable than what is suggested in the blog.

Just to add...because xattr it's also "easy" (for some definition of easy ;)) to add/modify tags from the shell making mass-tagging feasible (as requested by OP). Second answer here should give a decent starting point....

Cautionary

I settled with German because finding English took way longer than I care to admit ;)

http://divajutta.com/doctormo/ubunchu/c.html

here you go ;) (well, not really a translation of what OP linked but the first chapters of the manga in....lots of different languages)

As much as I dislike snaps..it's not true that everything snaps do can also be done with flatpaks. snaps can be used for system updates (system libs, kernel, drivers...)..something flatpaks can't do (and are not meant to do). (Sorry, best link I can find in a short search...but it doesn't get in very many details :()

Isn't that what's the articel about? An alpine contributor calling out projects that "supposedly" target linux according to themselves but instead only target gnu/linux.

As gentoo user I can't argue with that... ;)

But I think there are reasons why someone would want to build suckless tools manually...namely that their configuration is mostly done in the source-code (Damn, it's so hard to not write anything too opinionated about suckless but I really try my best). But even then I agree with your other post that it's far better to use the distro facilities for building the the distro source packages just with your own patches applied.

Lets set aside my personal belief that suckless is a satire that too many people started to take seriously...

Always using the latest git version as done in the article doesn't strike me as the most sane thing to do if you "just" want to use the software especially as suckless offers version tarballs.

But suggesting sudo make clean install to build is really not okay...(and also not how the suckless tools I checked suggest it). You cloned (or better extracted the tarballs) as user...there is not a single reason to build the software as root. If you have to install then do it in two steps, build as user and only "make install" as root.

https://www.amazon.com/Computer-Programmer-Strong-Birthday-Relaxation/dp/B09P1CNRRM

"Distro recommendation" questions aren't usually very useful...all you get is everyone recommending the distro they use. It's unlikely you can get anything useful out of the answers.

I wanted something with support and with people that care for the code

Applies to pretty much every major linux distro that isn't a derivative and also some of the derivatives that do more than just add some cosmetics (unless you specifiy a bit more in details what you mean with "care for the code").

Also all distros can be configured, there is no real reason to switch from something like ubuntu because you don't like how the "Files" manager works to another distro...you could get pretty much the same on ubuntu as other distros offer and in most cases easier than by doing a reinstall. Really, you are better off trying to fix an issue you have on one distro that distro hop at every little problem you run into...

Yes, in a way that's pretty much how having multiple local terminals work as well. For ssh there is no difference there, you can have multiple logins with the same credentials even from different remote locations.

Might want to have a look at KDE's dolphin and baloo framework as it covers most of the tagging part. However where possible tags are saved as xattr directly in the filesystem so no need for a database for the tags making it easy to keep tags even while copying files to different locations. So it won't be a perfect fit for your case...but can still be good for "inspiration".

I guess a mixture of POSIX compatibility, backward compatibility and non-interactive shell use-cases.

Being somewhat POSIX compatible offers a way to write scripts that work on many systems independent of the actual shell implementation (bash, dash, zsh...). But this means major overhauls of the shell "language" are out of question...

Backward compatibility gets important for things that ignored the first point and used features only available in bash. Given that bash is the default for 30 years for linux now there are probably plenty of examples.

And while bash is not the smallest shell it is also not the largest one...and rather configurable at compile-time when it comes to supported features. This makes it a viable option as "shell-script" interpreter for systems that hardly have any interactive shell usage. It's not a completely bare-bone shell so you get a bit of "comfort" for scripts but you can remove unnecessary things like interactive command line editing with lib readline..I can imagine some embedded systems find uses for such a shell.

And it's not that there aren't alternatives...Microsoft's Powershell is probably the most successful one "recently". But changing all existing "workflows" from a text-based one to an object based one is not a trivial task...and in addition you run in new problems with any new shell design (For example I really dislike the overly verbose interactive useage of powershell but that's rather subjective)

The Splash screens seems to be written in QML (check /usr/share/plasma/look-and-feel/org.kde.breeze.desktop/contents/splash/)...so it should be possible to re-write them to display any information you want...including the journalclt output during the login what is probably the closest you get to a log of the KDE startup. Might need you to write a c++ plugin for QML though...no clue if there is already a QML module that can access journald logs.

(And of course you probably want to do that in a local copy in ~/.local/share/plasma/look-and-feel)

I assume the major reason is trying to get pre-compiled proprietary software to work...hardly any proprietary software offers a musl version in addition to their glibc version.

Ahm...the article is about how that package creates a broken system and that the alpine project doesn't want to be associated with images that have the package preinstalled. So not sure if suggesting to try it is a good idea ;)

As gentoo user running systemd...I think it's a bit oversimplified to say gentoo doesn't use systemd. Gentoo comes with profiles for systemd and openrc so I would say both are supported. And on a personal level I dislike the comparison to arch as both distro's hardly share any overlap in target audience (But I think the article compares them on the one thing they have in common...gui less install)

Also kind of misses one elephant in the room...openSuSE.

Played this one with Proton for some time already...and I love it. I like most of the additions to the engine but it stays definitively a X³ game. If you enjoyed Terran Conflict or Albion Prelude this one is a must (And free (as in cost) for owners of Albion prelude)

No clue if that fits your needs but there are plenty of latex templates for resumes