I don't know Mint. I know Debian. I've also used Arch, BTW. I adore that all I need is an ebuild, and the new package is managed like every other package on my rig. Or I can faff about and build a .DEB for myself, but that depends on having Debian packaging, which tends to be old. Arch has the AUR, which is a notion I hated even more, because AUR packages don't update like everything else. ebuilds are typically designed to pull either stable or bleeding-edge git. When an ebuild exists, it's usually in the overlay list. Ebuilds are also a lot more compact than Deb packaging and don't involve the arcane build structure that AUR packages do.
OK so first of all I initially wrote "can" where I meant "can't" and that's been edited. But now I just looked in my
~/src directory while running Linux Mint. I see only:
Probably not. On Gentoo you wouldn't need to worry about hunting down all the
somelib-dev packages so you can compile your thing from source, because they are not separated (on a source distribution, you always need them). So that part is a little bit easier. But you have to go through the rest of a Gentoo install/setup to get that advantage. It just depends on what your priorities are. 
The
/etc/portage/patches directory is really powerful for developers, and you can post patches in forums for others to try out before submitting updates.
That's a very routine task on Debian and Red Hat derivatives so I think I can handle it. At least historically, the case made for Gentoo has been highly bespoke binaries that match the host architecture very closely but I'm not aware that they're a lot more efficient than generic binaries. The single most powerful non-distributed supercomputer installation in this world, El capitán, is based on Red Hat Linux. Don't you think these sites would switch over to compiling everything from source if it made some huge difference? I think stuff like user privacy was mentioned earlier in the thread but you can do all of that on a largely binary-based distro.
OK so damn that's a real mouthful. I'm actually a little familiar with Gambas, which is meant to be a superior version of Visual Basic that runs on Linux. But there's still a lot going on here. Seems like you want to do both front-end and back-end here. That's probably the most crucial thing here. Gtk and Qt both can't be guaranteed. I know that wxWidgets for example can use Gtk and Qt alike. But it seems like you need a back-end too. That of course wouldn't use any set of graphical widgets. I guess the point is, what is it you want to achieve exactly? Let's hammer out all the details.
I mean, if you actually regularly feel you 'get behind' on packages with Linux Mint then a better option would be to move to Debian or Devuan 'testing'.
Practically speaking, they probably chose RHEL for better support for some shitty Oracle database that they're running for the NSA. And for anything that's actually going to take any real CPU time on those machines, it will be compiled from source with the same optimizations that Gentoo nerds can implement for everything that runs on their system.
I think the core of it is an app that I can use to run batch file renames on my server, either through ssh or a web interface but preferably ssh as it's more secure and this app would have deep filesystem access. It's primary function would be to be able to take a list of files, and be able to run a bunch of conversions to the filenames (and paths) and export that to a to-from list which can be run through by calling mv for each line of the list. That way the command can be run at the user level for the preparing of the list, with sudo (if needed for files the user doesn't have permissions for) only being called for running the file renaming, after the user has a chance to preview the list.
Are you mainly aiming to use this to rename audio files, for which a number of X11 applications exist for? Like Quod Libet/Beets/EasyTag? Or for video files+audio, FileBot?
Gentoo is also following this decision. Difference is that non-systemd profiles continue to have the option of keeping split-usr without a forced transition, but all new stage3's are merged-usr. The problem with fighting that decision is that systemd made it mandatory while split-usr hasn't had a strong reason to exist other than momentum. I still continue to use split-usr as long as its viable, but I understand that it might become unsupported the next time there is a profile upgrade (~6 years).
Not ideal, but this just means toralf needs to hold the torch. It has been a point of contention that ago as chosen to keep his scripts private. So any new tinderbox runners will most likely base their own setups on toralfs (and before him flameeyes's) scripts as he keeps them public.
Gentoo binaries have an actual likelyhood of being less optimized nowadays. Binary distributions have been already been building their packages with cpu agnostic optimizations like lto and pgo, while glibc-hwcaps has been a somewhat recent addition that would get most of the the -march= benefit on newer cpus. On Gentoo it all depends on the end user configuring their cflags to be optimal as the default on gentoo is just -O2 -pipe.
It could do audio files, but it's generally intended to be pretty universal. And it would be used solely for local ssh logins.
Okay so I sort of figured out what is causing this - the colors are being replaced by something else. I tested this by setting the defaultbg to green, now, when I open ST, the background is green for less than a second, then it turns back. What could be doing this?
Does this happen with xterm/uxterm? Does it happen with eterm? If not, then st is doing something weird. If yes, then pywal16 is not operating how you would expect.
Yep, just tested it. I change the colors using
xterm -fg black -bg whitexbps-query -Rs xresources
# or
xbps-query -Rs resources
LTO is very easy to turn on in Gentoo, and I've yet to find a situation where it breaks anything.