So essentially people are abandoning the memory/speed efficiency of the .so ecosystem, and seeking exe/msi style convenience... You know... a dump of legacy dll/static-so-snapshot versions with endless CVEs no one will ever be able to completely fix or verify.
Should be fun, and the popcorn is waiting =3
They also gain substantial amount of security by being sandboxed by default unlike majority of native packages.
1. Current release applications on deprecated OS (Mostly good)
2. Deprecated applications on current OS (Mostly bad)
The Windows style packaging architecture introduces more problems than it solves. Fine for running something like Steam games with single shot application instances using 95% of system resources each power cycle, but folks could also just stick with Windows 11 if convenience and security-theater is their preference.
Some people probably won't notice the issues, but depends what they do. Arch Linux itself is a pretty awesome distro for lean systems. =3
Source? There is no measurable energy or efficiency difference at least for flatpak on any semi recent hardware. I know that snaps do take couple seconds longer at first start.
I prefer flatpaks for proprietary and internet facing applications because of there easy sandboxing capabilities. There is also the advantage on archlinux not needing to do a full system update for a single application.
https://tldp.org/HOWTO/Program-Library-HOWTO/shared-librarie...
Getting into why the community argued for years while Debian brought up deb version controlled packaging is a long dramatic conversation. Some people liked their tar ball mystery binaries, and the .so library trend started more as a contest to see how much people could squeeze out of a resource constrained machine.
In a single unique application running context, the power of a cached .so reference count are less relevant. As a program built with .so may re-use many resources other programs or itself likely already loaded.
> ldd --verbose /usr/bin/bash
> ldd --verbose /usr/bin/cat
Containerization or sand-boxing is practically meaningless when punching holes for GPU, Network, media and HMI devices. Best of luck =3
Many applications don't need these permissions and even the ones that do will be much more secure than having full user space access by default.
Someone could exploit the system to gain more access vs someone does not need to do anything because they have full access by default. It's like arguing you don't need a root password because sudo is insecure anyway.
Qubes, Gentoo, and FreeBSD are all a better place to start if you are interested in this sort of research. Best of luck =3
But thanks for trying to post actual relevant data on the topic. =3
"Popcorn Music Video" (The Muppets)
An outdated old package library relies on people understanding/tracking the complete OS scope of dependencies, and that is infeasible for a small team.
If someone wants in... they will get in eventually... but faster on a NERF'd Arch install. =3
That is exactly the strong point of flatpaks. It's a lot easier to use toggle in a GUI for permissions than write whole new profiles. Not to mention that many even disable selinux because it is difficult.
>An outdated old package library relies on people understanding/tracking the complete OS
It takes 0 understanding to copy paste a outdated package warning and report that to the repo listed in flathub. It explicitly tells you as much.
Some programs take a huge memory and performance hit on non-Linux machines. =3
You're implying without stating it (or providing any evidence) that programs perform worse when statically linked than when assembled out of ELF DSOs, even when each of those DSOs has a single user.
That makes no technical sense. Perhaps you meant to make a different point?
A 34MB static built version will cost that amount of i/o every single instance on a system that did not cache that specific program previously. Also it will take up that full amount of ram while loaded every single time it runs.
Inefficient design, but works fine for other less performant OS =3
I've forgotten how to count that low.
Also, static programs are demand paged like anything else. Files aren't loaded as monoliths in either case. Plus, static linking enables better dead code elimination and devirtualization than is possible with an AOT-compiled and dynamically linked setup, which usually more than makes up for the text segments of shared dependencies having been pre-loaded.
I'm not sure you have enough technical depth to make confident assertions about linking and loading performance.
> =3
The "blowing smoke" emoticon isn't helping your argument.
If .so reuse is low, or the code is terrible... it won't matter much. Best of luck =3