Will LTT Labs be making Linux packages benchmarks?

[Pm_me_nude_pc_parts]

In Linux there are so many different kind of packages and it'd be fun to have a list of what package manager is best for what applications. Obviously for a lot things you'll be using your inbuilt package manager, Apt, dnf, pacman etc.... But for Flatpaks, Snaps, and AppImages it'd be fun to have like launch speeds, performance, and general responsiveness benchmarked and clearly displayed on a website hence my question.

 

Benchmark it thrice on like Arch, Fedora, and whatever popular Ubuntu/Debian derivative like Pop_OS and for applications just pick the more commonly used ones. Personally I'd use this for deciding which package manager to use for each applications. Thoughts?

[Levent]

There is no such thing. You use the one that has the package you need with or the one you got. 

9 minutes ago, Pm_me_nude_pc_parts said:

what package manager is best for what applications

 

9 minutes ago, Pm_me_nude_pc_parts said:

launch speeds, performance, and general responsiveness benchmarked and clearly displayed on a website hence my question.

This is also pointless, it makes absolutely no sense to benchmark precompiled application binaries in that way. 
 

IMO none of the metrics you are asking here have of any value.

[Pm_me_nude_pc_parts]

9 minutes ago, Levent said:

There is no such thing. You use the one that has the package you need with or the one you got. 

 

This is also pointless, it makes absolutely no sense to benchmark precompiled application binaries in that way. 
 

IMO none of the metrics you are asking here have of any value.

 

So pictures of people benchmarking these things are just them lying

 

Gotcha bro

 

Also snaps are notorious for being slow because they have to decompress before launching or something like that.

[igormp]

FIrst of all:

 

18 minutes ago, Pm_me_nude_pc_parts said:

In Linux

LTT isn't even a linux-centric channel to begin with, so those results would be useless for most of the watchers. And even for the ones who care, LTT doesn't have much credibility when it comes to anything Linux or compute, it's mostly a gaming channel.

19 minutes ago, Pm_me_nude_pc_parts said:

what package manager is best for what applications.

As said before, there's no such thing.

19 minutes ago, Pm_me_nude_pc_parts said:

But for Flatpaks, Snaps, and AppImages it'd be fun to have like launch speeds, performance, and general responsiveness benchmarked and clearly displayed on a website hence my question.

Then it's not a matter of package managers anymore, but rather software sandboxing/containerization, and has many different variables.

20 minutes ago, Pm_me_nude_pc_parts said:

Benchmark it thrice on like Arch, Fedora, and whatever popular Ubuntu/Debian derivative like Pop_OS

That'd be a shit ton of tests already. Why not just take a look at openbenchmarking?

 

4 minutes ago, Pm_me_nude_pc_parts said:

 

So pictures of people benchmarking these things are just them lying

 

Gotcha bro

That has nothing to do with the package manager, but rather how the software was compiled.

 

25 minutes ago, Pm_me_nude_pc_parts said:

Personally I'd use this for deciding which package manager to use for each applications. Thoughts?

If you just want the fastest, install gentoo and set your build settings to always use -O3 -march=native with LTO and you'll beat anything else while having a binary that won't likely work anywhere else other than your machine.

[Guest]

Package manager is dependent on what distribution of Linux...  So is the format of the package..  And they aren't even necessary for any of them.   That screenshot is how long a Gimp render took, which is a graphics program like Photoshop but open source.

[Levent]

Just now, Pm_me_nude_pc_parts said:

 

So pictures of people benchmarking these things are just them lying

 

Gotcha bro

Passive aggressive eh, gotcha bro.

That data is built with binaries coming from different sources which can use different compiler options. 

[LAwLz]

5 hours ago, Pm_me_nude_pc_parts said:

 

So pictures of people benchmarking these things are just them lying

 

Gotcha bro

 

Also snaps are notorious for being slow because they have to decompress before launching or something like that.

People misunderstood you because those benchmarks aren't really testing the package managers. They are testing the package formats. 

 

Anyway, it seems like you already have a source for these tests so I'd just stick to that one.

Knowing LTT, they would probably fuck up their tests anyway without even knowing, and thus publishing incorrect data. It's an entertainment channel. Watch it if you think they make funny jokes and interesting videos, but I wouldn't recommend looking to them for information.