Schaller On Linux In 2018: Rust Rules, Apple Declines, Linux Graphics Compete

[L_A_G]

Man is he being wildly optimistic with many of these predictions.

Considering the amount of work related to moving from GNU Make to Meson I really don't think Meson is going to become the standard or even close to it. Sure, some heavily cross-platform projects are going to be adopting it and so are new projects, but considering how heavily most big projects are built around GNU Make, I wouldn't be surprised if GNU Make is going to be the go-to build software for at least another decade.

Rust supposedly being about to take off as a C/C++ replacement has become a running gag very similar to the "Year of the Linux desktop" that has supposedly been just around the corner since the late 90s. The reason why Rust was created in the first place was due to how slow progress had become in C/C++ and now that it's sped up considerably Rust is becoming less relevant with every new version of the languages. Sure, people who went ahead and learned it will refuse to admit they ultimately wasted their time and keep signing it's praises for years to come. However like the people who kept singing the Amiga's praises well after it had gone, they're not really going to change anything, except add gunk to projects that needs to be re-written for complexity reasons a decade or two down the line.

As for the graphics prediction, there's really no way Aspyr and Feral's games are going to perform on par with their Windows counterparts when so many of them use translation software rather than being native ports. Translation layers, no matter how "thin" they can be made using Vulkan, are going to cause an inevitable performance penalty.

Finally, Apple declining as a PC maker is also kind of questionable seeing whenever the market is going downward they tend to decline less or stay stable, when the market is stable they gain and when the market is growing they grow faster. The main threat to their PC business, in which the most successful products are their laptops, are the tablets and smartphones, in which they have a considerably bigger market share than in PCs.

[tildearrow]

What do you think of Christian's predictions?

Meson becomes the defacto build system in the Linux community.

Possibly true for Autotools projects. The only thing that annoys me a little is that Meson takes ~100ms to output something (CMake does it instantly).
I'm not sure if it is faster than CMake with Ninja though (at least in one project, RazerGenie, CMake is faster)

- Rust puts itself on a trajector to replace C/C++ for low-level programming.

Please, no. I haven't heard of any uses of it in Windows.
Also, why is Mozilla ignoring Linux (not enabling OpenGL in Firefox by default), but distracting us with Rust?

- Apple declines as a PC vendor.

I doubt this will happen. Most of their customers are happy with the macOS platform, and doing this will only be hurtful for Apple.

- Traditional Linux distribution packaging for desktop applications will start fading in favor of Flatpak.

This would be great.

- Linux graphics will become competitive across the board. Though part of this he thinks that Aspyr/Feral Linux game ports by the end of 2018 will begin performing as well as the Windows games.

True, as long as Feral decides to put more effort in their wrappers.
Also, one major problem for Linux is NVIDIA, which sadly is more popular in the Windows world.

- The H.265 format will be considred a failure and moving forward all new codecs will be open-source / royalty free.

There is a typo. I highlighted it.
I really doubt this. H.265 is successful in the proprietary world (plus it allows for DRM, which we all hate but companies love, which will keep it alive).

[przemoli]

i don't understand in which century rust will put itself on trajectory of replacing c++ in its own compiler by rewriting llvm

what c++ app including the linux kernel were rustified during first decade of rust existence?

Rust was released in 2012 so we are only 5y into it's first decade.
As to what apps were rustified so far I would point out Firefox engine. Gecko is being replaced with Servo (or components researched under Servo umbrella). But that's to be expected. Rust was created as better tool for Firefox development after all

What next 5y will bring? I'm not sure.
One problem Rust have it's competition in form of Go, which can do similar feats. That means some will choose Google over Mozilla instead Go vs Rust. Some will go with bigger ecosystem (Go). And most certainly cool factor is on Google side

Still Rust looks to have gained good recognition. There is trend towards more functional programming, and Rust solve parallelism, and that's a hot topic for the next 5 years.

[przemoli]

[QUOTE=tildearrow;n995829]
Also, why is Mozilla ignoring Linux (not enabling OpenGL in Firefox by default), but distracting us with Rust?
/QUOTE]

Parallelism is hard. Web browser such as Firefox really need parallelism. Thus follow that any improvement to tools to manage parallelism is of great benefit to Firefox.

[starshipeleven]

Also, why is Mozilla ignoring Linux (not enabling OpenGL in Firefox by default), but distracting us with Rust?

Because the switch to their new engine written in Rust isn't benefiting also Linux?

Really, they do what is the best way to spend their resources, and improving only their linux part is (far) less important than a cross-platform change like Firefox Quantum.

[starshipeleven]

ah come the fuck on, vbullettin blocked a post for tildearrow above this.

[Michael_S]

A lot of the value of a programming language is not purely what it can do. Instead, it's what it makes default.

I know Rust is riding high on the hype cycle and lots of people are sick of it.

But the enormous advantage of Rust over C++ is that the borrow-checker validates reference ownership in the compiler and you have to use rarely used and not recommended escapes to get around it. Plus, variables are immutable by default. With C++ you have to remember to use shared_ptr and unique_ptr and even if you do remember a mistake leads to a runtime error and not a compilation error (which is what the borrow checker in Rust provides).

Rust isn't going into the Linux kernel. It may never get into LLVM, either. But it is growing, because it's arguably the safest player in the fast-as-C++ space.

[leipero]

Traditional Linux distribution packaging for desktop applications will start fading in favor of Flatpak.

Not so sure about that, as far as I know flatpack still did not solved the problem package managers did = repositories. You only realize how useful repositories are a while of using GNU/Linux and messing with Windows. Package managers are really good solution for that, just search with it, you don't even need to know the name of the package.

[andreano]

Rust & Meson adoption:

This could be interpreted in 2 ways, and it's obvious to me that Christian meant the second: Not adoption as in rewrite (i.e. replace) what already exists and works fine, but adoption as in write new stuff in it. Old code is only rewritten if it's actively worked on. That's how programming languages are adopted – people are lazy after all.

He's not saying that Redox OS will replace Linux in 2018. He's saying that new projects of 2018 will use Rust or Meson.

Corollary: Popularity percentages is not an indication of adoption. Derivatives are.

[Leopard]

Not so sure about that, as far as I know flatpack still did not solved the problem package managers did = repositories. You only realize how useful repositories are a while of using GNU/Linux and messing with Windows. Package managers are really good solution for that, just search with it, you don't even need to know the name of the package.

Flathub is available. On Linux Mint 18.3 , software center has Flatpak support and it is linked to Flathub.