And, if something like that has been said, there is not marketshare, like marketshare. There may be OS marketshare of DAW users, or marketshare of using profesional interfaces....

These details are ultimately secondary. I don’t recall any such statement—it would have stood out to me. I assume it was never made in this form. The crucial point is, if Linux support were possible, it would already exist. Everything else remains speculation and wishful thinking. Hoping is fine, but nothing can be forced; it also has to fit for RME as they need to develop and support it.

The article linked by the OP does not tell anything about the Linux users.
Is it a webserver computer in a bank, a computer in a factory for steering a machine or is it a homecomputer of a private person working on multimedia stuff ?

Relevant would be how many homerecording people want to use Linux, and there is no measurement.

According to that same source it's back down below 5% in the US at 3.49% in October.
https://gs.statcounter.com/os-market-sh … f-america#

What if...
https://youtu.be/SLKtWi6NxmI?si=_dZIZ9b1fq2CzHY7

In complex-systems and technology-diffusion research, major transitions are often preceded by very small early-stage indicators. These signals usually look insignificant in real time, but become clear only in hindsight.

Well-studied examples include:

Soviet Union (1991):
Economic decline, regional pressure and reduced state capacity were measurable years before the collapse, even though most observers expected long-term stability.

Berlin Wall (1989):
Data already showed rising emigration pressure and sustained protests prior to the political breakthrough, but these patterns were not widely interpreted as a structural shift until afterwards.

Internet adoption (1990s):
Early usage numbers seemed marginal. Later analysis (e.g., Bass diffusion models) showed that disruptive technologies typically follow slow-start S-curves, where initial growth appears statistically small.

Smartphone penetration (late 2000s):
Early percentages were low, yet developer activity, standards formation and acceleration rates already indicated an upcoming transition.

Across these cases, the consistent finding is that macroscale change begins with microscale signals. Small percentages do not rule out larger future shifts, especially in technology ecosystems.

This doesn’t imply that any specific outcome is guaranteed. It simply means that early indicators should be interpreted within the broader context of how technological transitions normally develop.

Soviet Union and Berlin Wall are not relevant examples as they are examples from theory of chaos and theory of catastrophs, not applicable to product penetrating market....

You name succesfull stories of penetration... For each of them one may name tens, hundreds of those that have failed...

Linux has been here for almost 35 years... It has its place, it is gaining its marketshare. Maybe that people will leave Windows because of reasons.... But what if they switch to Mac, instead of Linux?

But still, my guess is Linux will gradualy gain marketshare, but not substantially. Why should people switch? They do not care much, which OS. If computers with preinstalled Windows will be sold, they will use those. Most of them.

If one has computer with Win, he will stay on Win till computer is obsolete. And when he buys a new computer, it may be Win again or Mac or maybe Linux.

There may be few individuals that uninstall Win and install Linux because they dislike Win. Or make dualboot computer with Win and Linux, but it will be tiny fraction of people. If one wants Linux, buys computer without OS and installs Linux or buys computer with Linux preinstaled, does not buy computer with Win and then reinstalls everything....

One may stay on Win10 or even Win7. I have moved from Win7 to Win10 two years ago. I still use WinXP. One may wait for Win12, maybe it will be better than Win11, maybe not.

Disagree

The extra cost is the time you have to spend for learning and collecting software.
A Mac has all in, is ready to use for private multimedia work. With Linux you have nothing.
But it's alternative. you are so powerful you have tricked the big players, you have tricked the system, everything alternative must be better...?

Are you worried or just surprised?

I see it differently: although Linux has only about 5% of the desktop market, its influence and potential are steadily growing in servers and the cloud, through the open-source ecosystem, and among younger users. Small niche areas can trigger a snowball effect, and market growth is not always linear.

Sure, one may expect exponential growth at first. Be it 10% a year, which is a lot, Linux would reach 10% in 2032. Be it rise 7% a year, still pretty much, Linux would reach 10% in 2035.

Look at yearly numbers of Linux marketshare and one may calculate, what the growth rate is. And then one may extrapolate it to future.

And do not forget, that the world has not been so close to atomic war or at least world war since 60 years ago. So one may have more serious things to deal with soon than Linux vs. Windows vs. MacOS.

Why should a company invest 50+% more resources to enlarge possible marketshare by 5% points? Maybe it is better to invest 25% more to current market and gain  more marketshare within 95% of market against competitors. Like making new versions of TotalMix and DigiCheck or making three versions of each driver for Mac, because Apple... Or making ARM driver for Win.

I do not see many chances for Linux real penetration to desktop anytime soon... Unless big tech gets involved and makes something like Android did. If there is one widespread distribution of Linux, it could get focus of developers.

Just imagine the nightmare of RME support if they support Linux as whole with so many variable distributions. Customer calls it does not work. Hmm, and what kernel, what distribution, what HW do you have? And customer, you know, I made my own build and has tweaked it little bit. And so on. It is no go...

One cannot reliably support can of worms. Widows with its diverse HW and Apple with its security everything is enought to make wrinkles. No need to add few more for marginal benefit.

I understand Linux users, that they would love to have RME support. That is natural desire. But the reality is different. I bet RME is getting been prepared for Linux, when right time comes....

Why and why. I’m not demanding anything. I’m just telling where the world is heading. You can always just stay lying still.

The fact that you’re reading the Linux section of this forum says something — even if you think it’s “bad” or “pointless.” Curiosity or awareness seems to sneak past your skepticism.

Just to clarify: the video in post #11 wasn’t fictional — it was interpreted as such. That’s fine, everyone reads things differently. But that interpretation doesn’t actually say anything about the points I raised here.

Labeling something as “fantasy” is an easy way to sidestep the substance of the discussion, but it doesn’t engage with any of the arguments about adoption patterns, industry shifts or long-term trends. I’m completely open to corrections if any of my points are factually wrong — but simply comparing them to a misinterpreted video doesn’t move the conversation forward.

Peace

Sorry, Mike, I do not want to hurt your feelings, but I think you seem to miss some important points.

Linux offers too many distributions, kernels, setups, and maintenance models, creating a level of fragmentation that makes life difficult for both end users and vendors, and prevents an attractive ecosystem for commercial applications.

As a result, there is no single system for all types of desktop applications, such as office, gaming, audio, video, and other areas. This makes it hard - even for experts - to choose the best compromise out of 800+ distributions and then customize everything that is still missing.

If you later discover that it was not the right choice of distribution, you will have to customize everything again, and because the systems can be very different, you may find that the customization steps are not the same, creating even more work for you.

For vendors, the situation is even worse, because they also have to support whatever system the customer chose—including whatever the customer customized on their own.

And since Unix systems, in particular Linux systems, can differ so much, finding the root cause of a problem becomes significantly harder when the list of potential issues is far longer than on the more standardized Windows or macOS platforms, where there is only one system and not hundreds of variations.

This explains why Linux never achieved the unified platform quality of Windows or macOS and remains a burden for everyone involved when it comes to commercial or professional support. It becomes even more critical if SLAs are involved, which is why larger companies prefer to develop and maintain their own Linux distributions. For example, Red Hat and Oracle do this for the enterprise, service provider, and data center markets. They cannot build their business and support model on a random release that is not under their control.

Use Linux as a server or for reading mail, social media, scripting, or development if you want—fine—but for real desktop systems with commercial applications, you need a standardized platform as a solid foundation for development, support, and ultimately for the customer, who wants the OS to be easy to use and maintain.

As Kubrak already mentioned, Linux has been on the market for a long time but still offers far fewer commercial applications compared to Windows and macOS. There is a reason for this: too much diversity creates too much extra effort on all sides.

P.S.: I am customer like you, this is only my $0.02 worth on this topic (based on my experience of the last 30+ years).

P.S. 2: After decades of chaotic proliferation, it is long overdue to take stock and guide Linux toward becoming a genuine alternative to macOS and Windows. What is currently lacking is a consolidation project—something akin to a Linux Foundation under the leadership of key developers, with the explicit goal of creating one Linux distribution with sophisticated package management that encompasses all essential functionality, thereby preventing the emergence of unnecessary derivative distributions.

All contributors should then pool their expertise to truly make it a “better system.” Crucially, optimizations for gaming, audio production, and other professional workflows—similar to those built into Windows—must be integrated directly into the system to establish a real standard.

Above all, the system design must change. It cannot consist solely of the Linux kernel with the rest being incidental. Linux must be a complete system in the traditional Unix sense: meticulously maintained, with comprehensive manual pages. Anything extending beyond the core Unix functionality should be provided exclusively through the package management system, with third-party tools installed in paths separate from the operating system’s standard directories.

This way it is better possible to enforce a standard that can be well maintained.

Of course this needs also a good Release management with migration/update paths. You can look how the FreeBSD project organizes things nicely to enable to stay up to date with the system, be it in the current release train or even when migrating to the next release.

But for RME it is not just release drivers. One may already use CC mode of interface and use Linux stock CC driver. If it is not sufficient then Linux developers should develop better CC driver, no need for RME...

But Linux users probably also want TotalMix, DigiCheck and so on...

Sure, it will come one day. And I bet RME is more or less prepared.

And concerning standardisation. EU has forced phone manufacturers to use USB, not proprietary data connection, it also forced them to use USB for charging and not using proprietary PSUs, and later on USB-c to be used for data and power delivery.

USB sort of converged with Thunderbolt (not to 100% but considerably) and Firewire became rather rare.

You are also right about this. RME devices do work in Class Compliant mode with Linux’s standard CC driver, and if the driver isn’t sufficient, then yes — it’s mainly up to Linux developers to improve CC support, not RME. And it’s true that CC mode still lacks important RME features like TotalMix and DigiCheck.

But the situation can always be improved with good cooperation, taking into account the interests of all parties and respecting their rights to industrial secrets and other proprietary information. Hopefully there will be plenty of goodwill and collaboration in the future.