When will it "actually" happen?

[Shibblet]

We always have this buzz about "The year of the Linux Desktop." When will Linux (in general) become a more dominant player in the industry?

We have watched Microsoft slowly push people away over the course of Windows 8 through Windows 11... and brought the Percentage of Linux Desktop use up to 4.55% according to:
https://gs.statcounter.com/os-market...ktop/worldwide

So, I was having a discussion with a friend of mine, who is a PC Gamer running Windows 10. We were talking about why he chooses Windows over Linux. We've done some testing on his PC, and was able to show that most of library of games work just as well, if not better on Linux. We've also shown that most of the software he uses, is available on Linux... and most of the Windows utilities that he uses are irrelevant on Linux.

He explained that you could get games to work on Linux... sometimes you need to tweak them a bit, but they will work. But it all feels a bit like a house of cards. One wrong move, and the game will come crashing down. He also explains that mods and utilities will work as well, but you REALLY have to tweak them to work, and it feels more like a house of napkins.

However, he makes a strong point when it comes to this question. His answer was as follows: "Linux will have more of a market-share, when it works the way Windows does."

i.e. I download my game with GoG, Steam, EA, Ubi, etc. And I play it. It just runs. No tweaks, no launcher, no compatibility layer, etc. "Until Linux works that way, Windows will always be dominant."

My question to him was "Don't you think PC owners / gamers have a problem with constant acceptance of these EULA's... telling them that they have to agree to all sorts of data recording just to play a game?"

His answer to that was "They don't care. Do you think teenagers give a crap about EULA's? They start mashing the controller buttons as soon as the game starts to get past all of the agreements, and publisher screens, and login menus as fast as possible."

Regardless... Do you think he's right? Does Linux have to basically be a better form of Windows in order to truly succeed?

[TheFu]

It happened about 10 yrs ago on Linux phones and tablets. Did you miss that? A number of people in my family don't own a desktop or laptop computer. They have Android tablets instead.

If you are a purists and think only an x86-64 "desktop" or "laptop" machine qualifies, then it won't happen until the huge commercial OS maker implodes. I don't see that happening in my remaining lifespan.

Normal people don't care about the OS. They care about the 10-30 applications they want to run working on the OS. When Adobe drops that other OS and starts development for Linux, that will be one of the early signs. Adding Linux won't be enough. They have to drop the other OS.

[Shibblet]

It happened about 10 yrs ago on Linux phones and tablets. Did you miss that? A number of people in my family don't own a desktop or laptop computer. They have Android tablets instead.

Nope. Didn't miss it. I was talking specifically about the desktop.

If you are a purists and think only an x86-64 "desktop" or "laptop" machine qualifies, then it won't happen until the huge commercial OS maker implodes. I don't see that happening in my remaining lifespan.

So, you think it will require Microsoft to fail before Linux becomes more prevalent? Would SuSE or Fedora (RedHat) move in?

Normal people don't care about the OS. They care about the 10-30 applications they want to run working on the OS.

My friends point exactly. They don't really care about the OS, they just want their games and programs to work.

When Adobe drops that other OS and starts development for Linux, that will be one of the early signs. Adding Linux won't be enough. They have to drop the other OS.

Personally, if Windows fails, I believe Adobe will make their own OS. If you've ever used the extent of their software, you can see how they are really just a couple of modules away of making their own OS.

[him610]

... won't happen until the huge commercial OS maker implodes.

Fat chance of that ever happening (in our lifetimes)!

Have any of my esteemed forum colleagues ever worked for large corporations? (I know several have, or maybe more.)
Most, if not all, large corporations have entire, for a better phrase, techno-systems and eco-$ystems based on the OS software that is produced in Redmond. Those large companies will not ever willingly move to another OS. Maybe to the cloud.

[TheFu]

I've worked in and around huge corporations and govt organizations for most of my career, even when I was actually working for smallish companies directly.
Most used whatever OS was necessary inside their data centers - some mainframes, lots of UNIX and some MS-Windows servers. On the desktop, most had MS-Windows, but a few also had UNIX workstations. I had both in all companies, except 1 where I only had a DECstation in my office at the company, but in my lab at the client location, I had a PC with WindowsNT and an Alpha running OSF/1 UNIX.

My last "real job" had over 100K physical UNIX servers in data centers around the world and about 30K physical MS-Windows servers for software that would only run on MS-Windows. Typically, we would deploy 10 to 100 virtual machines on each physical server, but it was completely based on loads. Some DBMS servers used the full physical server with 64-CPUs running HP-UX, no virtual machines at all.

I worked as an architect addressing internal corporate client system design needs. My boss was the Sr director over all the project-based architects, but I generally didn't work with anyone else in my team. We were farmed out to other project teams for 3-24 months as they designed and deployed new/upgraded systems to solve their business problems. My boss asked me for an estimate to replace 25K MS-Windows laptops for 25K Linux laptops to perform the same task. I spent about 8 hours creating an estimate for the Linux replacement, including talking to application vendors for WAG to support Linux. In short, we were spending about $4M/yr for MS-Windows software on the systems and the migration costs to port the 10 applications and drivers from MS-Windows to Linux were either not possible or somewhere over $100M and 5+ years to complete. So, the conversion from Windows to Linux didn't make any sense because it would take over 20 yrs to break even assuming the annual maintenance costs were the same for either OS. I had lunch with my boss and provided a 1 page WAG estimate for him to look over the 2 different prices. He never asked me about again. Sometimes Linux isn't actually the better choice. I didn't include re-training for 25K employees who already knew how to use Windows and all the infrastructure software changes like for VPNs, patching, and normal "Office" document handing under Linux that would also be necessary.

Anyway, I did say that it wouldn't happen in my lifetime. The first step is to get laws changed that require open document formats for all interactions with the Federal govt, not mandating specific, proprietary applications as is done today in most countries. Politicians don't care enough to make a law like that and they don't want to make MSFT unhappy, so it won't happen, at least not in the USA.

From a govt or corporate standpoint, they want to pay someone to handle these things. Their core business isn't about dealing with software, so they want to minimize the effort and maintenance around that. If they had their way, they'd never touch any OS once installed. If it is working, don't touch it unless a feature that will help them make at least 10% more money is included in the change. Security is a cost-of-business and a risk, not a way to make a profit for most corporations.

So, MSFT software on the desktop isn't going anywhere, unless MSFT really messes up. Their attempt to force a subscription model is making some corporate customers very unhappy. This unhappiness isn't well publicized related to MSFT, but it has been related to Oracle and VMware. Companies don't want annual or monthly subscriptions to software, unless THEY are the company offering the subscription.

[nicolasdentremont]

Year 2062

According to that graph. In August 2009 Windows had 94.69% whereas Linux had .69%.
In August of 2024 Windows had 71.46% whereas Linux had 4.55%.
Over the 15 years, Windows lost 23.23% and Linux gained 3.86%.
If we extrapolate that into the future, the two intersect at some point in year 2062.

Now that's assuming that:
A. The website is accurate (it might not)
B. The pattern is linear and predictable (it probably isn't)
C. My math is correct (I kinda suck at math)

In all seriousness, I think Linux based operating systems are already the most widely used OS just not on traditional desktop PCs. I doubt that Linux operating systems are ever going to outnumber windows for desktop use but I can definitely see it slowly becoming more and more popular for personal and work computers for small businesses. Free and Open Source software has really become an important part of my small welding business.

The user friendliness of Linux distributions like Ubuntu has improved over the years and so has search engines. So more things work right out of the box and if they don't finding a solution is usually fairly easy.

What got me to switch over and ditch Windows was Steam. I would install a Linux distro every few years just to see what was new. Last winter I decided to check out Ubuntu again and found that Steam released a Linux client and developed Proton to run Windows games. Between Steam and Lutris I can now run the few games I play on a regular basis without issue and without needing to switch operating systems. All the software I was using was already free/open source so switching over was easy.

As far as ELUAs go, yes I think most people just click agree without a care about what they agreed to. I've read many and they all seem to boil down to the same thing.

But anyway. I think Linux itself is already a success regardless of it's popularity as a desktop OS choice compared to those made by huge companies.

[TheFu]

If a company doesn't start out using MSFT software, then they won't become trapped using it forever. But most companies just starting out don't think about those initial choices and just want to solve a problem the easiest way. It is only later, when they've grown to about 50 people that it is already too late to change and MSFT software costs balloon on them. They didn't know that in the beginning. It was cheap enough back then and without much expertise, they could point-n-click enough in MS-SBS to have a working central server for their 3-10 client machines.

A few companies start out using Linux because the founder(s) made a conscious choice in the matter. They use Linux desktops and have remote access into Linux servers, mostly running F/LOSS tools. When they need to interface with other companies, it is likely they run into compatibility issues around documents, because the other companies all have been using MSFT proprietary software for decades or simply didn't know there were any alternatives. This forces the non-MSFT companies into having at least 1 MSFT system setup with remote desktop access to create documents that can be sent to clients in the format they prefer.

[werewulf75]

The server - and beyond - market is already largely 'hooked' on one or other of the large Linux distros. The desktop lags far behind still both in the corporate/business and the domestic user market. A lot of things would need to happen for the desktop to actually be able to compete or even overtake the Redmond company's excuse for an OS.

First, there are two huge obstacles, or rather three, standing in the way. Two have been identified you could say by two very prominent individuals in the Linux world a few years ago now. They are none other than Linus Torvalds and Mark Shuttleworth. The former expressed dismay that there were so many Linux distros, far too many really, rather than a unitary system or maybe a handful. The latter was disappointed that there are too many WMs/DEs. The third obstacle follows on from those two - a complete absence of a really strong, single brand image.

At a secondary level and following on from those obstacles, there are the problems of all manner of multiple dependencies and standards. Then there would be a need for a simple, single installer system - and uninstaller - for apps and drivers etc. so that third party developers can easily distribute to all major distros.

Last but not least by any means, a truly modern, innovative WME or WM/DE would be needed that just kicks everything else's ***, that is so user-friendly a 100-year old granny could use it in a day and that knocks people's socks off in terms of looking 'cool' as well without distracting from tasks. And it needs to be immediately identifiable as the brand.

Further, security and absolute privacy would have to be practically buit-in, including also something along the lines of Proton AG's suite of (open source) privacy apps.

At the same time, the Terminal would of course need to be retained, with an option to boot straight into this for sys admins etc. and not least for us more 'geeky' folks.

Perhaps an option for users to buy in-house support by email subscriptions might also be good.

There are a whole lot of other things that would be important, but these are the fundamental changes needed if Linux really was to have a chance to become a or even the major desktop player. I for one would be all in favour, because Linux has to grow or else it'll keep fragmenting into ever more distros and flavours and spins etc. etc. with fewer and fewer users, until the desktop will be pretty much lost.

[iiwiars]

But will we have the same pleasure in using Ubuntu if one day everyone uses it and it becomes completely mainstream?

[dacha]

Year 2062

Of all the regression types LibreOffice offers as trend lines for that graph, the best fitting one is the exponential curve, with an R^2 of 0.924 (a close correlation). Extrapolating from that curve, desktop Linux reaches 10% market share in August 2035 .

But it's hard to tell when, if ever, Linux will be the most popular desktop operating system. After all, many people - even professional IT people who should know better - see Apple as the Microsoft alternative , a company whose products cost more and who treats them even worse than Microsoft (incompatibility, lock-in, spyware, parts pairing, walled garden, etc.).