Here Comes Windows 11....

[Bobby Henderson]

Software developers really should step-up their game and make their stuff work under Linux.

They're not going to bother for a variety of reasons. Here's the two most obvious ones:

First, Linux is not an OS that is consumer friendly. The insult that Windows is a "toy OS" is really a double edge sword that turns into a damning judgment against the marketing potential to sell Linux to the general public. Linux has an image of being only accessible to hardcore computer geeks. Any casual user should not even dream of attempting to try. A great deal of the Linux user base even cultivates and promotes this for experts only image. Well, here's the thing: the overwhelming majority of the world's computer users are not experts at all. It's scary to consider how many don't even understand how to organize files and folders in the computer's file system. I think Microsoft and Apple understand this dilemma very well, perhaps even more so for Apple. If a potential customer is scared of using the product that person is not going to buy it or use it.

Second: there are too fucking many different distros of Linux out there. They're not all compatible with each other. Why can't there just be one Goddamn standard for Linux? Software companies do not have endless amounts of resources to be able to develop applications for multiple operating systems. Many can't even go beyond just developing for Windows. Making a Mac OSX version is too big a burden for many firms. In the graphics market I look at Corel and their CorelDRAW application as an example. They can't even keep up to speed with their Windows version (Adobe is running circles around them, providing far more frequent updates to Illustrator), much less fix the bugs that are present in their fledgling Mac OSX counterpart. Adding a Linux version would be a bridge too far for that company. Even Adobe had only fleeting amounts of experimentation with UNIX variants in the 1990's and did away with it going into the 2000's.

[Frank Cox]

Linux/Unix is designed with the idea that you have a system administrator. Your system administrator sets up your computer for you and you, as the user, use that computer in the way that it was set up for you. End users were never intended to be able to do things like installing new software, updating the operating system and the like. As a user you use the programs, and if you need another program you tell the system administrator. It's a great paradigm for the office since the receptionist and the machinist won't be screwing up the computers that are on the desk.

When you bring Linux home with you, now you can't just be the user. You also have to be your own system administrator, which requires at a modicum of know-how.

I can set up a computer for grandma to use to check facebook and read email from the grandkids. It'll be perfectly suited to the task at hand, she'll be happy with it and I can come back in a year and the thing will still be working just like and look just like it did the last time I saw it.

On the other hand, "I want to play the latest games and install random crap from the Internet." That's not going to work well with Linux unless you know what you're doing.

[Martin McCaffery]

I remember very early on in the personal computer age, pre-DOS, pre-Mac, I heard a computer guy on a talk radio show, pitching the glory of personal computers. He said "You need to know how to program your computer like you need to know how to change your transmission."
I still think that is about 90% correct and would be perfectly fine with most personal computer owner/ operators. Just make the damn thing work the way I want it to.

[Bobby Henderson]

I don't think needing to know how to change your transmission is a very good analogy. That might work with a really old car. With any newer vehicles the DIY garage mechanic would need some speciality tools made only for that brand of vehicle and access to the computers made only for that brand of vehicle. This nonsense has even filtered into commercial vehicles like big rig semi trucks. That limits DIY garage mechanics to doing only basic maintenance tasks, such as changing air filters. And even that can be a big pain depending on the vehicle model, which is one reason why I drive a pickup truck (more room under the hood, parts are easier to replace, etc).

It's much easier to build a custom desktop PC from parts bought individually or do things to service a desktop or notebook computer. Unfortunately, OEMs have been doing things to make products with fewer components that are user-serviceable. This is especially true for a growing number of notebook models. They're doing sleazy things like soldering the RAM and other components onto the motherboard. If something breaks in a MacBook Pro the customer will likely have to ship it off to an Apple-certified repair center, and pay top dollar for the repair work. More OEMs that sell Windows-based notebooks have been copying Apple's moves in recent years.

40 years ago it took a bit more careful planning and paying attention in order to work with and manage files on a computer running MS-DOS or anything else with a command line rather than a graphical user interface. And that's what I really don't understand with so many people these days who can't figure out how to manage their files in Windows or Mac OSX. The Finder or Windows File Explorer isn't all that hard to use.

[Martin McCaffery]

"That might work with a really old car."

Well, it was about 40 years ago;>

[Marcel Birgelen]

Apparently, Microsoft has pulled the Health Check (haven't tried it yet, but I read it somewhere yesterday). Also, apparently, the whole thing got so much backslash, they're now, apparently, also considering to support "a select number of 7th generation Intel CPUs that conform to their ideas about 'security'.", whatever that may be...

They're not going to bother for a variety of reasons. Here's the two most obvious ones:

I'd say it depends on the company and their underlying strategy. The big elephant in the room here, at least from my perspective, is Adobe, which never really committed itself to the Linux desktop. Even back when Flash was somewhat relevant, they never managed to deliver a single good release of their player for Linux, let alone a 64-bit version.

Then there is Microsoft itself, but interestingly, Microsoft is increasingly adopting Linux into its own stack of solutions, going even so far as releasing a MS SQL for Linux, that actually seems to work better than its counterpart on Windows. This is all server-side, so it doesn't hurt Microsoft's desktop business, but nevertheless, those are striking developments.

Another thing that's severely lacking on any relevant Linux distribution is integration into corporate networks. This is were Microsoft stole all the good ideas from Novell NetWorks and integrated it directly into Windows. Even though Novell became a major player in the Linux ecosystem, they never managed to deliver anything compelling to replace the Microsoft domain and policy infrastructure. This is what eventually caused Microsoft to become a dominant player in the corporate and enterprise networks infrastructure and where it largely remains till this day...

First, Linux is not an OS that is consumer friendly. The insult that Windows is a "toy OS" is really a double edge sword that turns into a damning judgment against the marketing potential to sell Linux to the general public. Linux has an image ofA being only accessible to hardcore computer geeks. Any casual user should not even dream of attempting to try. A great deal of the Linux user base even cultivates and promotes this for experts only image. Well, here's the thing: the overwhelming majority of the world's computer users are not experts at all. It's scary to consider how many don't even understand how to organize files and folders in the computer's file system. I think Microsoft and Apple understand this dilemma very well, perhaps even more so for Apple. If a potential customer is scared of using the product that person is not going to buy it or use it.

Yes, many Linux distros were notoriously hard to install, but if you look at one of the biggest current distros: Ubuntu Linux for Desktop, then there is nothing really hard about installing it and using the out-of-the-box functionality provided within, at least if the hardware is fully supported. Honestly, the experience may even beat Windows 10 in many aspects, one being the fact that the installer automatically pulls the latest packages from the Internet during installation and you don't need to go through 5 separate update/reboot loops for example.

Things start to go south once hardware isn't fully supported and you need to start tweaking around with stuff. It's not like this can't happen with Windows, but most people get their computers pre-loaded with Windows on top of it and hardware suppliers make sure their hardware at least works with the then-current iteration of Windows.

Second: there are too fucking many different distros of Linux out there. They're not all compatible with each other. Why can't there just be one Goddamn standard for Linux? Software companies do not have endless amounts of resources to be able to develop applications for multiple operating systems. Many can't even go beyond just developing for Windows. Making a Mac OSX version is too big a burden for many firms. In the graphics market I look at Corel and their CorelDRAW application as an example. They can't even keep up to speed with their Windows version (Adobe is running circles around them, providing far more frequent updates to Illustrator), much less fix the bugs that are present in their fledgling Mac OSX counterpart. Adding a Linux version would be a bridge too far for that company. Even Adobe had only fleeting amounts of experimentation with UNIX variants in the 1990's and did away with it going into the 2000's.

Personally, I totally agree, although many people in the community say that having many distros and many choices is part of the "freedom". In my opinion: An operating system should just work. Not having to mess around with nitty gritty details that should be trivial and having more time to do other stuff, that's what I consider freedom. It looks like the community is maturing though and there now is a more common understanding that less is sometimes more. Many distributions are being abandoned and there is a pretty strong consolidation going on right now, with Ubuntu and Debian being strong favorites.

Maybe we only see Linux on the desktop getting hold when the desktop itself has become a mostly irrelevant concept, but stuff like this Windows 11 thing may just be the push it needs to drive enough people away from Windows, into a completely new ecosystem. You can see how quickly stuff can change, while Microsoft may be the dominant player on the desktop market by far, they're all but irrelevant in the mobile OS space as of now and it's not like they didn't throw billions and billions of dollars at it. Sure, they were late to the party, but I also blame part of the failure simply due to the fact that people simply didn't trust Microsoft.

[Bobby Henderson]

Apparently, Microsoft has pulled the Health Check (haven't tried it yet, but I read it somewhere yesterday). Also, apparently, the whole thing got so much backslash, they're now, apparently, also considering to support "a select number of 7th generation Intel CPUs that conform to their ideas about 'security'.", whatever that may be...

My guess is Microsoft got a reality check on just how much of the user base is relying on aging or really old PC hardware. It's now pretty common to find PCs that are over 10 years old in both office and home environments. It sounds like the TPM 2.0 and Secure Boot requirements would not only knock out all of those users, but also quite a few people who consider their computers still fairly new. We're many years removed from the 1980's and 1990's when PCs were replaced more frequently.

I personally don't care if I'm unable to install Windows 11 on any PCs I currently use. Obviously other people feel different about the issue and want to be able to upgrade their existing PCs. I think Microsoft is possibly looking at much greater levels of backlash in 2025 when they're scheduled to end support for Windows 10.

Most of us understand any computing device in general has a limited life span. Nevertheless, many people buy a computer in the same mindset as they would buy a TV set or some other kind of home appliance that can deliver 20 or more years of use. They get pretty angry if something forces them to replace a device that otherwise seems to be working properly. They look at the "upgrades" as nothing more than ploys to make more money. Honestly, that is part of the deal for Microsoft and other PC vendors. They're not going to stay in business if the customer base has to buy a product only one time ever.

The big elephant in the room here, at least from my perspective, is Adobe, which never really committed itself to the Linux desktop. Even back when Flash was somewhat relevant, they never managed to deliver a single good release of their player for Linux, let alone a 64-bit version.

From Adobe's point of view Linux has all kinds of problems. First: there is the previously mentioned issue of all the different versions of Linux out there. That fractures the potential Linux customer base into multiple smaller pieces. Which distro(s) do you support? Second: a pretty big chunk of the Linux user base is hardcore about open source ideology and want people making applications for Linux to do it all open source. There is no way Adobe is going to give away applications like Photoshop, Illustrator or After Effects for free.

Then we get to add in the variables of new CPU types and having to write code native for them. Adobe is a huge company, but even they are having to move slowly, application by application, creating M1-native versions of their applications for Mac OSX. And they're having to maintain the Intel versions for at least the next couple or so years. That's a fairly big burden. I could just see Linux getting even more scattered if there was Intel, ARM and other CPU code bases for various distros.

Maybe we only see Linux on the desktop getting hold when the desktop itself has become a mostly irrelevant concept, but stuff like this Windows 11 thing may just be the push it needs to drive enough people away from Windows, into a completely new ecosystem. You can see how quickly stuff can change, while Microsoft may be the dominant player on the desktop market by far, they're all but irrelevant in the mobile OS space as of now and it's not like they didn't throw billions and billions of dollars at it. Sure, they were late to the party, but I also blame part of the failure simply due to the fact that people simply didn't trust Microsoft.

The mobile phone market and desktop computing are two completely different paradigms. Mobile phone users didn't need anything from Microsoft there. Google and Apple managed to consolidate much of the mobile space to themselves. I vaguely remember Microsoft's attempts there; IIRC, it was a mess. At first the devices appeared to be aping the Blackberry or other handsets with less or more physical buttons. Once Google and Apple were running away with the market it was too late for Microsoft to adjust.

The advantage Microsoft has on the desktop is their position is historically entrenched going all the way back to the days of MS-DOS.

When anyone uses a personal computer for any significant amount of time they'll develop a growing collection of files. Some files might be simple media files that can be viewed in different applications, different operating systems or even on mobile devices. But other files are often dependent on a host application to open and edit. That's where Microsoft and other commercial software vendors like Adobe really build their leverage.

Asking a computer user to switch OS platforms can be one hell of a big ask. The user's applications have to be available on that alternative platform. And the versions of those applications on the different OS need to be able to open the user's existing files accurately. That's not so easy. Little things like font files can gum up the works there. For most computer users the act of switching to a different platform is often more trouble than it is worth.

[Scott Norwood]

On the other hand, "I want to play the latest games and install random crap from the Internet." That's not going to work well with Linux unless you know what you're doing.

Agreed. I _am_ a Linux systems administrator (and AIX and, previously, Solaris, HP-UX, Irix, and others), and I think that it's most useful for either very simple tasks (user logs in and runs one application) or very complicated ones, but probably not for the average home user.

Nevertheless, many people buy a computer in the same mindset as they would buy a TV set or some other kind of home appliance that can deliver 20 or more years of use. They get pretty angry if something forces them to replace a device that otherwise seems to be working properly. They look at the "upgrades" as nothing more than ploys to make more money.

I think that most people are OK with the idea of upgrading their computer (which can range from buying new RAM to replacing everything) when the new one can do something that the old one cannot, or can do the same thing faster or better. If one owns a machine that needs to be upgraded to run Fun New Game X or work with Whizzy Bangy Device Y, I think that most would be fine with that. The problem now is that improvements in performance happen in very small increments. If a machine from five years ago works perfectly well, then most people will balk at the idea of upgrading it because an OS upgrade which is mostly cosmetic in nature will not run on the old one. And even stuff like UEFI is often buried in obscure BIOS settings, such that a compatible machine will appear to be incompatible because the setting is set wrong. What does a TPM offer to the average home user who isn't using encrypted filesystems? Not much. And Secure Boot makes it difficult to boot multiple operating systems and impossible to, say, boot a custom Linux kernel, so people who need this functionality are hurt by it.

[Ed Gordon]

I have taken a serious look at moving to Ubuntu several times in the past years. Every time I consider it, I find at least a few of my applications are not compatible with unix. The desktop market share for unix comprises just over 2% of the market. Developers have not shown any real interest in porting their apps to Unix, so the only real choices are Windows or a MAC.

It is far too early to judge how Windows 11 will be greeted. I have not seen anything about it that gets me remotely interested in it. Microsoft has shown an almost unique capacity to repeatedly shoot themselves in the foot, so I expect to see Microsoft stubble once again.

The only good news about the announcement of Windows 11 is that it will redirect M$ activity away from Windows 10, so that Windows 10 will finally become stable.

[Leo Enticknap]

The one thing I don't like the sound of is that UEFI will be mandatory for a W11 system drive. It's an absolute pain in the poohole to set up dual booting from different partitions under UEFI, though I suppose I'll have to bite the bullet and figure out how to do it.

[Bobby Henderson]

I wonder how hard it will be to set up a virtual machine on a Windows 11 system. Lots of people have old PCs and old install discs for past versions of Windows. I have an old PC in my 2nd bedroom I sometimes fire up to run vintage software that's no longer developed (Macromedia Freehand and the Kai's Power Tools plug-in for Photoshop are two examples). VMs can be set up in Win10 and Win7.

[Leo Enticknap]

Afterthought on the UEFI thing: I guess that if I use two physically separate drives for Windows and Ubuntu, that might work, as long as the Windows UEFI bootloader will let me launch the legacy Ubuntu drive, or vice-versa. Not much of a problem for desktops, but very much a problem for most laptops. I've relied on having Windows and Ubuntu available on the same computer for many years now, and am not looking forward to that becoming significantly more difficult. I use both, all the time.

[Marcel Birgelen]

From Adobe's point of view Linux has all kinds of problems. First: there is the previously mentioned issue of all the different versions of Linux out there. That fractures the potential Linux customer base into multiple smaller pieces. Which distro(s) do you support? Second: a pretty big chunk of the Linux user base is hardcore about open source ideology and want people making applications for Linux to do it all open source. There is no way Adobe is going to give away applications like Photoshop, Illustrator or After Effects for free.

Then we get to add in the variables of new CPU types and having to write code native for them. Adobe is a huge company, but even they are having to move slowly, application by application, creating M1-native versions of their applications for Mac OSX. And they're having to maintain the Intel versions for at least the next couple or so years. That's a fairly big burden. I could just see Linux getting even more scattered if there was Intel, ARM and other CPU code bases for various distros.

It's a bit of a chicken and the egg problem, but in my humble opinion, those big-ticket companies could've forced some kind of consolidation of distributions. If Adobe would've said: See, we're only going to support the APT-packet infrastructure and nothing else, that would probably have resulted into a big shift towards APT-based distributions like Debian and Ubuntu early on, as having the Adobe stack of productivity/creativity products available on a certain Linux distribution, would've been a huge draw.

I don't know how many assembly code there is left in Adobe software, but I doubt it's all that much those days. I guess the more hardware/OS specific optimizations nowadays are mostly in the GPU space, since most compilers are pretty much optimized for their given platforms. It's also increasingly hard to program for multi-pipelined CPUs on a low-level basis, something compilers are getting better at in each iteration. It's like the modern automatic gearshift in your car. In certain situations, manual shifting may be more efficient, but in general, automation nowadays achieves a better result than the average driver.

I wonder how hard it will be to set up a virtual machine on a Windows 11 system. Lots of people have old PCs and old install discs for past versions of Windows. I have an old PC in my 2nd bedroom I sometimes fire up to run vintage software that's no longer developed (Macromedia Freehand and the Kai's Power Tools plug-in for Photoshop are two examples). VMs can be set up in Win10 and Win7.

If they don't drop the TPM 2.0 requirement, then you'll only be ever able to run it on a host that also has TPM 2.0. Also, if the Secure Boot stuff remains in place, it will only boot if the UEFI bootloader of your hypervisor has been whitelisted by Microsoft.

Afterthought on the UEFI thing: I guess that if I use two physically separate drives for Windows and Ubuntu, that might work, as long as the Windows UEFI bootloader will let me launch the legacy Ubuntu drive, or vice-versa. Not much of a problem for desktops, but very much a problem for most laptops. I've relied on having Windows and Ubuntu available on the same computer for many years now, and am not looking forward to that becoming significantly more difficult. I use both, all the time.

You can't rely on the Linux boot loader (GRUB), because that won't be signed by Microsoft, hence, GRUB will not be able to boot Windows. I don't know how well most UEFI implementations support booting from separate partitions on the same disk, but it should be a safer bet to install them both on their own disks and switch between them using the UEFI boot menu provided by your hardware.

[Mark Gulbrandsen]

I have taken a serious look at moving to Ubuntu several times in the past years. Every time I consider it, I find at least a few of my applications are not compatible with unix. The desktop market share for unix comprises just over 2% of the market. Developers have not shown any real interest in porting their apps to Unix, so the only real choices are Windows or a MAC.

It is far too early to judge how Windows 11 will be greeted. I have not seen anything about it that gets me remotely interested in it. Microsoft has shown an almost unique capacity to repeatedly shoot themselves in the foot, so I expect to see Microsoft stubble once again.

The only good news about the announcement of Windows 11 is that it will redirect M$ activity away from Windows 10, so that Windows 10 will finally become stable.

Exactly the same problem I had with it Ed. None of the standard Photo Apps like Photoshop or Capture will run. Those are the two main apps that I use... I do have a mobile (car) computer that can just run W-7, but it sure runs a lot cooler with Ubuntu on it.

[Scott Norwood]

The dual-boot thing is going to suck, for sure. As is the fact that PXE booting doesn't work the same way in a UEFI environment (I don't know the details...I've always just set everything to "legacy mode" and used it that way).

I'm fine with offering UEFI and Secure Boot as options, but forcing them is just going to result in more environments that stick with older operating systems.