In the post-SGI, pre-Second-Jobsian-Revolution world of the the mid-late 90s, there was a real malaise in workstation computing. The systems were dumbed down, the hardware was homogenized and commoditized into joyless beige, an ocean of "Try AOL Free for 100 days" CDs jostled and slushed amidst the flickering fluorescence of the cube farm, where CRTs and overhead lights would never quite hum along at the same frequency, zapping the air with an eerie, sanitized sick-static. This was the future of computing.
This gloomy world of palpable beige-yellow-purple is the world of Microsoft Windows. As a product, it caters exclusively to that clientele, in that environment. To be blunt, above all other design considerations, Windows was built to accommodate the lackluster office drone -- or, more precisely, their bean-counting overlords who wanted to save on user workstations.
Despite attempts to undo this, the design has proven impossible to build upon. How many legitimate improvements in the cutting edge of either academic or industrial compsci have been built on Windows technology, for any reason other than "MS signs my paycheck so I'm contriving this to appear like I'm happy to be using PowerShell"?
MS was able to paper over the deficiencies for a while through sheer force (pushing .NET as a semi-unified computing environment, Ballmer screaming "DEVELOPERS!", etc.), but Windows was in no way prepared for the revolution brought by virtualization in the mid-aughts, and any shred of a hope that MS would somehow recover was utterly and entirely obliterated by the widespread proliferation of containerization. Good luck getting a real version of that working on Windows.
I've said it before and I'll say it again: in about 5 years, I wouldn't be surprised at all to see the new version of "Windows" plumbed end-to-end atop a nix-like kernel, and driven by a hybrid userland comprised of a spit-polished copy-paste from WINE + a grab bag of snippets from the proprietary MS-internal Win32.
The flexibility of the nix model is the indisputable, indomnitable winner here, and it stunned and killed the Goliath in its tracks. This is the ultimate surrender to the open-source model. Windows's top-down, "report to your cube by 8:26am sharp and don't question the men in the fancy suits" approach to computing resulted in a rigid operating system that was unable to keep pace with the technological demands of the many pantheons of loyal corporate drones that comprised their user base. Even strictly-Microsoft shops are forced to give developers MacBooks now, or they're unable to get anyone competent to sign on.
It really couldn't get more poetic than MS desperately integrating Linux into their OS so that people won't switch to the more flexible nix-like systems for their workstations, although I anxiously await MS accidentally linking in a GPL module and thus becoming required to disclose the Windows source code. :)
Meanwhile, per usual, nix-like OSes have been humming for 50-ish years now and show no signs of slowing down. IMO, that's all the objective, borne-out proof one needs to say that for all practical purposes, Windows couldn't stand the test of time.
Please note that this is not about Linux per se, but the overarching design theory and development processes in use in major operating systems, and the massive success it represents for open-source, research-driven systems.
The chapter on "The Windows Way" has been written, and whatever its benefits may be in theory, they don't bear out as sustainable in practice.
It'd be fun to do a macro-scale timeline comparison to Sun. These invincible tech behemoths that cater to their narrow niche become rotting, hollowed-out fossils as the free systems continue to develop and evolve the cutting edge. Microsoft is right on time here, and I fully expect to see them struggle through the next decade or so until Larry Ellison finally puts them out of their misery.
I think I've read this free-beer explanation of yours in about 100 distinct comments. In most of them you explain why some particular programming language that you hate with a passion came to develop mindshare while it was so unlikely considering what a bad design it is.
For some balance, you could consider how Windows got preinstalled on most computers since the 90s and what huge dominance it had in the consumer and also development domains (also thanks to some evil capitalist practices, some might say!). That there are free versions of Visual Studio and other IDEs and dev tooling for various programming languages for Windows just as well. Also, consider that I can buy Windows 10 Professional for $8.90 with 5 seconds of googling.
This gloomy world of palpable beige-yellow-purple is the world of Microsoft Windows. As a product, it caters exclusively to that clientele, in that environment. To be blunt, above all other design considerations, Windows was built to accommodate the lackluster office drone -- or, more precisely, their bean-counting overlords who wanted to save on user workstations.
Despite attempts to undo this, the design has proven impossible to build upon. How many legitimate improvements in the cutting edge of either academic or industrial compsci have been built on Windows technology, for any reason other than "MS signs my paycheck so I'm contriving this to appear like I'm happy to be using PowerShell"?
MS was able to paper over the deficiencies for a while through sheer force (pushing .NET as a semi-unified computing environment, Ballmer screaming "DEVELOPERS!", etc.), but Windows was in no way prepared for the revolution brought by virtualization in the mid-aughts, and any shred of a hope that MS would somehow recover was utterly and entirely obliterated by the widespread proliferation of containerization. Good luck getting a real version of that working on Windows.
I've said it before and I'll say it again: in about 5 years, I wouldn't be surprised at all to see the new version of "Windows" plumbed end-to-end atop a nix-like kernel, and driven by a hybrid userland comprised of a spit-polished copy-paste from WINE + a grab bag of snippets from the proprietary MS-internal Win32.
The flexibility of the nix model is the indisputable, indomnitable winner here, and it stunned and killed the Goliath in its tracks. This is the ultimate surrender to the open-source model. Windows's top-down, "report to your cube by 8:26am sharp and don't question the men in the fancy suits" approach to computing resulted in a rigid operating system that was unable to keep pace with the technological demands of the many pantheons of loyal corporate drones that comprised their user base. Even strictly-Microsoft shops are forced to give developers MacBooks now, or they're unable to get anyone competent to sign on.
It really couldn't get more poetic than MS desperately integrating Linux into their OS so that people won't switch to the more flexible nix-like systems for their workstations, although I anxiously await MS accidentally linking in a GPL module and thus becoming required to disclose the Windows source code. :)
Meanwhile, per usual, nix-like OSes have been humming for 50-ish years now and show no signs of slowing down. IMO, that's all the objective, borne-out proof one needs to say that for all practical purposes, Windows couldn't stand the test of time.
Please note that this is not about Linux per se, but the overarching design theory and development processes in use in major operating systems, and the massive success it represents for open-source, research-driven systems.
The chapter on "The Windows Way" has been written, and whatever its benefits may be in theory, they don't bear out as sustainable in practice.
It'd be fun to do a macro-scale timeline comparison to Sun. These invincible tech behemoths that cater to their narrow niche become rotting, hollowed-out fossils as the free systems continue to develop and evolve the cutting edge. Microsoft is right on time here, and I fully expect to see them struggle through the next decade or so until Larry Ellison finally puts them out of their misery.