enorbet said:
Apparently you still have yet to look at the chart on OpSys Market Share as you would have members believe that Linux with a 35% Market Share on Enterprise Servers (compared to 32% Windows) and leaving out the other Unix based balance, and/or the 96% Market Share on Supercomputers, that these multi-BILLION dollar ventures rely on junk that breaks all the time from a simple update. That is utterly laughable.
You also apparently didn't visit CoreOS or look up the nature and value of containers. When all base system incremental updates are disallowed, even viruses and other malware are ruled out as well in any important area. The existing rules and security of a mature OpSys that began in 1969 as a multi-user, networked system are more than sufficient to compartmentalize and handle any incidentals.
This is in sharp contrast to a system that still doesn't allow concurrent users (nor apparently know how to properly handle them) began as a standalone single-user OpSys and only got a TCP/IP stack nearly 30 years after Unix (which Linux inherited from the start) and had user logins with root/admin privileges by default (easily available to any script let alone human hacker) until 2009. These are not my opinions or of some IT guy somewhere of unknown credentials. These are historical facts. I have yet to see one shred of evidence from you. Almost everything is your opinion and/or anecdotal. With your opinion on Linux I find it confusing and doubtful that you really allow Ubuntu anywhere near your PC.
Back on topic: Don't take my word. Don't take anybody's word. It is so easy to see for yourself. Perhaps try
-T H I S L I N K-. Even if you're not motivated to download it and boot it, just look at it and see how "not up to snuff" it is. It will likely surprise and delight you, and this is but a portable, temporary system. The full deal is even better.
Look, you're a Linux evangelist. I get it. But your fervor for this particular operating system is impacting your ability to discuss the issue objectively. The topic is operating systems for the end-user on a desktop or laptop, not servers. And yes, from all the system administrators I have talked to, Linux does break much more often than Windows or Solaris. It is widely used because businesses see the higher maintenance cost as a trade off for not having to worry about paying for licenses. Also, most of these companies that are sinking millions of dollars into servers that run linux are not using normal builds. They pay a third party company or someone on the inside to create a custom solution which undergoes extensive quality assurance before any changes are made. This stands into stark contrast with a proprietary solution like Microsoft or Oracle where most of the quality assurance is done by the companies that manufacture the software.
These guys are not just slapping in some distro they downloaded and updating as needed. They typically pay a lot of money to work around the flaws in the quality assurance of the various distros. Some of the big boys actually develop their own distro that is not available publicly.
And you accuse me of using anecdote, but I am the only one that has contributed meaningful empirical evidence. The desktop penetration of Linux has stayed around 1%. If Linux's consumer experience were really on-par with Windows and OSX on the desktop, given it's price, you would expect it to be gaining significant market share. When you cannot give something of value away for free to the average man on the street, that speaks volumes to its quality in the eye of the average consumer. Linux desktops are far too fragmented and the Linux OS is far too rooted in ancient server systems from nearly half a century ago to be successful with end-users. Unless someone fixes these basic flaws, I do not see this changing.
That is all I am going to contribute to this subject. You clearly are a very passionate evangelist. My only suggestion would be instead of talking about why people should be using Linux as a desktop OS (when they clearly are not going to), you should put that effort into some of the problems that I (and others) have identified and fixing them, just as a bit of a rundown:
1) Reliance on the buggy X-windows server.
2) Reliance on the buggy KDE and Gnome windows managers.
3) Fragmentation of both the underlying operating system (e.g. Debian versus Redhat) and the higher end user experience.
4) Poor software installation routines for third party binaries (treating user software like system software, not providing for a central installer for linux binaries similar to what Windows has with .msi packages or OS X Installer.app).
5) Limited or buggy driver support for certain hardware (could use a quality assurance program similar to Microsoft that requires signed drivers).
6) Fragmentation of the user desktop experience. The average consumer does not like to have to relearn anything. With Linux, desktops are fragmented by choice in desktop manager and then further fragmented by how that manager is implemented by the distro. After seeing how users rebelled against the evolution of the start menu in Windows 8, it is clear that end-users do not like having to learn a new way to do the same thing. If Linux cannot offer a new, clean, and consistent desktop experience, it is going to turn off users.
7) Fix the situation with proprietary hardware. Linux needs to license software for doing basic consumer tasks such as playing MP3's, DVD's, and blurays and require that it be included on any new non-volume purchase of a computer with Linux preinstalled. Consumers want to put in a blu-ray and have it just work. They don't want to be forced to download a program that violates federal law to do so.
8) Fix the fragmentation of various underlying aspects of the Linux OS, such as the audio system which has over a dozen implementations instead of a single easy to use one like on Windows or OSX.
These are just some of the major flaws which are obvious to me (and others). Given the state of Linux today, I don't see them being fixed unless some major corporation like Google is willing to sink billions into Linux on the Desktop, probably by pulling a Steve Jobs and building a completely new user experience on top of the existing stable Linux core.
