Re: One Of Life's Best Rewards Is Watching Linux Fail !!!!

On Thursday, June 9, 2011 5:59:12 PM UTC-4, flatfish+++ wrote:

I assume you mean "Fail on the PC Desktop"

Watching Linux Fail:
http://limuxwatch.blogspot.com/

Coverage of the program to convert PC users from Windows to Linux. A big problem here is that the PCs must be purchased with preinstalled copies of whatever version of Windows comes with them. It's harder to convince people to make the switch when Windows itself is "free". The actual cost to the OEM is about $30 and these days the OEMs are having a hard time recovering the cost of parts and labor on the hardware, let alone add big mark-ups for Windows.

The Microsoft OEM license agreement requires that the OEM include Windows on EVERY machine they produce, before it is shipped. Dell and HP used to get around this legally by shipping machines without hard drives. The problem is that if the end-user then installs Windows, regardless of how he purchased it, the Windows Genuine Advantage will see that the machine was never licensed for Windows and Microsoft can use this proof to demand the full retail price for each such machine. Even if only 10% of the "No-OS" machines end up running Windows, the cost of the pirated copies at $300 each would be more than the cost of the OEM licenses at $30 each for pre-installing Windows the cheapest hard drives you can buy. If Microsoft does identify this 10%, the burden would be on Dell to prove that these machines HAD been legally licensed. They would have to have each customer send them proof that they had legally acquired the correct type of license. A consultant who picked up a copy of Windows Student edition for $30 at the local college would not be properly licensed and Dell could still end up having to pay the $300 price. The cost of getting the proof would far exceed the $300 price, and may end up being cost in addition to the cost of the full retail cost of the license. With WGA, Microsoft has all of the leverage, leaving the OEMs in a vulnerable position.

The net result of this situation is that nearly every PC sold by the major OEMs has to be sold with Windows pre-installed. There is a relatively small market of "Computer Fair" PCs which are typically assembled from components selected by the customer at the vendor's kiosk. These kiosks offer OEM licenses for Windows. In the USA, the price is typically around $60. Roughly 1/2 day's pay at median wage in other countries. However, if the buyer is planning to run Linux on that PC he is likely to purchase a Linux-ready motherboard and peripherals, and will probably spend the extra money on a faster hard drive or a second hard drive.

One strategy that several OEMs have begun to adopt, is to create machines that will not support user-installed Windows. For example, Tablets made by Sony, Amazon, Pandigital, Archos, and about a dozen others, many in the consumer electronics business rather than the "PC" business, have been making devices powered by the ARM chip rather than the Intel chips. Microsoft does produce an Embedded version of Windows 6 and Windows 7 mobile editions for the ARM chip, but they are not user installable. Needless to say, this is a game changer, and has opened up a huge market for secondary devices such as cell phones, 7 inch tablets, and 10 inch tablets, all of which are far more profitable than most PCs in terms of actual profits earned by manufacturers and retailers alike. As a result, the biggest problem has been keeping the Linux devices in stock.

However, Microsoft still likes to see benchmarks quoted strictly in terms of actual "PCs" that are actually capable of running Windows. Very often, a monopolist who is facing stiff competition will attempt to maintain it's claim of market dominance to keep the loyalty of customers - by simply maintaining their claim over a subset of the market that gradually shrinks.

In the 1980s, Novell was a leader in the networking of PCs. Netware networks were found in most corporations. Netware provided basic file sharing and simple e-mail functions that were perfect for small teams, such as a manager and his direct reports. Typically, each network would have less than 256 users on it, and the users were often connected using 10-base-5 coaxial cable that could be run in a short loop from the server to the desks of the manager and his reports.

By the early 1990s, however, enterprises wanted to integrate the network into a larger corporate network so that all of the PCs could be connected and information could be sent throughout the whole network. Most corporations were finding that the best networking protocol for doing this was TCP/IP. Because if it's DARPA roots, military backing, and widespread use in the National Research Foundation networks, it had already been well tested, and many of the issues of security, scalability, reliability, performance, and operational support for identifying and fixing problems quickly, had already been addressed. As a result, corporate networks were being rapidly transitioned to TCP/IP, often over the protests of proponents of DECNet, SNA, and Netware. These early networks were often supported by UNIX servers, which provided e-mail, file transfer, and shell account access, as well as applications that worked through the terminal (telnet) interface.

Novell didn't want investors and existing customers to drop their support, or the cash flow might stop abruptly and end their business. As a result, Novell redefined their niche, by calling it the "Network OS" market. Novell carefully defined that market to exclude UNIX and also to exclude Windows PCs that might be running alternatives such as NetBIOS. This made it possible for Novell to now claim that they owned 95% of this self-defined market. The reality is that in terms of networked nodes, the Novell share had gone from about 90% of the 10% of the PCs and servers that were networked, to 10% of the 90% of the PCs and servers that were connected via some form of network, including SPX/IPX, TCP/IP, OSI, DECNet, SNA, and NetBIOS. By 1995, TCP/IP had about 80% of all PC users connected and networked, usually to UNIX servers which now also supported web browsers and web servers as well as e-mail, file transfer, and file sharing (SAMBA, Microsoft/IBM SMB, and CIFS). Many corporations also adopted Lotus Notes as a way to combine e-mail, teams, and document archiving into a single interface that was more efficient at storing data, and provided an audit trail that could be used for legal proceedings.

Microsoft has been doing the same thing. Microsoft is very sensitive about benchmarks, and especially ANY comparison between Windows and Linux. This includes such benchmarks as customer satisfaction, ease of installation, or number of devices running different types of operating systems.

Microsoft prefers to share it's market share in terms of "PC Operating Systems", with a PC being narrowly defined as a desktop or laptop device with an Intel processor and pre-installed Windows operating system. This pretty much means that the only Linux devices that would be counted would be systems where the end-user had removed Windows completely and installed Linux instead.

In surveys where Linux looks better, the survey is usually given in terms of "Client Environments", which would include Macs, PDAs, Netbooks sold with Linux preinstalled, Tablets, e-readers, kindles, and so on. Needless to say, Microsoft has a much smaller share of the overall client environment market.

Desktop Linux: The Dream Is Dead
"By the time Microsoft released the Windows 7 beta
in January 2009, Linux had clearly lost its chance at desktop glory."
http://www.pcworld.com/businesscent..._dead.html

Windows 7 has been suffering from a very slow adoption rate. Windows 2000 had the highest adoption rate, capturing almost 80% of the NT 4 devices within about a year. One of the key drivers for this is that it was a "give-away" release. Microsoft had promised corporations with millions of users that they would be entitled to free upgrades to Windows NT 5 if they purchased NT 4 in 1997 and 1998. Windows 2000 was released in early 2000 and supported USB, had much better performance, made much better use of memory, and didn't need memory upgrades. The upgrade was a drop-in, and as a result, most of the NT 4 machines were upgraded within 12 months.

When Windows XP came out, Microsoft had a problem. Corporate customers liked Windows 2000 too much, but they didn't have to pay for the upgrade. Microsoft was only getting the money for the annual support contracts. They tried to increase their revenue by pushing out Windows XP, and telling corporate customers that if they did not upgrade immediately, they would not be supported anymore. Furthermore, Microsoft dropped support for NT 4. Microsoft gave them a deadline, by which time, they had to accept the licenses, even if they didn't actually do the deployment of the software. This allowed Microsoft to claim the successful sale of millions of licenses in the first month and the first quarter, but deployment took much longer, almost 2 years. When SP2 was released in 2003, it broke so many third party applications that many corporations and organizations decided to take a closer look at Linux again. Many corporations first started looking at Linux in 2001 when Microsoft insisted on the forced upgrade to XP at as much as triple the annual rate for support for service that they would have gotten with the OEM license in the first place, such as automated updates and security fixes.

Microsoft had already taken a beating in the server market, with many organizations converting as many of their Windows NT servers to Linux when Microsoft substantially raised the price and CAL requirements, boosting the license costs for a Windows server from around $1500 per box to as high as $20,000 per processor. Many corporations converted many of these Windows servers to Linux servers. Microsoft was concerned that a push of Windows XP SP2, that broke all that third party software would result in a mass defection to Linux. Microsoft executives personally engaged in orchestrating the take-over of Caldera, firing Ransom Love, and replacing him with Daryl McBride, then having McBride attempt to create mass panic around using Linux by suing a bunch of corporations that used Linux, and suing IBM for $3 billion. It was a suicide lawsuit, and McBride knew it. On the other hand, it was also the perfect opportunity for a "pump-and-dump" stock scheme. McBride got 1 million shares of SCO stock options at a strike price of 75 cents/share instead of a salary. His lawsuit of IBM bumped the price from less than $1 a share to nearly $17 a share IIRC. At that point, McBride and the Microsoft placed CFO began selling off their stock, excercising their options gradually, to get the highest price they could, before it became obvious that the lawsuit had no chance, and that IBM might even end up bankrupting SCO. Even worse, SCO discovered that they didn't even own UNIX, that they only had non-exclusive unlimited distribution rights. SCO could sell as much UNIX as it wanted, if they paid Novell a lump sum up front, and a percent of each license, which SCO hadn't actually paid. Making things even worse, the code SCO claimed IBM had stolen from them, was actually code IBM had been using in their mainframes for decades - and SCO claiming to own that code meant they had admitted stealing it from IBM.

Desktop Linux on Life Support:

http://www.techradar.com/news/softw...ad--961508

This is probably true, because desktops are dead. Nobody wants a desktop anymore, they might want a laptop, that's big enough and powerful enough to run BOTH Windows AND Linux, but nobody wants to be chained to a desk anymore.

Seriously though, at this moment I'm running a laptop with 6 gig of RAM, running Windows 7, a dual-core CPU that can crank out nearly 8 billion instructions per second, and also runs Linux in a virtual machine. I can use VMWare, VirtualBox, or Virtual PC to run Linux and Windows simultaniously. I could open up browsers in both systems, but why burn extra CPU on virtualized Linux web browsers? My other laptop runs Red Hat Enterprise Linux, and runs Windows XP under KVM. It only has 4 gig of RAM, a slower CPU, and slower hard drive, but it outperforms the Windows 7 box, and has none of those annoying pauses where Windows just freezes. The other thing I've noticed about Windows 7 is that it gradually slows down. I've only been using it a couple of months now, and I've already noticed that it's getting slower, the pauses are coming more often, and are more frequent. I was able to speed things up a bit by turning off all indexing, cleaning out the index, cleaning out browser cache, and cleaning up some other files, but it didn't take long for the once snappy and responsive machine to get painfully slow again.

Vista was a nightmare. I really thought that Microsoft might have completely blown it. Steve Ballmer's first OS as CEO and he was so obsessed with the game machine features that he dropped the ball in the things that mattered most to corporate customers. Fortunately, Microsoft has been able to turn around and Windows 7 is a vast improvement over Vista. Unfortunately, there isn't the same overwhelming improvement over Windows XP. To get the same performance you get with XP running on 1 gig of RAM, a 5400 RPM hard drive, and a single core CPU in 32 bit mode, you need Windows 7 64 bit, with 6 gig of RAM, a hybrid 7200 RPM hard drive, and a much faster CPU. Worse, productivity doesn't improve. You still have the limits of the PC applications, and if you want better enterprise applications, most of those are simply web interfaces to Linux or UNIX servers anyway, so there is very little marginal gain.

For the productivity gain, you need a Mac. The higher resolution display, 1900 x 1200 means you don't have to dedicate the entire screen to one application, you can have a couple of different applications side-by-side. And having Windows and UNIX on the same system means you don't get those nasty pauses and waits when your are using your web browser to access those corporate applications.

All of this makes counting Linux a bit tricky. The surveys that Microsoft likes best are the 'All or nothing' scenarios. Which OS was installed when shipped by the top 5 OEMs - gives Microsoft 99%. If you add the Mac OS/X to the mix, you still get 99%, because Windows is installed on most of the Mac machines in one form or another. The most popular is parallels, which allows you to run Windows as a virtual machine on the Mac. But since we only care whether the PC had Windows on it when it was shipped, all those Macs get counted into the percentage.

Another popular tactics is "which IP addresses use Windows" - StatMarket and some of the other popular WinTroll quoted surveys like this tactic. In this case, if an IP is ever used by Windows, then that's all that gets counted. There might be 3 Windows machines and 5 Linux machines hiding behind that NAT firewall, but they will be counted as Windows. The only way Linux gets counted is if it's the ONLY operating system to used that IP address for the entire sampling period. The same is true with DHCP addresses. Is it any surprise that Linux would have such a small percentage? Even if a PC runs both Windows and Linux at the same time, on the same machine, the Linux wouldn't get counted.

Another popular measure in the server market is license revenues. Linux revenues are based on support contracts, the licenses are free. Ironically, if Microsoft demanded 1% of the license revenues for Linux, that would still be zero. If you buy a magazine that includes a DVD loaded with Linux, you paid for the magazine and the media, and the documentation - not for the license. Gartner eventually opted to start counting support revenue, but this means they also had to include support revenue for Windows as well.

The irony is that even though it may look, based on these surveys quoted above, like Linux has failed, but there are billions of Linux devices out there. In fact, there are far more Linux devices out there than there are Windows devices. You just don't see tux prominently displayed on the screen somewhere. Instead, you see a digital picture frame, an HDTV display, an Android Robot, a web page, e-mail, and "connection ready".

You may see that Windows logo in the bottom left hand corner of your screen, but there is BSD UNIX code in the kernel, the network drivers, and some applications. If you are using WiFi, the hub is probably powered by Linux, if you are using a cell phone, the phone may be running Linux or Unix, the tower is probably running Linux or Unix, the switching routers are Linux and UNIX, the firewalls of the servers you reach are probably Linux or Unix, and even if the Web Server itself is running Linux, unless it's just a vanity site, there's a pretty good chance that there are back-end servers running Linux and/or UNIX that are generating that content for you.

If Windows was suddenly destroyed, and Microsoft went bankrupt, or had to stop producing Windows, it would be possible to transition to UNIX or Linux in a few days.

On the other hand, if Linux and UNIX were suddenly destroyed - say by a massive virus that would kill every computer than ran Linux or UNIX, the consequences would be devastating. The power grid would shut down, telecommunications would stop instantly, satellites would go out of alignment, Nuclear reactors would go critical, airplanes couldn't land and would be falling out of the sky, you wouldn't know about it because your TV wouldn't work, the digital decoder wouldn't work, the cable box wouldn't work, the cable TV network wouldn't work. The radio stations would shut down, the banks couldn't do business, the credit cards would be useless, the cash registers wouldn't work, the GPS wouldn't work, and many cars would stop running or couldn't be started.

Even worse, fail-safes would be triggered that could lead to nuclear exchanges, but the guidance systems might be unstable, so countries that launched could end up bombing themselves.

Transportation would shut down, and eventually civilization would break down into gang rule. In less than 12 months, we'd be in the "Dark Ages", most of the people who understood technology would already be dead. Only the strongest and most violent, the most criminal element, would survive. Due to lack of food, clean water, working sewage systems, and climate control, disease, starvation, heat stroke, and freezing would kill most of the population in just a few weeks. The rest would resort to cannibalism, eating the dead, killing the weak so they can be eaten while they are still relatively "fresh". Anything that would burn would be used in a few months, so winters in the northern latitudes will kill most of those who did survive.

And you might be able to edit your Microsoft Office documents for about an hour, until the battery on your Windows 7 laptop ran out of juice.

flatfish+++