PDA

View Full Version : Microsoft Just Keeps Comming Back For More!



Sirin
February 7th, 2006, 10:06 PM
Microsoft has just made a new page of their FUDded "Get The Facts" campaign.


Windows Server: Less Maintenance. Fewer Patches. No System Failures.
Security Innovation compared Windows 2000 Server (upgraded to Windows Server 2003) to SuSE Enterprise Linux Server 8 (upgraded to version 9) in a simulated, real-world environment.

Three experts in each system were required to perform ongoing tasks associated with changing business requirements and upgrading systems running e-commerce applications. The results speak for themselves:
• Linux had 14 system failures, Windows Server, none
• Linux required more than 4 times as many patches as Windows Server
• Maintaining and enhancing functionality took 68% longer with Linux than Windows Server

http://www.microsoft.com/windowsserversystem/facts/lp1/01/default01.mspx

:evil: :evil: :evil:

xequence
February 7th, 2006, 10:11 PM
'Tis annoying.

On the desktop linux has only ever crashed/froze on me once.... When a GDM theme messed up and it went to default: it just froze at the default screen and I couldent type anything in. It did unfreeze however =P

Windows has crashed more, but not enough to be a problem...

But those figures are like... "Linux had 0 crashes, cant hurt to round it up to 1...". Then someone else in microsoft gets the report and rounds it up to 2, etc.

tseliot
February 7th, 2006, 10:27 PM
The comparison is between Windows Server 2003 and Suse Enterprise Linux Server 8.

I don't like Yast (Suse's package manager) and I had problems with it (it was Opensuse though).

They should try Fedora or Red Hat (which is aimed at servers).
BTW Red Hat is the most used distro for servers.

That would be a better choice than Windows Server.

public_void
February 7th, 2006, 10:43 PM
The really truth

Windows Server: More Maintenance. More Patches. More System Failures.
Security holes compared Windows 2000 Server (upgraded to Windows Server 2003) to SuSE Enterprise Linux Server 8 (upgraded to version 9) in a simulated, real-world environment.

Three experts in each system were required to perform ongoing tasks associated with changing business requirements and upgrading systems running e-commerce applications. The results speak for themselves:
Windows Server had 14 system failures, Linux, none
Windows Server required more than 4 times as many patches as Linux
Maintaining and enhancing functionality took 68% longer with Windows Server than Linux

mips
February 7th, 2006, 11:14 PM
If you want to run a server in a enterprise environment use OpenBSD or FreeBSD (which is not Linux by the way) and then compare. MS does not really stand a chance when it comes to stability & security.

Lord Illidan
February 7th, 2006, 11:16 PM
So they take an old version of Suse Linux to test it. Fine...
You know, what.. why not just let them spout their "facts", and get on with our lives.

Sirin
February 7th, 2006, 11:53 PM
So they take an old version of Suse Linux to test it. Fine...
You know, what.. why not just let them spout their "facts", and get on with our lives.

Yeah, but this is stopping people from finding Ubuntu. They might take this to the next level and start doing Anti-Linux commercials.

nocturn
February 8th, 2006, 08:42 AM
Basicly, the study was flawed.

Instead of leveling the odds, they used a commercial web application on Windows that had vendor support for everything.

To make this comparison fair, they could have done some comparison like using OSCommerce on both Windows/IIS and Linux/Apache.

It would just be like comparing Windows with MSOffice to Linux with VI (instead of OpenOffice).

That said, They also included an OS upgrade in the test, which has always worked flawless in Debian/Ubuntu, but I have had trouble doing this on SuSE too. Yast is no match for apt.

nocturn
February 8th, 2006, 08:43 AM
Can you upgrade from one Windows version to another?

I have always had to reinstall in the past, but Win98 was the last version I used personally (at work, we have XP, but I'm not the one maintaining it, I admin a Linux and HP-UX box).

SuperDiscoMachine V.5.7-3
February 8th, 2006, 08:48 AM
Is this the stupid study where they had the Linux admins install a software that was not supported on SLES 8, required them to change the glibc in order to do so and only after some naturally borked their system by doing this had them upgrade the borked install to SLES 9, on which the software was supported?

Very convincing indeed. :-D

nocturn
February 8th, 2006, 08:50 AM
Is this the stupid study where they had the Linux admins install a software that was not supported on SLES 8, required them to change the glibc in order to do so and only after some naturally borked their system by doing this had them upgrade the borked install to SLES 9, on which the software was supported?

Very convincing indeed. :-D

Yep, that's the one.

Virogenesis
February 8th, 2006, 08:54 AM
funny also that microsofr are talking about how easy it is too upgrade from what I remeber microsoft always suggested a fresh install so why do they even talk about how easy it is too migrate to a newer version.

Its sad....redhat are more convincing.... its also amusing that suse supports EB already.

briancurtin
February 8th, 2006, 09:07 AM
same response i give to any thread like this:

MS people say MS is better
open source people say open source is better
independent, unbiased evaluators are never unbiased

most results in these TCO studies were paid for

nocturn
February 8th, 2006, 09:12 AM
same response i give to any thread like this:

MS people say MS is better
open source people say open source is better
independent, unbiased evaluators are never unbiased

most results in these TCO studies were paid for

Off course noone is completely unbiased.
But in any study, you have to define the objective (in this case evaluate Linux vs Windows). This requires that as many factores as possible should be the same or the study is void.

I mean, let's say you compare the performance of two brand of petrol, one would asume that you use the same brand and model of car to execute the test.

In this case, you used two completely different cars to test the fuel quality, so the study is fundamentally flawed.

Leo_01
February 8th, 2006, 09:30 AM
:-k
The study bring alot of odd points.
how do you even define a simulated real-world environment?
diff companies = diff real-world environment...


*Security Innovation report sponsored by Microsoft. See full study for details. Results are representative of three experienced administrators for each platform under evolving business requirements. Maintenance and enhancement phase includes solutions based on Windows 2000 Server and Novell SuSE Linux 8. Full study included migration to Windows Server 2003 and Novell SuSE Linux 9, respectively. Your results may vary.

This got to be the most sickening part.
Most likely the admins are experienced non-linux admins...:evil:

Correct me if i am wrong but does those admins knows that there are other flavours of Linux other than SuSE?
Besides...
They should compare the time taken for a patch to be released by MS and a patch released by Novell...
i am sure it is faster...