Slashdot Mirror
Yankee Group Survey Says Windows, Linux TCO Equal
prostoalex writes "A new survey by Yankee Group analyst Laura DiDio shows Windows and Linux are viewed as equal by U.S. businesses. In the eternal OS wars, '88 percent of respondents said that the quality, performance and reliability of Windows was equal to or better than Linux.' Companies were also asked to rank the operating systems on security. On a scale of 1 to 10 'companies rated Microsoft's security at 7.6, double the rating in a similar survey conducted last year. Linux's rating was mostly the same at 8.3.' Conclusion? 'DiDio said that most companies -- whether large or small -- rarely take the huge step of replacing one operating system with another. Instead, they usually add a mix of Windows and Linux server software to expand functionality.' Microsoft used last year's Yankee Group survey results in their Get the facts campaign."
That's not something you see very often, usually its a landslide one direction or the other, depending on who did (or didn't) pay for the study.
From my experience, this seems to be fairly accurate (as far as company's interpretations). Can anyone else back that up?
"Instead, they usually add a mix of Windows and Linux server software to expand functionality."
Thus, they have the ability to directly compare between both. If they find Linux to be infinitely better, they would switch. Different tasks -> different tools, however, so they use both.
webpage
DiDio is a total shill for Microsoft. I don't know why /. dotes on her every word. She isn't an unbiased source, y'know.
The non-biased information all says the obvious: Linux has TCO ownage on Windows. That said, I'd like to see a TCO study where Linux and Windows are compared to MacOS X, especially now that Apple has a relatively cheap model that could be a great replacement for enterprise desktops.
Knowledge is power. Knowledge shared is power multiplied.
In most cases, both Linux and Windows are growing at the expense of Sun Microsystems Inc.'s (Nasdaq:SUNW - news) Unix-based servers
DiDio said that most companies -- whether large or small -- rarely take the huge step of replacing one operating system with another. Instead, they usually add a mix of Windows and Linux server software to expand functionality.
I know all these frontpage stories are framed in terms to churn up a large number of comments, but these quotes have always been true. For all the companies I've worked for/with it has been a mix of windows and linux/unix. The bigger the company, the more diverse the mix. There are actually running business systems that predate both unix and windows. Over the decades linux is taking the place of the Unix boxes. Windows largely rules the desktop. Linux largely rules the server room. And windows quality has gotten so much better with 2003. Windows being better is not a slam to Linux and does not threaten its ascendance in the server room.
I hate to say it, but these Linux vs. Windows stories have been the same fucking story forever. The only thing that changes is the write up.
I don't even have a problem with the price. What's keeping me from ever using XP is Product Activation. I hate the fact that I'd always be treated like a thief. That'd I have to essentially ask permission to change my hardware (which is something I do quite often, I go through at least three motherboards a year, at the least.)
That's the exact same reason I never bought Half Life 2. I just hate the idea that someone has control over my computer other than me.
That's why I'll eventually switch to Linux. I've been trying Suse 9.2 with great results. And since I have kids, I hardly ever play games anymore, so that's no longer a hinderance.
(I'd love to switch to OSX but I'm too much of a hardware junkie to switch to Apple. I love building them myself, and Apple won't let you do that.)
If someone says he and his monkey have nothing to hide, they almost certainly do.
I'm just tired of DiDio. Isn't she the one who said just last year, that it will take Linux close to a decade to even be percived to be competitive to Windows? I better become an analyst to. How do I start? And there is Gartner too spewing "facts" about Linux all the time...
"I'd say they are equal if you discount price, which this survey did."
Seconded. After the switch to NT, Windows was actually quite usable for both servers and workstations at my previous job. Our IIS webserver had an uptime of over 180 days. (Pity we had to move it, I'm curious how long it would have lasted.)
When the engineers switched to Linux, though, there were all kinds of stupid problems getting it working. Some of them were networking, some of them had to do with flawed implementations of stuff we needed to work. (i.e. on dual-proc machines, the clock would sometimes jitter back and forth a second or two.)
I'd like to mention a couple of things, though, since dues with mod-points are often vindictive when legitimate complaints about Linux surface:
1.) This was a couple of years ago. Those problems may or may not still exist. I think it'd even be fair to say that most of the problems were likely unusual. The workstations were both development stations AND custom software was being written on them.
2.) Some of the networking problems we had may not necessarily have been the fault of Linux on the workstations. It was, however, very difficult to tell. I remember watching the engineers googling for various networking tools just to narrow down the list of suspects.
"Derp de derp."
It's really too bad that we don't have access to the actual study. Without it it is hard to judge very much. I went to the Yankee Group web site and found their press release, which is a little bit more informative than the news item, but not much. Elsewhere on the Yankee Group site they reveal that the study will not be available until JUNE 2005. Funny that they are issuing press releases now about a study that won't be released for two months. I wonder if that is so that they can have their impact now and defer the hard criticism?
Anyhow, there was an interesting bit in the YG press release:
I think that this gives us a hint of what is going on. If MS Windows were really perceived as better than Linux, or even equal, the cost of making a change and general inertia would presumably result in little Linux adoption. The fact that the same businesses in which MS Windows has an overall reputation of being better than Linux are adding Linux or shifting partly to Linux suggests that there is actually a perception of Linux as better and/or cheaper. I suspect that what is going on is that the reputation questions were answered largely by managers with little firsthand technical knowledge, who have, however, been pushed by their technie subordinates to allow a shift in the direction of Linux.
A mixed environment is the best environment. using one or the other doesnt make sense.
There is a thing called a Cascading failure that has been hypothesized (sp?) over about using a single Os for everything.
One hole to bring them allll down.
Think smart.
In terms of security, Yankee Group's survey showed a sharp rise in companies' assessment of Microsoft's security level, bringing it closer to perceived security level of Linux.
May be more accurately phrased:
In terms of security, Yankee Group's survey showed a sharp rise in companies' perception of Microsoft's security level, bringing it closer to the assessed security level of Linux.
I think you underestimate just how much I just dont care.
Who the hell are "the respondents"? What was the methodology? What was the exact wording of the question?
The list goes on
Sorry, I conduct research for a living. This kind of drive-by "journalism," simply report-what-the-findings-were reporting pisses me off to no end.
Lorem ipsum dolor sit amet
Linux and Windows having perceived security ratings of 8.3 and 7.6 isn't interesting, but what is interesting is that Windows perceived security rating doubled in one year. It may be time for Linux advocates to find another argument besides "security". The "stability" argument is no more, the "security" argument is losing steam, so I guess "price" is the only argument left (although, if the perceived TCO is the same for Linux and Windows, even the "price" argument doesn't have much steam).
-- "I never gave these stories much credence." - HAL 9000
I've lost count of the MCSEs I know that LOOK for a reason to deploy Linux and run it at home.
This guy is way out there
Seriously, this is so void of legible response that it makes it painful to read.
Your conceptions are clearly misguided on the basis that you seem to not know what you are talking about.
This is obviously not a jab at you personally or professionally. I'm sure you are extremely qualified and good at whatever it is that you do.
However, you are clearly lost. Bill Gates personal life and his business acumen and behavior are two completely different things. Praising the man for his general charity doesn't expound to his or the companies he works for behavior in the computing industry. Please learn to seperate the two. That said you would do yourself good to try other operating systems.
Hosting via Windows is russian roulette. I say this because i've done the real life test myself. I'm no windows professional. Infact, in Dec of 1994 I stopped using windows when an OS/2 warp disk of mine died and an IBM engineer at the time gave me a copy of linux. There are so many black magic items in Windows that it would take an increasing amount of time for me to learn them all. Windows isn't easy to learn, it's easy to click alot of buttons and try to get it working but that doesn't enable me to understand what is going on or what is happening with my computer.
This has lead me to believe that windows administrators simply don't care about their systems enough to know how they work. They just want a patch or a quick fix or to press a couple of buttons and reboot. I've even extended myself to try and find a good windows administrator to learn from. I hate windows, but I'm clearly trying to understand where the low cost TCO and ease of use come from. If anyone is willing to help with this please feel free to contact me.
The documentation via Microsoft is often incorrect or not detailed. In most cases the behavior exhibited isn't what the documentation is really for and/or there are completely missing chunks of steps via documentation. An example of this would be smartcards. In Unix land, these problems are mitigated by having the source. Also, man pages usually are exacting, so my questions are answered. This isn't available via windows.
I switched over a heavily trafficed site to Windows based on contract purposes and it crashed, repeatedly. Why? I'm still amiss as to why. There was simply no way to fix this. Microsoft said they would look into this and I'm not sure if they ever did. My caring well ran dry way before that. I took the same site and put it on a Solaris box, that was in 2001. I haven't touched it since. It's still running from the time I powered it on, this is about 3 years and 9 months later.
So my experience with Microsoft has been the complete opposite. I'm not sure they've made a positive difference in the computing industry. Even with the low cost of hardware which can be traced back to hardware manufacturers and competition. Microsoft hasn't provided software that changes people lives or allows for general productivity in the work place. Computers are to be aids, tools to real life work. Has Microsoft changed my life or allowed me to be productive? To date, no, they've allowed me to be less productive. To get less work done. I spend more time fiddling with windows machines than trying to work on new interesting things. It annoys me.
In comparison, if I put a unix box up to do something. I walk away, and usually I don't come back. I go on to other things.
Unix/Linux/Open Source allows me to learn to enable productivity. It allows me to get my job done and go home and enjoy other things.
I used to subscribe to that newsletter because it has some interesting stuff in it regarding the windows world sometimes, and I'd say that it is very pro windows, but not as anti-linux as you might think.
I work in a Microsoft shop. We've got 600 servers, all Windows 2000 and 2003. We use the hell out of AD, we have 50 sites, and over 6,000 workstations. It's not a huge company but it's formidable.
Anyways, the bane of any Windows IT person these days is patches. Almost every single Windows patch requires a reboot. Even some of the IE patches.
Microsoft releases new patches every Month. On Feb 8th they released 12 of them. That's 12 patches we need to determine if we're going to install, test the ones we do, and deploy. We have tools to make this easier but the server environment is so diverse that it's a daunting task to face every month, with the coordination and downtime windows. Fortunately, they released none in March. But they usually do.
You could never run your IIS server for 180 days now, because you'd be patching it every 30 or less. I understand the need for the patches, and we do them to maintain the system, but if the system were a little more secure in the first place it wouldn't be so bad.
I really think they should tally up the patches for each system based on whether or not reboots are required and see how many Linux has versus Windows.
- It's not the Macs I hate. It's Digg users. -
I have done exactly what you are talking about. Taking a gaggle (or is it a murder?) of sales drones who know jack about computers and need to have a room full of computers they can log into when they are at the office, without it mattering which actual computer they use, profiles/home dirs are on a server.
The machines cost $200 each (walmart specials), and their cost to have me set it all up was $200 (2 hours work). It was 12 machines, but for comparison lets pretend it was 10 machines. That would mean $2200 vs windows $3790. No training was needed, these guys only applications are email, opening a single template word doc and filling in the blanks with customer details, and using a browser to look up and print maps on google to get to their sales appointments. Their desktop had nothing but "Email", "Internet", and the same bizzare name for the word doc they had always used. They had no problems and never needed any more help than they did with windows.
Yep, you hit the nail on the head. Serious scientists call these "researchers" who perform studies for-profit and for interest groups "whores". If you pay for a study, even one whose results can be objectively measured and the study is conducted properly, the official conclusions and the way it is presented to the media will push the desired outcome even when not borne out by the numbers themselves. An average perceived security rating of 7.6 vs 8.3 may actually be a signifcant difference if the survey was large enough but because this is obviously a Microsoft commissioned study as the last several DiDio/Yankee Group studies have been that difference is called insignificant and the conclusion is the two are not just percieved as having similar levels of security but in fact do.
Other than SCO and Darl McBride, I think DiDio is probably held in the lowest possible esteem over on Groklaw. They quote her a lot, and she seems to get it wrong nearly every time. The opinions that I have read by her are consistently pro-SCO, pro-Microsoft, and anti-open source, to the point that I don't think she can be considered an even remotely reliable source.
So it's particularly interesting that "TCO is equal" is the best she could come up with. If that's the best they can manage, it's a huge win for Open Source. When TCO is equal, why on earth would you pick the software that costs more up front?
The claim must be that Linux costs more to run, since it's free to install. That was the exact method that Microsoft used for ages to get ahead in the market... it was cheaper up front but cost more to run. That can actually be a very smart business decision, since presumably you'll have more money later than you do now, particuarly if your business is just getting off the ground. (That's part of why leases do so well.)
Of course, we all know that Linux is probably cheaper to maintain once you have the skills to do so, possibly by an order of magnitude, due to the absolute control you have over the system and the enormous power of the built-in scripting languages.
But even if you grant that it's more expensive to run, this study shows that Linux is a good choice for many businesses, particularly small ones, or companies growing very quickly without a lot of capital to do it.
The current direction of Windows reminds me of that old quote, "Those who don't understand UNIX are doomed to reinvent it - badly", although all things considered it may not be entirely accurate.
For many years now Microsoft has been patching NT, a single-user system only really suitable for small local networks, into a multi-user system that can cope with many large networks. All the while they're making an effort to retain backwards-compatibility all the way back to MS-DOS, and to avoid doing anything that makes them look like UNIX. This obviously adds up to a very troublesome equation, even when one does not consider the hybrid "Windows 9x" branch (thankfully abandoned, but not yet gone or forgotten).
Now, a few years ago the swiss-cheese nature of the system became so apparent that the general public grew aware of it. That quickly resulted in official denial, then a plethora of security-oriented PR campaigns and, a bit later, more intensive patching. It has recently resulted in some actual improvement, but it's still being made by attacking the effects and not the problems themselves. Longhorn naturally promises to fix everything.
Take a look at all the new toys in Solaris 10, as well as SELinux, and you'll notice that the UNIX world is not just standing still either. There's just not as much noise to be made about being relatively secure compared to the previous version.
The joke is on Thompson, he reinvented VMS badly.
For many years now Microsoft has been patching NT, a single-user system only really suitable for small local networks, into a multi-user system that can cope with many large networks.
Windows NT has always been at core a multi-user operating system. The kernel architecture is derived from VMS which is itself heavily influenced by Multics and ITS. This is not really a surprise since there are not all than many people who have designed O/S and pretty much everything has its roots ultimately in Multics and project MAC.
The problem that faces both UNIX and Windows NT when it comes to networking is that multi-user security and network security are two absolutely different things. The features you need for one do not help much with the other.
Most production Apache servers run on a system that has at most three active accounts. Root, apache and maintenance. To all intents and purposes the separation of apache and root does little more than help prevent the system partion being corrupted, it does not really do much for security since all the data assets of the machine are going to be accessible from the apache account.
If you wanted to actually use the O/S security mechanisms to bear in a meaningful way you would have to configure the Web server to respond to data access requests by spawning off a new process and locking it down with the appropriate system privs each time a privileged access was performed. This is technically possible in both Unix and Windows but it will grind the machine down if you try it with any appreciable load.
Looking for an Information Security student project suggestion?
Try http://dotcrimeManifesto.com/
Price isn't the only aspect of the idea of licensing software. I think many people don't realize how easy it is to get hardware, and how fluid the hardware situation can be.
Say you have some people doing a data entry job. Say that for whatever reason, you have a sudden excess of data that needs to be entered. With a Linux set up, you could take an old computer, put together a terminal, and have someone enter data for a day, and then you can throw that computer back in a closet. With Windows, you would need to buy a seperate license for that computer, even if you were using it for a day. Similiar situations exist all over, from small jobs like this, to someone who might have a temporary spike in web traffic for a week, and needs another server to cover it. Dealing with the technicalities of getting a license for these things would be somewhere between a nuisance, and a threat (if you do it wrong).
Many people, especially in management, wouldn't realize this is a problem, because they grew up in an era when your computer hardware was too heavy and rare to move. Now, when you could get a P-266 off of a pile, or at a garage sale, and turn it into a backup webserver in one hour, the entire idea of licensing specific computers makes less sense.
Hopefully I didn't put any [] around my words.
You're right - I shouldn't have taken liberties with Saddam's record in this discussion of linux & microsoft. ;-)
It was an appeal to tradition, as the phrase "call me a point a [sic] click nub" indicates.
While the OP is indicating that windows is easier to use, he is doing so in a way that appeals to what passes for tradition in the computing world.
Pointing it out as an appeal to tradition does not challenge or refute the assumption that windows is easier to use, but highlights the rhetorical technique being utilized.
Now then: "You don't need to remember commands in Linux ever since you could set "Start X on login" at the installation and go straight into KDE or Gnome (etc)." is a counter-argument against windows is easier to use. I've sat more than a few windows users in front of a nice KDE desktop and not one couldn't figure out how to click the mozilla icon. They generally remark on how pretty or cool looking the desktop is and then don't really notice much else.