Slashdot Mirror
Microsoft-Funded Linux Studies Benefit ... Microsoft
mr.big_pig writes "The Seattle PI had a front page article analyzing the Microsoft's Get The Facts website and related ads compairing Windows to Linux. The short and sweet: follow the money and see just how 'independent' is this research. What caught my eye was that this was on the front page and not buried in the business section."
Microsoft can never pay for a study to be done. Noone can pay to have a study done that involves them or a field they are in. How will the studies be done?
I've definitely noticed in the last few mnths that Microsoft seems to be REALLY ramping up its PR war against Linux. They've been talking about it for a while, and now we're seeing it.
/. story below this there was a link to e-week about the 2.7 Linux kernel, and guess who had a big ad on that page? Microsoft. And the ad tried to show that Microsoft Windows Server is 11-22% faster in '4 out of 5' workplace scenarios than Linux.
On the
Even Slashdot has been running Microsoft ads, and almost any tech news site you go to is crawling with them. Microsoft has a definite advantage against Linux when it comes to ad budget, as only IBM seems to be really pushing Linux in terms of PR and advertising.. and even then it's more about IBM's solutions than Linux, which is not surprising really. And so Microsoft is going to continue funding studies and surveys, slightly tweaking the questions to favor them ("How easy do you find it to connect to an Active Directory from Linux?"), showing the world the results which are good, and dismissing the surveys which are bad.
I wonder if there are any Linux mad advertising zealots with deep pockets to get some ads on those sites, and to generally kick up a stink and get us lots more stories in the papers and magazines. This is a PR war, and if you're a Linux devotee, make sure you fight back against it in some way (even if it's just winning your clients over to Linux even more).
"Dan Leach, group product manager for the Microsoft Office System, was asked at the time whether Microsoft would have publicized the results if they hadn't been positive. He answered that he had been so confident in the software's benefits that it "was never going to be a question." Is MS smoking the same stuff that SCO is??!?
First of all, any vendor TCO study is going to be completely bullshit. However, there's a glimmer of truth in the Microsoft stuff:
+ Realistically, the software & hardware costs aren't going to be significantly different between Windows and Linux. Yes, you can download Linux for free, but your boss is going to pay real money for RedHat or SuSE.
+ Unix admins are more expensive than Windows admins, although they generally have a much higher skill level. Maybe as Linux penetrates the market, this will equalize (both in cost and skill level).
+ MS selected specific scenerios to favor them. For example, File and Print have never been a strong spot for Unix -- Novell and MS have owned that segement for years and years. It will be interesting to see what Novell/SuSE puts on the market.
And attacking Linux on Mainframes is like hitting the broad side of a barn -- There might be some scenarios where it makes sense, but for the most part a mainframe has pathetic price/performance and is very expensive to keep running. (Although, that wouldn't stop IBM from selling you one.)
And as for J2EE -- some of the tools are ridiclously expensive, so that's a pretty easy cost study to rig.
Whenever I hear the word 'Innovation', I reach for my pistol.
I've refrained from posting on any garbage like that forever, but since not much people are posting (sick of it already, or too busy combing thru all the reports?), I'll chip it one.
Disclaimer: I've not read all the reports, just the article and the IDC report.
Microsoft's Taylor said that findings are also presented in such a way that they can be duplicated by others. I'm not sure. The reasoning for the quantification method is weird at best. On p.10 of the IDC report, every item for Linux is more expensive.
Let's take hardware. The hardware for Linux is more expensive because it is assumed that for the same hardware, Linux can handle less load than Windows, therefore, you need more hardware if you deploy Linux, hence higher cost. That's weird, how did they come up with that assumption? It's certainly not explained in the "open methodology".
Software: how did they come to the conclusion that Linux softwares are more expensive? I can't find the list of comparable softwares they used in their study. If this methodology is really open, let's provide the data, shall we? And they claim that Linux is used mostly for print, file serving, and web serving. Well, if that's the case, the softwares for those functionalities cost almost nothing, except for support, which is more or less the same for both platform. How come I remember I used to pay thousands of dollars for a Windows Server allowing only 5 connections?
Staffing: Sure, Linux/Unix admin are more expensive. That's true only if you assume that each Linux/Unix admin can only do the same amount of work as an MCSE monkey. You draw your own conclusion.
Downtime: Whoa, Linux cost more for downtime (in a couple of cases)? Real data please?
Training: That, I'm not sure. It's probably easier to pick up Windows, as every new kid is already familiar (more or less) with windows interface already, before the training? Ok, let's say the data here are correct, but I still want data.
Outsourcing: I can't seem to understand how did they come up with that conclusion. I'd like to see the raw data.
The funny thing is this: the report said that Linux is used only for "light workload on the edge", and not for the real stuff.
Hmm, I guess they didn't talk to the CIO of amazon.com (hint: based on their previous experience with Linux for other things with a $16M cost savings, they are moving their mission-critical terabyte database to Linux!)
Not to mention... How the studies are conducted... Mostly when MS is comparing Windows up against its like comparing Apples to Wax Apples... They can apear to be the same... But How they are used are 100% diffrent. The one study goes on comparing multiple windows boxes against a linux mainframe consolidation server... Umm... Well Linux can run on any platform windows can... why not use the same boxes? (Because it wouldn't produce favorable results for MS)
Where as I think most of the linux vs ms studies that get done are alot more closer to real apples vs real apples. But I haven't seen very many studies at all in detail poing linux over windows.. Its just my guess they are alot closer to a direct comparison than what I have seen in windows vs linux with windows on top.
Who needs WiFi when we can have Packet Over Sheep! http://datacomm.org/PoS-InternetDraft.txt
Just 30 posts so far, hours after an article on MS is put up on Slashdot. Guess it implies that study reports (funded, sponsored, sexed-up, or otherwise) carry little weightage with IT consumers, these days.
This could explain why Linux adoption continues to increase despite all the media hype and study reports - users and organisations are probly doing the study reports themselves..... consumers getting wiser is a highly undesirable phenomenon for the Corporat types - I think we'll soon see Ask Slashdot article on "How to Keep the Consumer Stupid?"
-
If you keep throwing chairs, one day you'll break windows....
It don't understand this. M$ can't be _that_ stupid, can they?
Step number one was completly negleting OSS and hoping customers wouldn't notice. *That* was a time when M$ should have prepared to sell it's own Linux distro with DX 9 and some other embrace and extend stuff. They missed it and screwed up. Lucky was we.
Step two was bashing the GPL as 'unamerican' and other bullshit and bringing customers to look twice at licensing where they used to give a hoot about the small print. Thus causing them to also look at M$ licenses and notice what BS they have been subscribing to for years allready. Ballmer backed of merely a half a year later and admited it was a bad plan to draw so much attention to OSS by bashing Linux/GPL in such a way.
Step three: Publish studies were everybody with more than 2 braincells notices in an instant that Linux/OSS is on top of things and M$ knows nothing other to do about it than flail the bullshitting-club left right and center.
Can a company of this size with marketing departments on a budget as big as the anual throughput of something like the third of afrika be so stupid and windows focused to pull such a mindless stunt?
Honestly, if I were a stockholder of M$ I'd be somewhat pissed and would want a question or two answered on that matter. M$ better get a grip and start preparing to change their business model or else they're gonna be in deep shit faster than any of us had ever hoped for.
We suffer more in our imagination than in reality. - Seneca
There is approximately 42^42 reports published yearly on any given subject. If you cannot find your truth in one of them you just don't know what you want your truth to be like.
All Windows vs Linux researches are meaningless because of one simple fact: Microsoft has billions available to throw at making Windows better, whereas the Linux community, even with RH and other prominent companies, is rather poor in funding in comparison.
If I was an IT manager, I would consider Linux to follow a "best effort" sort of approach, whereas Windows, being far less than free, should come with guarantees about stability and performance. However, at least in terms of stability and security, Linux certainly outperforms Windows, if not by much these days. And it probably isn't far from Windows in terms of raw speed.
So, basically, those studies compare apples and oranges. In the cycle racing world, if Joe Bike, with no funding, no top-level medical staff and a simple bike, was to finish a race second or third to Lance Armstrong, everybody would herald him as a great hero, and Lance Armstrong's team wouldn't put up a website pointing out Joe Bike's poor performances compared to Lance Armstrong. Fact is, Microsoft has much vaster resources than the Linux community, and only manages to perform marginally better than Linux at best. That's what the studies should show. Throw Microsoft's billions at Linux and Windows would instantly become a laughable joke.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
Looking at the
.NET vs J2EE/Linux study
.NET and SQL server'. And it probably is true. But that is what comes of not embracing open source more fully. Adopt JBoss instead of WebLogic, save nearly $160K. Adopt Postgres or MySQL instead of Oracle, save $40K. end result: open source wins hands down, provided development costs are roughly comparable.
The reviewer compares the cost of WebLogic+Oracle versus Windows Server+ SQL server. While the OS is much cheaper (and they omit costs of securing the platform against repeated worms), it is the cost of the proprietary software that gives MS its 25% cost saving.
The thing is, the cost of the app server and database are huge; they dwarf everything. So a large size company would only pay $5K for Redhat versus $40K for windows, but then pay $160K for WebLogic and $40K for oracle (versus $0 and $20K for the MS solution). And of course the annual maintenance fees are simply a fraction of the software costs, so they are more on the j2ee system.
Really the survey says 'J2EE using Oracle and WebLogic is more expensive than
So yes, the study was utterly rigged. It makes a valid critique of using WebLogic and Oracle, but says nothing about Linux/JBoss/mysql.
Adopt JBoss instead of WebLogic, save nearly $160K. Adopt Postgres or MySQL instead of Oracle, save $40K. end result: open source wins hands down, provided development costs are roughly comparable.
You've got to be kidding me. Postgre, while a capable midrange database is nowhere near the performance or capability of Oracle, and MySQL isn't even in the same universe given that it lacks such basic features as stored procedures and transactions. This isn't even counting all the "value add" that oracle provides with their support tools and programming environments.
If your needs can be met by MySQL or Postgre, sure... don't buy Oracle, but then you'd be stupid to buy Oracle even if those open source tools weren't available. You'd probably use Access on Windows instead.
As for JBoss vs. Weblogic, i don't have enough experience with either to make a valid comparison, but Weblogic is ceratinly a much more capable product by features alone.
If you need web hosting, you could do worse than here
The real problem is that the studies are geared toward benchmarking the few strengths of the Windows platform. A perceptive reader would note that the IDG study actually confirmed that Windows was significantly less reliable that Linux.
The IDG study mentioned that among file and print servers, Linux servers on Intel platforms averaged a higher workload with a lower failure rate. However, faithful to their master, IDG goes on to contradict itself by noting that Linux admins get paid more than Windows admins. Which means that taken on a per machine basis, Windows is cheaper, but when taken on a workload volume basis, Linux is less expensive.
The problem with such studies is that they are slanted toward the situations in which Microsoft's products do perform reasonably well. Consider for example the Windows-server against Linux-on-mainframe benchmark: a totally useless comparison. In the first place, companies don't buy mainframes for web-serving; they buy them for corporate datacenters. Then, when they want to provide web functionality, they either augment with Windows boxen which must connect to the mainframe for database access, or they run Linux servers on the mainframe. The first case involves hiring additional Windows admins, the second, merely training the existing mainframe systems programmer on Linux. Furthermore, you will never find a situation in which a company's mainframe-based webserver is outperformed by a Windows box. The reason? In real world corporate environments, business critical data is always stored on the mainframe simply because it is the most reliable platform. Thus, the screamingly-fast webserver on a windows box can never run faster than the mainframe simply because it must wait on both the mainframe database and network latency when filling requests.
In reality, the studies are worthless because they simply don't address the manner in which businesses actually use the systems. They ignore the crucial questions of reliability, robustness, compatibility, and support.
The society for a thought-free internet welcomes you.
Here's a fair way of doing the comparison...
Have Microsoft and a Linux sponsor (IBM?) each have a competition to pick a team of 6 knowledgeable Windows admins and 6 Linux knoweledgeable admins.
Put two identical empty servers in a room in a neutral place with an independent analysis company.
Give the two teams all the tools they need and, say, 24 hours to build their respective Windows and Linux environments - even allow the Linux team to build their own custom kernel / distro if need be.
Then do the performance testing on the servers.
Result:
1) Fair test results that will probably show Windows is better at some things and Linux at others.
2) Microsoft gets some glory showing a willingness to compete in an "open" trial.
3) Microsoft and Linux both end up with "things to do" to improve their software.
4) We all benefit as a result.
5) We all stop bickering over a marketing campaign that is no different to Mercedes comparing its cars to BMW or Macdonalds comparing its burgers to Burger King.
Gentoo Linux - another day, another USE flag.
This is voodoo art. Statistics are not a scientific analysis of facts. Any function that takes one data set and can produce two diametricly opposed results is not science. That would produce one answer. Statistics is nothing more than rumor, gossip, and bullshit! Just like the current version of economics, it was created as a separate field in the fifties by some out of work mathamaticians to create jobs for themselves.
There are lies, damn lies, and then, statistics.
Professional Politicians are not the solution, they ARE the problem.
So how about it? It looks to me like Microsoft may have better throughput on these tests. Who can prove them wrong?
I like my women how I like my sugar.. granulated.