Submitted by “Marking what could have been a summer-long hiatus in its ‘Get the Facts’ campaign, Microsoft is re-igniting the flames on the argument over whether enterprises spend less to manage Windows systems than Linux systems. This morning, the company touted a study it commissioned from independent analyst Mercer Management Consulting, which made the case that companies that implement migration programs away from UNIX systems based on the need to adopt new applications – what Mercer calls ‘transformational migrations’ – now tend to choose Windows over Linux.” I just bought some salt, now I can put it to good use.
Replace “Microsoft” with “HP”, “Dell”, “Apple”, “Magnussoft”, “Republicans”, “Democrats”, “USA”, “UK”, “France”, or “Paedophiles” and you get an idea of the credibility of the study. But note!: Microsoft has now gone from saying that Windows is *cheaper* than Linux to saying costs are *equal*. That’s pretty significant considering the fact (among others) that *individual* Linux admins are paid *more* than Windows admins.
“IBM commissioned the frances study”
Who commissioned this one?
Apple.
😉
over here we have _one_ guy who admins our email servers (sendmail, cyrus, and a bunch of other related foo running on linux/solaris). he has one person assisting him (who full time does the blackboard admining), and one full time developper that works on cyrus code, from home. licensing costs $0 (ok, the sophos puremessage spam filtering costs aplenty, so you got us there…) we have about 30,000 accounts, give or take, with a growing global presence…
all things considered, you really think you could pull that off with windows + exchange?
And how much would it cost you to replace that _one_ guy if/when needed?
“And how much would it cost you to replace that _one_ guy if/when needed?”
Well, out here in Los Angeles the same as the cost for a Windows Admin..does not matter the OS as the pay is the same for the title of ‘System Administrator’. Roughly $70K/year.
I don’t mean how much the new guy will cost in salary.
I mean how much does it cost to replace the _one_ guy who knows your system in and out, knows where all the hackish solutions are, etc.
Is it so standardised that you can lift the _one_ guy up, and put another admin there and get no additional costs at all?
Take into account also that you may not be able to let the _one_ guy teach the _new_ guy.
This is something I often see missing in TCO accounts (probably because systems that TCO are calculated from in studies already have many admins that can cover up for each other, as well as teach new people in case such a situation comes up).
knows where all the hackish solutions are, etc.
I’m willing to bet a lot of the “hackish” things you are talking about are only “hackish” to an admin who doesn’t know *nix.
The great thing about linux is that it’s a hackish culture. A linux admin generally understands that a setup would be customised for a particular businesse’s needs, looks for the documentation and reads it.
Things hacked together on linux tend to give the feeling that they aren’t really out of place and make sense.
Hacks on windows tend to be very ugly as an admin attempts to get around the structure that windows imposes.
“And how much would it cost you to replace that _one_ guy if/when needed?”
Less than hiring an overcertified and underqualified Exchange administrator.
It depends on how you look at it. Do you want a customized solution or do you want a prepackaged solution?
Your customized solution is working very well for you. A good management decision on your company’s part. Now the other side of the coin is the prepaackaged solution where the admin has to license windows + exchange for X amount of dollars but doesn’t have to pay a developer to customize it. It will comes down to how you want to run the operation.
However, in a lot of situations open source solutions are not customized, but rather used as is.
I always laugh at these studies. I have been a sysadmin for two Windows-only companies and for two Linux/BSD companies. The Linux/BSD companies required less hardware to run and had no licensing costs. I got paid a lot more at the Linux/BSD companies, but I was also required to maintain a lot more, so in the end, the Linux/BSD option was much cheaper.
I’m sure it is possible to have a situation where Windows and Linux would end up costing the same, but that situation, if it exists at all, is a rare exception.
I can one-up you… we have about 45,000 mail accounts here, all running with vpopmail/qmail/courier-imap. Spread out over four systems, only one of which is at all taxed by the load (it’s got the majority of them and is *very* busy with 200MB spools).
Licensing costs: $0
Server maintenance costs: perhaps $500/mo in labour
Spam/virus prot. cost: $2500/yr for Barracuda subscription
We also use Qsheff with all spam protection turned off. It’s job is to have clamav check every incoming message (whether from the Barracuda or from people relaying), which has turned out to work quite well.
Maintenance costs are for the once every few months that we have to compile and install an update to some piece of software (usually Clamav, sometimes the Linux kernel) and costs to back up the mail spools to a removable disk. OS updates are done through apt-proxy, so only one copy of each package comes in off the Internet and then is fed to the machines at LAN speed. I may be over-stating the costs, if I don’t include general server maintenance costs that you would do with any system you might have to look after.
Why Barracudas? We have two BC 400s doing our work for us because they have real-time intent analysis, hourly virus and spam updates, block lists; provide customers with customization and quarantines. All for less than $10K investment and a few thousand dollars a year (that’s for both). They’re also very flexible. I can’t see any better solution for a site of our needs.
We use Squirrelmail for the main server (it’s very lightweight and quick even over satellite connections) and IMP/Horde for more advanced mail features. It provides some collaboration and scheduling features (Squirrelmail just has a nice little editable calendar).
Cost of doing all this on Exchange? Maintenance costs involved? The mind boggles.
Look, regardless of whether IBM or Microsoft or whoever commissions a particular study, that doesn’t mean that we should totally discount what it’s saying. Take it with a grain of salt? Sure. But simply turn our backs and refuse to examine the implications of the study because we don’t like who commissioned it? No. That’s a mistake. Examine the study’s data and conclusions, and form your own opinions. Assess the methodology.
There are some things that I agree with in this study. It has been my experience that, while its initial cost is much lower, Linux administration simply costs more. Salaries for Linux admins are higher, it usually takes longer to coordinate and perform maintenance, and there’s higher complexity, generally speaking. These are all factors which contribute to TCO. What’s not clear to me is whether the cost for porting applications from UNIX to Windows is factored into TCO in this study. That makes a difference.
“The high cost of Linux is misleading, and does not indicate a difficulty in managing JVMs on Linux,” states the Frances paper. Although it says the higher efficiency of Linux-based hardware does enable clients to purchase fewer servers to perform the same workload, it goes on to say the real cost savings is strategic: specifically, by employing fewer people to administer services, and creating fewer administrative departments to be delegated among a larger IT staff.
Look, regardless of whether IBM or Microsoft or whoever commissions a particular study, that doesn’t mean that we should totally discount what it’s saying.
You mean that we should use logic? Methinks you have much too high expectations of OSNews and its visitors.
it’s not clear to me is whether the cost for porting applications from UNIX to Windows is factored into TCO in this study.
License costs 🙂
whas OSS being used on UNIX and will closed source be used on windows?
But simply turn our backs and refuse to examine the implications of the study because we don’t like who commissioned it? No. That’s a mistake.
Agreed,only there isn’t much to study.The paper is very thin.
In addition to the License vs. Labor debate, I believe there’s a different way of thinking between the two camps. When confronted with a software problem, a Windows admin is more likely to purchase a solution. A Linux admin would be more likely to create his/her own solution. Neither of which is *always* right — it’s just the “build it or buy it” dilemma. What does this mean for the TCO winner? Nothing but more complexity in determining the TCO. Regardless of the credibility of the story, I think the result is accurate. In the long run costs of both will be equal.
Well, I think, administration of Linux based servers is much less complex and/or time consuming, than Windows based, although you need more knowledge to get into that state, but this should be reflected by the higher salaries already.
Let’s suppose Linux does in fact have a higher TCO.
So what?
If you really believe something is better, what does cost matter?
I work in a K12 School District. Considering how many PC’s and servers we have spread over 40+ sites, we are spread pretty thin.
If I or one of my colleagues has a problem we can research it and try to fix it. But if it still baffles us, and it is Microsoft software, we can call them for about 250$ and they will work on it until they solve it. Sometimes 2 or 3 teams are called in for help. Sometimes an hour, sometimes 20 hours or more. Same price.
Thats an amazing deal.
If we wanted the same service from RedHat, we would need to run RedHat on each of our 200+ servers at 1500$ per year per server.
What happens if/when they say “No”?
“What happens if/when they say “No”?”
They don’t say no. If it turns out to be something they can’t fix (which is rare) you get a refund. If it turns out to be a legitimate bug, you get a refund. Otherwise they work it until they can fix it.
But if it still baffles us, and it is Microsoft software, we can call them for about 250$ and they will work on it until they solve it.
I’m sorry, but I just don’t buy that. Microsoft isn’t stupid, and only charging you $250 for a job that might take multiple hours is a stupid thing to do. They would be losing money. I’d guess that to get the privilege of this $250/solution deal, you’re paying some sort of yearly fee. It’s simple economics, nobody can provide professional onsite support for that little without losing money.
Think about it, their costs for just having one guy work on something for an hour is probably already $250. Perhaps your school threatened to migrate to a different platform, and they gave you a sweet deal, but I don’t believe this is their standard pricing scheme.
You obviously haven’t a clue what you’re talking about.
http://support.microsoft.com/?LN=en-us&x=18&y=23&scid=gp%3Ben-u…
Professional support is 245.00 during business hours (and “mission critical” products such as SQL Server/Exchange/Windows/Etc are supported 24×7). There is no yearly fee involved unless you step up to Premier support. The sky is the limit there but you are paying for dedicated resources then.
That 245.00 buys you a solution. If one isn’t provided, or if it turns out to be a legitimate bug (which is rarer than one would think) you get a refund.
It’s well known that MS PSS (Product Support Services) loses money. All support centers do regardless of the company (i.e. they are all cost centers within the company)…the simple economics is that they make their money elsewhere, and PSS gets a budget that they need to operate within.
If fact, most divisions within companies are cost centers. Out of MS’s 7 core businesses, only 3 make money.
When I worked in PSS as a SQL Engineer the average cost per incoming call to our group was around 800 bucks, and average time to resolution was about 6 hours.
You obviously haven’t a clue what you’re talking about.
Partially yes, I neglected to look it up. Mea culpa. But my statement that they lose money is apparently correct.
That 245.00 buys you a solution. If one isn’t provided, or if it turns out to be a legitimate bug (which is rarer than one would think) you get a refund.
Not quite. From the site you linked to: “A Problem Resolution incident is defined as a single support issue and the reasonable effort needed to resolve it.”
So you get a reasonable effort, not necessarily a solution. They do not say you will get a refund if they can’t solve it in a reasonable amount of time. Of course this might just be an omission on their site.
Anyway, so Microsoft may lose money on their support, but your claim that all companies lose money on support is silly. Many software companies survive based on their support contracts. The initial sale is just the beginning.
“So you get a reasonable effort, not necessarily a solution. They do not say you will get a refund if they can’t solve it in a reasonable amount of time. Of course this might just be an omission on their site.”
We’ve had them on calls for 20+ hours over several days when servers have had catastrophic failures.
It is a real bargain that has to be counted as part of the cost of choosing Windows. It is an amazing benefit and they throw a lot of resources at it.
The thing that makes this MS study, if not ludicrous (sp?), then irrelevant, is that what the costs and benefits are for a given situation is highly dependent on the specifics of that situation. In other words, the MS proprietry route is appropriate for some environments, the OSS/Linux route is appropriate for others. A heterogenous solution is appropriate for yet others.
If the original poster in this thread finds that Microsoft servers and support are better in the situation he works in, then so be it. If it so happens that for his particular situation, it would be far to expensive and/or too much work to both migrate to and maintain a RH infrastructure, so be it. Noone here but him can ever know the full details of the situation and therefore he is the only one qualified to make that call. If it works, it works.
On the other hand, if for someone else, using a customised linux solution is the best way forward, good on them. They’re getting what they want the way they want it.
No large organisation is seriously going to base its decision purely on a study like this anyway. They’ll do their own internal study, perhaps approach the vendors to see what their best offers are, and decide themselves. Or they’ll keep doing “what we’ve always done”, whatever that may be (admittedly, this inertia tends to favour Microsoft).
Edited 2006-09-27 03:22
You obviously haven’t a clue what you’re talking about.
Partially yes, I neglected to look it up. Mea culpa. But my statement that they lose money is apparently correct.
That 245.00 buys you a solution. If one isn’t provided, or if it turns out to be a legitimate bug (which is rarer than one would think) you get a refund.
Not quite. From the site you linked to: “A Problem Resolution incident is defined as a single support issue and the reasonable effort needed to resolve it.”
So you get a reasonable effort, not necessarily a solution. They do not say you will get a refund if they can’t solve it in a reasonable amount of time. Of course this might just be an omission on their site.
Anyway, so Microsoft may lose money on their support, but your claim that all companies lose money on support is silly. Many software companies survive based on their support contracts. The initial sale is just the beginning.
All support centers do regardless of the company (i.e. they are all cost centers within the company)…the simple economics is that they make their money elsewhere
You’re grossly generalizing. Yes, non-chargeable customer support and tech support are cost centers but smart companies engineer that overhead into the cost of the product, if they even offer those services. Smarter companies focus on quality control that reduces the overhead involved in supporting the product. Less smarter companies use support centers as a band-aid for poorly engineered products by simply saving on development and investing in post-support.
But paid/contract services are a goldmine for almost every organization. Only Microsoft could lose money doing so, and frankly I doubt they are. They’re probably playing a little three-card monte with the bookkeeping, by transfering the expenses for support into a seperate division and leaving the profitability of the software products untainted. It’s an accounting game that balances out on the overall P&L anyways.
Incident support isn’t as lucrative as the annuity stream you get from maintenance contracts, but they capture revenue from customers that wouldn’t be purchasing annual support anyways. Your team may have taken 6 hours to resolve the average $250 incident, but there are probably others that take minutes to resolve and cost-balancing comes into play. Support centers are generally built upon a database of past incidents in order to reduce the cost and effort involved in problem resolution over time. The longer a product is out there and the more widely deployed it is, the overhead for providing support diminishes since incidents will tend to repeat or have common fundamental similarities.
At least that’s how most companies operate. I can’t believe Microsoft would be that different, only the scale and scope due to their size would. But you can bet that other giants like Cisco or IBM are not losing money on product support.
However, getting back on topic, this does underscore one of the biggest challeneges enterprise linux faces. Companies are not necessarily adverse to paying for high-level vendor support, in fact many insist on it. If something goes wrong with Microsoft products, you can ultimately go back to Microsoft. Organizations need to build up the same comfort level in believing that linux can be equally supported.
Modded up, nice response.
“But paid/contract services are a goldmine for almost every organization. Only Microsoft could lose money doing so, and frankly I doubt they are.”
I neglected to mention that the support contract division of Microsoft (aka Premier Support, which targets fortune 1000 companies) does indeed make money for Microsoft; most of those companies pay millions annually for a team of dedicated Microsoft resources (so in essence they are paying their salaries and then some) such as Support Engineers, Account Managers, training, etc. It’s also worth mentioning that those companies get an enormous discount on software licenses as well so it balances out for them. So yes they do make money elsewhere on support, but not within the realm of “Professional Support.”
“They’re probably playing a little three-card monte with the bookkeeping, by transfering the expenses for support into a seperate division and leaving the profitability of the software products untainted.”
I wouldn’t call it a monte as this is common practice in all companies…in this case each team in PSS has a direct correlation to a product team (i.e. the developers who wrote the product) out in Redmond, and that’s who subsidizes their quarterly budget. What’s even more interesting is that when a bug is indeed rooted out in a product, the support team bills the product team for each call until the bug is fixed (and they don’t make a profit off of this, but it’s motivation to the product teams to get it fixed ASAP).
“Your team may have taken 6 hours to resolve the average $250 incident, but there are probably others that take minutes to resolve and cost-balancing comes into play.”
Yes, it was an average. And usually for the quick fixes (perhaps there was a KB article with a remedy on the MS support site) we would choose to refund the customer’s money. Some incidents took 30 minutes, and some took 30 days (or longer). But generally speaking it took a total of 6 hours, with about 12-15 man hours when you include consulting with other teams or teammates.
“Companies are not necessarily adverse to paying for high-level vendor support, in fact many insist on it.”
You nailed it. And that’s why it peeves me when I see folks say something like “well, with open source I can fix it myself, or I can find the developer who wrote the product and they’ll fix it.” If I took that sort of attitude to my CIO or dept head I’d be out of a job, plus it’s a shoddy attitude to have towards getting issues resolved. Companies often put more emphasis on product support than the actual cost of the product licenses themselves, which is why they are willing to pay millions annually for dedicated resources.
That’s the cost, doesn’t matter if you believe it or not. MS is not in the support business, they make their money on software licensing, the cost of MS support is very reasonable, unless you are a home user
uh dude, if you’re an academic institution, redhat will license you redhat enterprise at $60/server. that’s quite a chunk less than 1.5K.
http://www.redhat.com/rhel/details/academic/
Edited 2006-09-26 21:18
“uh dude, if you’re an academic institution, redhat will license you redhat enterprise at $60/server. that’s quite a chunk less than 1.5K.”
I quote from the RedHat site: “but they are provided without RHEL technical support. For production and business application usage of Red Hat Enterprise Linux, academic institutions are advised to use supported Red Hat Enterprise Linux versions.”
If I want RedHat support, I pay 800/1500$ per year for each server. We have over 200.
OTOH, we pay about $125-150 for Windows 2003 R2 Server under Microsofts Academic licensing program (plus CALs if needed). We then buy support incidents for $1225 for 5. We use about 10-15 per year. About 2500$ or so.
A real bargain compared to RedHat.
Edited 2006-09-27 00:00
even for the full package deal, that page does say academic discounts are available (it’s to their advantage to get a foothold in the schools.)
anyhow, seriously, if pay-for support is that much of a constant and predictable cost factor for you… perhaps you really aught to be considering migrating to a different OS anyhow.
“anyhow, seriously, if pay-for support is that much of a constant and predictable cost factor for you… perhaps you really aught to be considering migrating to a different OS anyhow.”
Ha ha ha ha ha ha ha ha. You can’t be serious.
Which “magic” OS doesn’t need support? I’ve even had to call for support for an OpenVMS server and OpenVMS is a lot more reliable that Linux or Unix.
Lets see … we could spend an ungodly amount of time and money to switch 3000 desktops and 200 servers to another OS, or we can budget 2500$ a year for support for Microsoft software.
Duh!
I work in a K12 School District…
If we wanted the same service from RedHat, we would need to run RedHat on each of our 200+ servers at 1500$ per year per server. -NotParker
This is a bunch of baloney, I also work for a K12 school district, I don’t know what redhat charges per desktop/server but I do know that they have discounts for educational institutions as well as site license deals once you get above a certain amount.
For a real-world example, we have a site license with Novell that runs about $32,000/year, this license covers every single site in the district and includes not only an unlimited number of servers but also the Groupwise mail server/clients, disk imaging software, desktop management , etc etc. And if you want to run SLED10 on all the desktop computers it’s only $0.50/machine/year.
Lets see … we could spend an ungodly amount of time and money to switch 3000 desktops and 200 servers to another OS, or we can budget 2500$ a year for support for Microsoft software.
First of all, if you have 15 desktops per server you are doing something wrong, unless perhaps you are doing thin clients but since you said you are using windows I’m guessing you aren’t using thin clients. Secondly windows costs money, office costs money, microsoft does not make billions of dollars off nothing. The OEMs’ pay microsoft around $75 per copy of windows xp pro, assuming retail costs for a moment the customer ends up paying somewhere between $75-$125 for each copy of windows.
If you want an example just pick an OEM that sells machines with and without windows and see what the cost difference is. For example take the HP dx5150, this is a mid-range current athlon64 desktop model sold by HP that you can get with freedos/suse or windows. If you get the machine with windows xp pro it is $130 more than the same exact hardware without windows. My point is you have already spent tens of thousands of dollars on copies of windows and acting like they are just giving the stuff away.
“I don’t know what redhat charges per desktop/server”
No. You don’t.
“Secondly windows costs money”
Yes. But it isn’t costing us 30,000 a year.
As I’ve said, we pay about 125$ per server. We buy 20-30 per year. This year, probably closer to 0.
Thats about 3750 a year for the server OS.
And then 5 – 10 support incidents per year = 1200-2400.
CAL’s are on top of that. But aren’t that expensive.
Exchange and SQL are more, but we only buy those every 3 – 5 years.
That 30,000 you are spending on server OS etc sounds outrageously expensive.
Edited 2006-09-27 02:43
“The OEMs’ pay microsoft around $75 per copy of windows xp pro”
I don’t have the numbers in front of me, but academic licenses from MS are dirt cheap, and much less than $75. MS practically gives them away. They don’t make much money from academic licenses.
If you get the machine with windows xp pro it is $130 more than the same exact hardware without windows. My point is you have already spent tens of thousands of dollars on copies of windows and acting like they are just giving the stuff away.”
I doubt they went to HP and bought 3000 desktops, and *if* they did they no doubt got a HUGE discount A) for buying in bulk and B) because they are an academic institution. More than likely they went through an MS channel and probably spent less than 20,000 dollars total for their entire system.
“Thats an amazing deal.”
And how much does your school districts yearly support contract with MS cost?
“And how much does your school districts yearly support contract with MS cost?”
We pay by the incident.
Huh. I’ve done quite a bit of work with both companies, and I can honestly say neither has ever worked on an issue for 20+ hours for me, that I’m aware of. If it gets that far, generally hardware is blamed, the install is blamed, or some other crap. With redhat, a few years ago after upgrading to a newer kernel, our I/O went to hell, and they did work on that one for a while, but we were only one of several (hundreds?) who called in about it. A new kernel was available soon after that which fixed the issue. That’s about the closest I can think of.
Hire the best sysadmin you can find: Sometimes it’s hard to find a smart one who isn’t a lying weasel looking to play counter-strike on your dollar, but they’re out there.
Once you’ve hired him. Give him 3 months to figure out which way to go (possibly after 4-5 months working with the current setup so he knows your problem domain(s)). Tell him to research cost so he knows how much each will cost. Tell him to figure out how many people he needs to do it.
Then ask him which way he wants to go. Don’t listen to yourself, listen to him. If it costs twice as much … pay it.
Hopefully your company already has a bright sysadmin, but if you don’t you’re going to have to start by finding one.
It’s not the cost to buy, or to run everyday, it’s the cost when it breaks. If your company doesn’t lose much from a lost day of work, great, you’re very free from this cost, but I bet most businesses lose a lot of money during downtime: Much more than their IT infrastructure.
These cost analysese are for cheap businessmen looking to impress their bosses by doing basic math to save him small sums of money. Saving money is great, but if you’re not getting your job done the money you saved is worthless. And if you’re saving at the cost of your employees you’ll pay later.
I’m going to weigh in with my own two cents here, and say… it depends.
If you have classic small bussiness setup with say five workstations, and an SBS server for file, print, mail, and a database with a custom app for whatever the bussiness does, then yes, Windows is probably cheaper.
On the other hand, if you have developers that require access to multiple machines, operating systems, databases, languages, etc, then you will probably find a pure MS solution, even with an MSDN subscription, will probably be a lot more expensive.
Linux – Tomcat – PostgresDB: Free – Free – Free.
Nobody was buying the last load of fairy dust on how linux costs more than windows so now we’ll try another propoganda carpet bombing saying how it costs the same, sorry but nobody will buy that either, you should have put that money into development so you can hang onto what little, “it’s easier”, you have left.
Every day in the jungle a penguin wakes up and starts running….
“Nobody was buying the last load of fairy dust on how linux costs more than windows so now we’ll try another propoganda carpet bombing saying how it costs the same, sorry but nobody will buy that either”
Actually the study recommends avoiding “seat of the pants” attitudes like yours and says companies should do a real TCO analysis before listening to the guy who has Linux at home claiming its cheaper.
I tend to think of Linux the same way I think of Paris Hilton. Sure, they both look cheap … but maintenance costs will kill you.
You sure have made a huge emotional investment in defending Microsoft. Too much to be credible or for me to believe that you are not directly associated with the company.
Why has Linux gone from a small little OS put together by a few people on the internet to a solution deployed by millions of people worldwide?
Because it works and it is indeed much cheaper than anything currently offered by Microsoft. I have clients in the SMB sector who will never go to a Microsoft solution on the server even when some of them are still running Windows clients.
Give us a break with the rhetoric and made-up numbers that do not match reality.
And that little Paris Hiltion retort sounded a bit too contrived for you to not be yet another astroturfer.
“You sure have made a huge emotional investment in defending Microsoft. ”
But a small financial investment compared to the offerred conrete examples.
“Too much to be credible or for me to believe that you are not directly associated with the company. ”
Well, Linux supporters always make emotional argumetns about freedom and run for cover when people point out RedHat’s cost and the cost of other Linux offerrings.
“Why has Linux gone from a small little OS put together by a few people on the internet to a solution deployed by millions of people worldwide?”
It was cheaper than Unix. But sales are slowing as the number of easy Unix conversions has come to an end.
And remember, Apaches lead over IIS is exclusively because of unused parked domains. And of course Apache 2.x has 11 times the number of security holes than IIS6.
“And that little Paris Hiltion retort sounded a bit too contrived for you to not be yet another astroturfer.”
Contrived? No. Accurate? Yes.
Edited 2006-09-27 18:48
Well, Linux supporters always make emotional argumetns about freedom and run for cover when people point out RedHat’s cost and the cost of other Linux offerrings.
So don’t use Redhat. There’s choice in the Linux world.
It was cheaper than Unix. But sales are slowing as the number of easy Unix conversions has come to an end.
B.S. Provide references or shove off. Use Redhat’s latest sales figures if you want – profit is down but revenue (what counts) is way UP.
And remember, Apaches lead over IIS is exclusively because of unused parked domains. And of course Apache 2.x has 11 times the number of security holes than IIS6.
Absolutely false in every sense of the word and every way stated or unstated. Provide concrete references or be struck by lightning. In fact, Microsoft spent a great deal of time and money to convince companies to park domains on Windows just so they could get a better foothold in online stats. Apache was the choice of those companies… Microsoft had to actually pay them to use IIS.
Contrived? No. Accurate? Yes.
You spelt “no” correctly the first time.
Astroturfing shills can shampoo my crotch.
These TCO studies are idiotic. The best OS for you, depends on your situation.