With a long-awaited security update to Windows XP now complete, Microsoft is preparing a holiday season push for the 3-year-old operating system–and is set to revisit ambitious plans for the next major revision, News.com has learned. Also, Microsoft is expanding a plan to draw mainframe customers to Windows for high-end applications.
Microsoft preps XP push, mulls Longhorn ‘priorities’
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Why do I think Microsoft will begin to abandon it’s XP update plans and concentrate on Longhorn?
I have serval customers owning a mainframe and I see none of them thinking about replacing them with Windows. It is no question about technology but more a question about money. They have spend so manny million swiss francs (I am from Switzerland), that they are not throwing that away. The only option often left to them is to run Z-Linux on the mainframe.
One of my key customers is eaven on the way to enroll a new server infrastructure, where everything will be migrated to Solaris (timeframe is 2006/2007). I don’t get it! Linux is already in the company (okay… Solaris and Windows as well) and they want to replace the Windows servers with Solaris. Anyway.. the reason for that is, that they had a internal security audit done by one external company. The company told them, that hacking Solaris is much more harder, since not every script kiddy can own at home a Solaris box. But with Linux everyone can install it at home and search for exploits. Let alone Windows.
Well… I stay quite! No way to convince the responsable person of his wrong thinking. Not that Solaris is bad, but a 100% decline of Windows and Linux in flawor for Solaris is in my viewpoint not the best solution. (btw: they had most of their Solaris systems runing Solaris v4 for SPARC. Can you imagine the exploits available for that systems? The security company found manny of them.)
Anyway… at least the mainframe will run Z-Linux. Now tell me how Microsoft want’s to beat the IBM guys in the mainframe area? IBM offers guarante uptime and manny manny SLA’s serval things on the mainframe. Microsoft can’t beat that. Microsoft just sells software. That’s all
Never underestimate the power of marketing (or the stupidity of consumers).
Actually Microsoft is more like a financial company, like a bank. Most of the annual revenue is dereived from ROI on their short term investments etc. And not in sales of their products, which only have two-three money making brands: Windows and Office. Both under heavy siege, i wonder.. Will we see Microsoft as a competitor to Citibank someday?
.
Probably not, now that they are paying out truckloads of money in dividends. I fail to see what good it does Microsoft, when they are loosing ground in many markets, and the future dont look so bright (for their growt, not that they are going bankrupt anytime soon).
Why would anyone pay per-cpu-priceing on Windows for a mainframe. The keywords on a mainframe is RAS and atleast five nines
– something, i have never come to expect from microsoft, and something they will have a hard time knocking of their brands.
I already have a few customers myself that run NT Clusters and have been considering moves to Windows Server 2003 Data Center so this should be very interesting to see where it goes.
“The company told them, that hacking Solaris is much more harder, since not every script kiddy can own at home a Solaris box.”
I hope you are aware that Solaris for x86 is available for free, and every script kiddie can download & install it and search for exploits?
Do YOU have any non-Windows customers?
we have just updated all of our nt 4 boxes to xp on the desktop side, and we are adding 5 more windows 2003 servers. I really don’t care when Longhorn will be out, becasue our software will run on any windows pc. I dont see windows going away any tome soon, We do have a few linux boxes in house as compression servers and network monitoring, but our work horse pcs are windows.
A year ago I was only windows, slowing however Linux has been becoming my OS of choice. I still have a windows 2003 server and stuff but I prefer Linux over windows now since every time I use one to take over a windows server function the results have been very possitive.
Windows certainly isn’t going anywhere however.
“Windows certainly isn’t going anywhere however.”
It’s not on the decline either. Simply because someone converts their entire operation to Linux, is by no means an indication of the decline in Windows.
While someone is huddled in their IT dept. installing Linux and considering the irreversable and permanent damage they are going to MS. Consider how many new Windows systems got sold while that few moments of installation was running.
One step forward, two steps back…. Thats not going to change anytime soon either.
“Now tell me how Microsoft want’s to beat the IBM guys in the mainframe area?”
Not just Microsoft, to be 100% correct, but Linux on servers, too.
Here is the real story of a company that wanted to buy IBM mainframe to run Linux on it.
A company has a set of applications that are coded in C/C++ and very much platform agnostic: can run on Linux and Windows with no big efforts from developers.
A presentation by IBM salesperson was impressive: run many Linux instances, start them at will, replicate them, all that on top of virtual layers- so they can’t harm each other, and so on, and so on.
Then, salesperson left, and company started to look at numbers.
The cheapest mainframe goes for close to $1 million, before discounts and mail-in rebates.
Same time, a company got a quote for dual-Xeon 2.4 GHz rack mounted server with enough RAM to run Linux or Windows2003, for $5,000.
So, a cool million could buy it 200 servers or one mainframe.
These 200 servers bring, combined: 1,000 GHz of computing power, 800 Gbytes of RAM, 8,000-12,000 GBytes of hard drive space.
Can IBM mainframe, in $1 million range, match it? No, not really.
On top of that, virtual layers on top of virtual layers is fine, but each is, by definition, a hit on performance. Why would you want virtualization when you can get Linux on a server talking to the raw metal?
Then, redundancy. Yes, IBM may guarantee uptime and all that stuff, but if they can’t deliver- they just pay money. Not too much, by the way. Can’t cover company losses from its highly demanding contracts.
That company estimated it can put 30% of these 200 servers aside thanks to overcapacity, and keep them for urgent replacements.
So, at any given time, if up to 30% of their servers go down- they can still manage to run the business using replacement pool.
Tell me, can you guarantee that if up to 30% of random mainframe parts fail, a mainframe can still be running? No, not really.
After that, what is left for company is not rack mounted servers vs. mainframe, but Linux on server vs. Windows on server. The mainframe is gone from evaluation. It may be nice and it looks cool in a server room, but it can’t deliver. In the case of that specific company, at least.
Well, now when you have Linux vs. Windows in the server room, it is the game that Microsoft knows how to play. Will Microsoft win? May be. Will it loose? Quite possible.
The outcome of Linux vs. Microsoft on server is irrelevant, what is relevant is a company not buying IBM mainframe.
What is interesting in all of that is a business that compares Linux on mainframe to Linux on servers, picks pool of servers for better performance, and then evaluates Windows vs. Linux on the server.
If IBM were not offering Linux on mainframe, that company would be left comparing something like AIX to Windows, and you understand that in such comparison Windows does not stand a chance.
“Will Microsoft win? May be. Will it loose? Quite possible.”
Yes Microsoft has been known to be quite loose, after all it has a lot of customers
[cx]
You could guess, from my name, that English is not my native language.
I speak five different languages, English is #4 in that list. Can read on two more languages.
If you would like to post something in Russian, be my guest. I promise to be generous enough to overlook your mistakes.
**************
Speaking on the topic: 200 servers are much larger challenge for IBM than it seems.
First of all, these were not from IBM. Their competitor offered better prices. So, not only IBM lost on mainframe, it lost on hardware. IBM is gone.
Secondly, on these 200 servers it is unlikely the company I mentioned will be running just Linux. They have UNIX/Linux apps, they have Windows apps. For them, because they are not forced to port everything on Linux any more, the natural path is to have mixed Win/Lin environment and see how it goes.
So, at least for now, not only IBM didn’t make hardware sale, but its new software pet- Linux- did not gain 100% server base.
Which means: IBM got $0, Microsoft got few bucks in license fees (which is better than $0 IBM got), and Microsoft Windows is running in the server room of that company.
Sure, Linux is there too, but Microsoft makes money on Windows while IBM loses money on Linux- because IBM is paid $0 by that company I am talking about for its billions of dollars investment into Linux.
Will Microsoft win and take 100% of servers in that company? Will Linux? Who knows.
The fact is, at the end, dealing with that company, IBM has $0 in sales, many many dollars in Linux R&D expenses- and that, my friends, is bad for a corporate bottom line.
So, what were you saying about Microsoft beating IBM at mainframe area? Linux beats IBM at mainframe area even more.
200 Linux dual-CPU Xeon servers (not from IBM) running WhiteBox Linux will beat IBM mainframe running Linux 9 times out of 10. That is the reality.