[Leroy S. Mort]

"For those users, migration costs will be very high because all Windows applications must be replaced or rewritten,"

I'm guessing he means in-house applications here, not Office-type suites, etc.

[Mind-numbed Robot]

Aren't most operations of any size running off a server? How many companies are using individual, unconnected desktops?

[SES1066]

I believe this because there is a cost in throwing out paid-for software. The cost would be a lot less for a virgin installation. What cannot be so easily quantified or qualified is the continuing rise in cost of a Microsoft monopoly that is publically seeking to move to a subscription basis. My personal belief is that MS would already be doing this for its OS and Office products if not for fear of Linux. Having defeated DR-DOS, BEOS and OS/2 and reducing Apple to 5-6%, the coast would have been clear except for the guerrilla open code Linuxers.

[staytrue]

Abandonning Linux is an invitation for MS to double or triple its software prices. Right now, MS fears linux and keeps its price low and competitive, if linux dies, you can expect to pay double or triple for software.

[smith288]

I can think of the complex porting that would have to be done for the insurance company I work for. Hundreds or even thousands of applications would need converted... Hardly a justifiable cost.

[FastCoyote]

It's not the next couple years of costs I'm worried about, it's when you go to Longhorn and DRM and Office requires 2003 server or better to run. Plus new licensing plans, backward incompatibility, no access to underlying code. With the browser being written into the OS, and the move to automated updates, the whole thing becomes a security nightmare subject to catastrophic failure. All in all, I see Microsoft becoming a monster within the next five years.

So, I move to Open Office (spreadsheet and wordprocessor for 90% of the office), make my server FreeBSD (solid as a rock), write my aps in Java or PHP with Mysql in the backend, and when the big crash happens, my little world just keeps on going. What a relief. And all the software is free.

So, I take a hit on the learning curve of Unix (which is not negligible), but the application changeover won't be too bad if done over a few years and I've already started that. So, all in all it comes out about the same costwise over say 5 years, but I'm way ahead on the catatrophic failure front and not having to worry about the next killer virus. And over 20 years there's no planned obsolesence to pay for (needed to keep Microsoft in Business)

[Rain-maker]

I don't see the report mention the lack of programs for Linux.

I use graphics oriented software and there's no Photoshop, FlashMX, Director, Lightwave, etc. in Linux format. Until the support from software manufacturers comes in, the author is beating a dead penguin.

[Dataman]

Windows 2000 and Windows XP include more modern technology than Window 95 and are generally more stable and incur lower costs,” said Silver.

What? Linux will crash 4 or more times a day compared to XP's 3 times a day?

The very fact that the comparison has to be made shows that Linux is eating away at Gates.

[Steely Glint]

Have you run both Linux servers and Windows servers? I have.

The great difference between the two is stability. No Linux server I have been associated with has EVER carshed.

[AgThorn]

Microsoft-sponsored survey

Yeah, that's believable ..

[pyx]

The report comes in the same week that a Microsoft-sponsored survey of just 12 companies claimed firms could save up to 28 per cent by developing certain programs with Windows rather than Linux.

"Just twelve (12) companies", huh ? ... Filed in the round drawer ...

[Gee Wally]

"For those users, migration costs will be very high because all Windows applications must be replaced or rewritten,"

I'm guessing he means in-house applications here, not Office-type suites, etc.

Isn't Microsoft pushing for those apps to be written/rewritten in .NET anyway? So how is it an added cost if Microsoft is pushing for migration to its .NET paradigm? I think a more accurate assessment would be achieved by comparing the cost of converting to where Microsoft wants those apps to go, .NET, versus Linux and Java or whatever.