It was intresting to read Scoble's take on Cringely's article and his feelings about the future of Microsoft. It might be cool if there where separate companies presenting the different products as a defense against feelings of hate of all things Microsoft. However I don't think it's all that necessary or even wise. The Xbox is marketed in such a way that a lot of folks don't think of it as Microsoft, even if they do know it. The Tablet and Media Center Editions of Windows, is Windows. They would suffer (the tablet especially) without having the company thinking about them and implementing support.

In regards to the dividend comments, remember that Microsoft is currently increasing its R&D investments, and I believe that both things can co-exist. An interesting point that I think I read on Motley Fool was that by paying a dividend, the companies stock holders (which is almost anyone who owns a stock fund) gets a more “me” sense versus a “that giant cooperate machine” sense when Microsoft is sued. At least Scoble knows that there are actually people and not mythical demonic beasts that work here.

Rest assured that folks are working on the stuff you ask about. Windows XP was to a large part of doing better at handling photos off of a digital camera, and that team isn’t done. (Side note, I watched a usability study of a Linux user try to do some of the digital camera style tasks with Linux, and got an appreciation of what I hadn’t had to worry about when doing it with Windows XP). The Tablet and Cell Phone groups haven’t gone anywhere. The Windows Media Center Edition (tivo like) was recently released and they are already working on the next version. There is no end to the work that my group has to do right now either. When Longhorn planning started we got a glimpse of what windows could be like someday, and then the next day I saw what the social computing group was up to (work that may not be ready for the release after longhorn). There are a lot of things to invest in, and I know that folks here are investing in them, and will continue to for a long time to come.
8:47:16 PM    comments ==

Slashdot is linking to a Robert Cringely article that attempts to present the idea that Windows should be ported to Linux. This article reminds me why I don't bother to read Cringely (and may well have destroyed his credibility with others), but I've heard the sentiment before, so I should perhaps look at the issue more completely. Cringely makes the following arguments:

1. Windows is mostly a brand implemented with a little bit of windowing system sitting on top of DOS (but the DOS is very well obscured).
2. Apple did something similar with MacOS X
3. Wouldn't have to maintain the Kernel Code which would be a cost savings.
4. The GPL wouldn't be a threat since Microsoft would do very little to no kernel development.
5. Since this would be just another competing windowing system on Linux, the antitrust stuff goes away.
6. Stability gain running on Linux instead of Windows XP
   
   I'll add one more argument:
   
7. Reimplementing means a chance to end up with something powerful yet cleaner (see the visual effects on MacOS X and some of the BeOS speed, API and architecture features)

Each of these ideas are pretty flawed:

1. Windows is a brand, but it's also the definition for a large set api's, interface and code that can't be trivially ported to Linux.
2. Apple sacrificed a lot of their backwards compatibility to go to MacOS X. This is expensive on two fronts, stable drivers/hardware support and application support. Apple's control of the hardware minimized the hardware problems, but Microsoft would get slammed hard with them.
3. Of course Microsoft would have to maintain the kernel for its distribution. One of the more useful features of Windows XP was 802.1x and good wireless support. A good chunk of it I believe involved stuff that happens deep in the network stack in the kernel. which brings us to the problem with point 4
4. Microsoft does a lot of kernel work and would continue to doing new kernel work, this would be extremely problematic from an IP and kernel control stances.
5. Redefine the relevant market from OS's running intel compatible platforms to Operating Environments running on top of Linux. Blam! Same problem.
6. I don't think you get a stability benefit moving to linux from windows XP.
7. You can do a lot of the same stuff without a rewrite, Microsoft has just been slow about doing it. Ex: API -> .NET

It took a long time for Windows NT to be built up and polished enough make the transition, and even then it was almost too painful. That bought some significant reliability and security gains. What is left for Linux to offer, especially at it's cost?
10:29:56 AM    comments ==