Someone needs to write a pointed humor piece about Gates called the Emperors' New Clues, ending up with a little child saying clearly in the silence, "But the Emperor -has- no Clues!"
I've been presuming Microsoft's positions, being self-interested, are examples of legally induced stupidity, but recent evidence suggests that Microsoft is in fact legitimately and sincerely st00pid, which puts a different light on things.
E.g., Microsoft has been arguing that Internet Explorer needs to be integrated into the Windows Kernel in order to operate. Now, any 10-year- old with two clues to rub knows that good software design depends on modularity and clean interfaces, and hence that the IE UI component desperately needs a clean interface separating it from the kernel -- this has been obvious to everyone for at least thirty years, since Unix demonstrated clean separate of the commandline interpreter from the OS kernel, allowing the shell to be a cleanly separate application. In the current case, Microsoft is being given the clean separation on a platter, since browsers are in fact cleanly separated from the OS already. X Windows demonstrates that the sort of serial-line protocol WWWeb browsers use is in fact perfectly sufficient for running the entire system UI interface: All Gates has to do is have enough of a clue to maintain the existing interface and take advantage of it, and he can both have IE as his system-wide interface, and also have a clean API allowing other UIs like Netscape to be plugged in to replace IE. Everyone happy, right? Except Microsoft would rather have a monopoly, so it pays Microsoft to exhibit legally induced stupidity and insist on producing a monster piece of spaghetti code with no internal API firewalls anywhere.
So far so good: Microsoft crazy like a fox, no?
Except Microsoft is producing equally shitty basic software "designs" even when it hurts it instead of helping it.
E.g., the Windows NT kernel services calls have a horror of a structure which virtually guarantees not just lots of security loopholes now, but a virtually unending supply of them into the future. One -might- suppose this is just so Microsoft can sell an unending series of upgrades, but that case seems weak, especially given Gates' vehement insistence that there's no money in fixing Windows bugs, hence that they don't exist and won't get fixed. (Sounds like at least incipient mania to confuse existence with profitability. Gates' daughter isn't profitable, does it follow that she doesn't exist?)
In the most recent example of Microcluelessness, it turns out that the Windows filesystem uses both long and short names for files... and that when permission bits on the long filenames don't allow one in, one can often access the same information using the short filenames, which often have different permissions set on them. This is today's Computer Emergency Response Team bulletin on the Microsoft Internet Information Server -- remote users can often access supposedly forbidden information.
Now, the Unix filesystem is a third of a century old, and it deliberately avoids keeping an y sorts of permission bits in the directory entries precisely to avoid this sort of duplication of information and consequential maintainance nightmare: Directories hold only name->inode# pairs, and all other information about the file is stored exactly once in the inode for the file, not replicated in every directory entry.
This wasn't a brilliant insight even thirty years ago: it's just a really fundamental programming principle that one never replicates information without cause, precisely to avoid this sort of maintainance problem.
Bill Gates, popularly acclaimed boy genius, has had thirty years to study the Unix filesystem and learn from it, with full access to the source code almost all that time. He's got the money to hire the entire Unix design crew if need be to read the code to him and explain it in words of one syllable -- and still he fucks up the new Windows filesystem design in just this absolutely elementary fashion, which will probably haunt all users of the Windows filesystem for decades to come. What possible advantage does Microsoft derive from such fundamental botching of quite basic design issues?
The most parsimonious explanation is that in fact Microsoft is not pretending to be st00pid in court: It simply IS that st00pid, and its software designs exhibit this across the board, whether or not the results favor Microsoft. Microsoft is run by a college dropout who never had any real computer science education, who has never had any opportunity (or inclination) to rectify that, and who insists on throwing software out the door every six months come hell or high water, creating an environment in which doing it right the first time and fixing it later are equally unthinkable -- the solution to every broken piece of software is a new broken piece of software thrown atop it, producing an overall Windows design strikingly reminiscient of a rubbish heap.
In short: The Redmond Emperor has no Clues.
And if that is the case, things look even bleaker for Microsoft than the objective situation might initially suggest: Not only is Microsoft the Last Dinosaur in the sense of being the last surviving vendor of proprietory solutions in the brave new open standards era -- but at that it is a giant-size peabrained Dinosaur competing with smaller, faster mammals with bigger brains. Because there are clueful folks out there running competing companies.
And the supply of dinosaur food is fast running out: The steady supply of first-time computer buyers on which Microsoft has fattened over the years is rapidly dropping to a trickle as market saturation is reached and children of the Computer Age reach positions of purchasing power -- no longer can Microsoft count on the average purchaser being even more clueless than its CEO. If there is one clear message from the new clueful customers, it is that they are tired of dinosaur droppings: Java is one way they are saying this. The percentage of first-time buyers is now crossing 50% headed downward, and in a few short years will drop to negligible proportions. What then will preserve the Last Dinosaur, other than tarpits and sedimentary rocks?