Interesting idea, that Microsoft should gradually break compatibility in stages, allowing software developers to just issue patches that let their existing software work with the new OS, while still gradually improving security and performance. If Microsoft could do this at a rapid pace (as in, breaking a similar chunk of compatibility with every release), then they quite possibly could fix most of their problems, saving Windows in the long run. The only problem is, I don’t know if Microsoft was intending to break this level of compatibility, and if they didn’t, would they be willing to do it again? Still, I’ll hold off on making any more comments about Windows until the Vista optimized software comes out (it might take a while for truly optimized software). There are still some things that I don’t like about Windows that they couldn’t get rid of without completely breaking compatibility (cough the registry cough), but significant improvements over the next few years would be satisfactory.

I’ve read several reviews and articles on Vista but I can’t recall a single significant thing it’ll do for me that I can’t do with right now with XP and some third party software without sacrificing oodles of resources just to have their glass UI. The Vista hype is completely unjustified . Someone commented above that it doesn’t support OpenGL… this whole thing is not even funny anymore.

I am sure Vista will some day justifiably oust XP from the top spot but that’s a LONG way off.

Alright, now on to clarifications. I did not mean to say that Microsoft should smash every hard drive that has ever had Windows source code on it and re-write from scratch. What I meant was that they should re-work the code they already have into a better product, even if that means breaking compatibility. While Windows at heart is a good operating system (yet with a few major flaws that could easily be fixed in a re-work, such as how the registry currently works), it is plagued with old bits of code that are not accomplishing anything. In fact, some of this code only serves to bloat the OS, slow it down sometimes, and occasionally cause security problems. Cleaning up Windows would not only improve the OS, but also make it easier for Microsoft to maintain and improve upon it, seeing as the source would be more streamlined.

Hrm, now you’re delving into the realm of the subjective. I mean, Vista was certainly reworked a lot, but maybe you think they should rework other parts they didn’t? I can’t say I understand the entire OS (no one really does…), but to really know which parts needs reworking requires a lot of study in those respective parts. Microsoft is big enough that I am sure they have people looking into almost all aspects of the OS – and they are smart people – so this suggests to me that prioritization was an issue and some things were just more important to the PMs than others. And again, we’re back what subjectivity – what SHOULD the priorities be.
With regard to compatibility, Vista did break quite a bit of this, so I wouldn’t say they’re not taking the route you suggested – although maybe not to the degree you would like. But from a business perspective, I think they have broken just the right amount, so they don’t abandon their software base, which is their greatest strength.
If you look at other technological markets, you might think Vista’s rather progressive with regard to breaking compatibility. I mean, we’ve been stuck with the crappy x86 ISA for way too long. Everyone knows it’s a hacked up piece of crap, but we stick with it because of the software all built for it. Heck, Intel made a whole hunk of hardware to translate x86 instructions into better micro-ops because it sucks so much.

Anyway, I think my point here is that although Vista certainly isn’t perfect, I believe with regards to the kernel that it is a step in the right direction, and probably as far a step as can be expected at this time.

Alright, now on to clarifications. I did not mean to say that Microsoft should smash every hard drive that has ever had Windows source code on it and re-write from scratch. What I meant was that they should re-work the code they already have into a better product, even if that means breaking compatibility. While Windows at heart is a good operating system (yet with a few major flaws that could easily be fixed in a re-work, such as how the registry currently works), it is plagued with old bits of code that are not accomplishing anything. In fact, some of this code only serves to bloat the OS, slow it down sometimes, and occasionally cause security problems. Cleaning up Windows would not only improve the OS, but also make it easier for Microsoft to maintain and improve upon it, seeing as the source would be more streamlined.

One more thing. Thank you for the article. I enjoyed it.

@Ryan – I always try to be civilized. I see no purpose in having pointless flame-wars over every minor disagreement. Community is very important for a blog.

Judging by the quick responses in the comments I almost thought you guys were going at it, but it looks like you kept it civilized. I really appreciate it because it’s nice having a community here that can learn from each other considering I just learned nearly all of that Mac info.

You know I am not alone in my opinions that Windows is a terrible operating system, with sloppy code left to stagnation for many years and missing many of the features of other, alternative operating systems. I do not like many of the ways Windows operates, and I know many people who think that Microsoft should go “screw compatibility, let’s re-write the OS from scratch”. That’s what I was trying to get at in my previous post, though I now realize that it was very poorly phrased. I was not trying to suggest that the kernel has never been upgraded, but rather that it is old and has never been redone, like many Unix-like OSs. You will also remember that Apple has already left their old “classic” OS in favor of the more modern OSX.
On another note, your right, the “services”, DRM, and lack of support for competing standards sickens me.

Agreeing with a lot of other people, certainly doesn’t make it true, especially when those people are ill-informed. I do agree that Microsoft has a tendency to try to “retain its assets” (paraphrased from a Microsoft interviewer) – i.e., never do a complete rewrite. However, most software companies do the same thing, and I can’t think of one proprietary software company that doesn’t (maybe a small one might). When you’re stuff is proprietary, it really makes you view it as an asset. Whereas open source projects get to do all the great innovating and renovating because of the intense Darwinian battle-field they live in. But even then, larger projects tend to retain most of themselves – only being replaced if they were poorly designed not to scale anyway. (A good software project – as does any technical project – knows what is coming and makes is able to scale through time well.) A good example of a poorly designed system that was eventually replaced in the Darwinian OSS world is CVS. I won’t go into the details, but let’s just say it’s a giant load of HACK.

Now, Apple did do something as dramatic as you suggested – they ditched their old, inherently crappy and outdated OS for a new Unix-based one. But those were much different circumstances. First, since Jobs was fired shortly after releasing the Macintosh, Apple slowly evolved their OS – allowing Microsoft to shoot ahead drastically technologically. (Instance: they did not even have time-share multitasking until OSX!) Second, OSX wasn’t really a complete rewrite, but really similar to the OSS model of replacement – when Jobs was fired he founded NeXT, a company that made an OS based on BSD. 1997 rolls around, Apple is dying, and what do they do? Buy NeXT and rehire Stevie to save them! A little changes here and there, and BAM OSX joins the rest of the world in the 90s. Even if you consider NeXT – it was not built from scratch but off another unix system! So no rewrites there, just replacement. And when you consider where Apple was at the time, their current OS wasn’t really an asset anyway.

Now, is Vista a complete rewrite? No, of course not. But I’d say they got dirtier than they ever had before – that combined with the size of the project is really what lead to the delays. For the next OS, Microsoft might need to re-think its current model of development so the law of diminishing returns doesn’t catch up to them and force them into a humility lock again. (Last one was IE)

We’ll see where they go now…although from what I hear, the NEXT Windows will just be all the features that didn’t make it into Vista, so dramatic stuff won’t happen until after.

Random article on this stuff: [informationweek.com]