Vista rant ---- MS must dump it...

[Albuquerque]

I'm not sure if Vista is a stop-gap. I don't really think so tho.

NT6.0 (Vista) is a technology foundation for the next several OS releases from Microsoft. NT 6.1 (Longhorn Server) has the same underpinnings but with obvious changes directed at server extensibility, such as clustering, more (quantity, specialized, disparate) processors, more memory, more storage, and more performance tuning.

Vista actually brought forward a new "modular" concept to the OS foundation components such as kernel, HAL, driver model, network and the like. Going forward, this foundation is supposed to allow MS to create new OS updates on a much more rapid basis because the above-mentioned individual components are not so intertwined... (Well, that's what their development teams are saying)

We shall see.

 

[N00b]

http://techreport.com/discussions.x/13303

Good article. This quote pretty much sums up my experiences:

As someone who's been using Vista on his primary PC for ten to twelve hours a day since February 17, I can attest to this first-hand: it just isn't anywhere near as bad as everyone says. It's not even a little bad—I can honestly say I haven't once in the past seven months felt like moving back to Windows XP. In fact, I feel like I'm missing out when I have to use XP on another machine

 

[Moloch]

Which is cool. I mean, he(she?) can obly be killed by acid or fire+regeneration's included...what more can you ask for?

*shakes head*

 

[Rainbow Man]

Its more like 7% but your argument is just as valid

MacOS 7% of US market I can believe.

7% worldwide - NEVER. Impossible!

In which Euro country is there even any Mac retail presence to speak of? Out of sight out of mind is very much the case in the PC market I'd say.

Nevermind asia africa, and south america which aren't as prosperous as ye typical western nation generally speaking.

Vista is penetrating the market slower than XP did but so did XP compared to 98.

I think this is because each successive MS OS is increasiingly more mature and well-functioning towards thhe end of its life so there's simply less incentive and outright need to switch.

It works, better and better. Why change?

Of course MS sells most of its licenses with new PCs. That's how they make people upgrade.

Peace.

 

[L233]

MacOS 7% of US market I can believe.

7% worldwide - NEVER. Impossible!

In which Euro country is there even any Mac retail presence to speak of? Out of sight out of mind is very much the case in the PC market I'd say.

MacOS market share might have increased in Europe because Apple's notebooks have gained some very limited popularity in recent years. A few years ago, I think the UK and Switzerland were about the only countries where Apple even had a measurable market share (like 1% or so) and that was only because educational facilities bought some Macs.

iPods aside, Apple barely even exist in Europe in terms of consumer mindshare.

Worldwide market share... probably less than 5% and I don't think that will change any time soon.

 

[Rodéric]

MS just write an OS from scratch with modern design !
That's the only thing people need say again and again and again to MS.

Get rid of that legacy junk once and for all, make something GOOD for a change !
(After all, Visual Studio is good, so MS can do good things...)

 

[L233]

Why reinvent the wheel?

Also, what do you mean when you say "modern design"? 1970s unix-style or what?

 

[Albuquerque]

MS just write an OS from scratch with modern design !
That's the only thing people need say again and again and again to MS.

Get rid of that legacy junk once and for all, make something GOOD for a change !
(After all, Visual Studio is good, so MS can do good things...)

Well, that's the rub, isn't it? If they truly did something 100% brand new with zero legacy "junk", who in the world would buy it? Not business. Not government. Not home "upgraders". The only people left would be some OEMs, and if it had no legacy support, the OS would be shit-outta-luck until the software developers came in and wrote all the new software -- so that kinda kills the OEM sales too.

In other words, there's no way. What you ask is simply not possible for the mainstream.

What might be possible is to have a new parallel OS specifically designed to be new and different, targetted to science or something. The problem becomes again -- in 10 years, do we start yet again from scratch and drop ALL legacy support?

 

[Rainbow Man]

MS can't drop legacy support, the only reason the PC market exists in the dominating form it does today is becausr the damn thing's backwards compatible to the late 70s.

If MS would drop legacy they'd implode - and so would much of the PC market too. There's simply no widely adoptable alternative to MS today. Linux is a joke - STILL after so many years! - from a consumer perspective and IBM laid OS/2 to rest ages ago. MacOS is too small and fringe to fill the void.

PCs and MS exist in a kind of vampiric symbiosis one might say.


Possibly MS could relegate legacy support to a virtual machine running under their new theoretical legacy-free framework. And then gradually phase out support for the legacy stuff over a period of many years.

Realistically this proceass would have to take upwards of a decade though.

And even then many would complain it's too fast haha.

Peace.

 

[swaaye]

Heh. Look at how long it's taking for x64 to really penetrate the market. And it offers extremely flexible legacy support.

 

[Bouncing Zabaglione Bros.]

If you can't get companies to sort out their drivers and software for a new mass market and backwards compatible OS, you've got a real uphill battle getting them to do it for an OS that breaks with the current market and all backwards compatibility.

 

[Rodéric]

Two things (pretty late so don't expect much)

-modern operating system : look at the BeOS, and it was with th 80s tech/paradigm, I'm sure we can have better than that today.

-dropping legacy support doesn't mean you can't write an emulator on your system for legacy stuff, just like Virtual PC, or WINE.


Food for thoughts...

 

[Albuquerque]

Two things (pretty late so don't expect much)

-modern operating system : look at the BeOS, and it was with th 80s tech/paradigm, I'm sure we can have better than that today.

-dropping legacy support doesn't mean you can't write an emulator on your system for legacy stuff, just like Virtual PC, or WINE.


Food for thoughts...

Counter argument: where do you draw the line / how do you define legacy? Is everything before this point considered legacy? What pieces do you eliminate, what pieces do you depricate, what peices do you allow? Who makes this decision? How much time do you spend on virtualizing those legacy components? What is an acceptable performance tradeoff (Virtual will never be at "full speed", obviously)

As for progress on "groundbreaking OS paradigms", Microsoft Research is the responsible Redmond party for doing such things. A recent example would be the Singularity operating system, which is only open to certain tech colleges, but is at least one example of them thinking a bit further outside the box.

Again, this really boils down to a huge problem: you have "computer morons" who will complain when the computer doesn't act like an appliance. You have "computer savvy" folks who complain when they can't immediately figure out how to do the stuff that they generally think they know how to do. You have the "tech weenies" who will complain that it never needed to change. And finally you have the "tech gurus" who will see it and love it.

Why cater to the upper eschelon? Any new OS they create (from Win31 upwards) is going to face opposition from at least two of those over-generalized groups I just lumped together. Does that make their OS bad? Nope, but someone will think so and will shout loudly to anyone and everyone with ears or a browser.

You can't please 100% of the people 100% of the time -- I think, so far, Microsoft has done a great job of pleasing most of the people for the majority of the time. If Linux could say the same, it would be gaining marketshare. Maybe that's why Apple is (incredibly slowly, mind you) starting to pick up a little? Dunno...

 

[Rodéric]

Counter argument: where do you draw the line / how do you define legacy? Is everything before this point considered legacy? What pieces do you eliminate, what pieces do you depricate, what peices do you allow? Who makes this decision? How much time do you spend on virtualizing those legacy components? What is an acceptable performance tradeoff (Virtual will never be at "full speed", obviously)

Well as IMO the OS should be rewritten from scratch with the most up to date knowledge and paradigm, nothing should be kept.
(Obviously you'll have to keep the lessons you learnt from your mistakes in mind when writing the new OS.)

As for progress on "groundbreaking OS paradigms", Microsoft Research is the responsible Redmond party for doing such things. A recent example would be the Singularity operating system, which is only open to certain tech colleges, but is at least one example of them thinking a bit further outside the box.

I am aware of this.
But they state clearly that it's an experiment, if it makes it to a brand new OS, that'll be interesting...

Again, this really boils down to a huge problem: you have "computer morons" who will complain when the computer doesn't act like an appliance. You have "computer savvy" folks who complain when they can't immediately figure out how to do the stuff that they generally think they know how to do. You have the "tech weenies" who will complain that it never needed to change. And finally you have the "tech gurus" who will see it and love it.

Simple is beautiful ?
As long as the OS does its job well, fast, reliably and have a nice, solid, user interface, with well set defaults, and simple controls... I don't see who would complain...

Why cater to the upper eschelon? Any new OS they create (from Win31 upwards) is going to face opposition from at least two of those over-generalized groups I just lumped together. Does that make their OS bad? Nope, but someone will think so and will shout loudly to anyone and everyone with ears or a browser.

You can't please 100% of the people 100% of the time -- I think, so far, Microsoft has done a great job of pleasing most of the people for the majority of the time. If Linux could say the same, it would be gaining marketshare. Maybe that's why Apple is (incredibly slowly, mind you) starting to pick up a little? Dunno...

Pleasing everyone doesn't seem possible, but pleasing most people and not displeasing many is possible...

More is not better, the OS should only provide basic functionnality and leave room for applications to be sold as replacement.
(Basic media player, basic burner...; basic as 'incomplete')

 

[Blazkowicz]

Heh. Look at how long it's taking for x64 to really penetrate the market. And it offers extremely flexible legacy support.

not that extremely flexible, else 16bit support would be retained on 64bit Windows.

 

[INKster]

not that extremely flexible, else 16bit support would be retained on 64bit Windows.

The 16bit mode is still supported by x64 (AMD64 and Intel 64/EM64T) processors, so Microsoft's decision was really made in order not to compromise 32bit/64bit stability just for the sake of a -really old and decadent- software "baggage".

 

[Blazkowicz]

http://en.wikipedia.org/wiki/X86-64

Removal of older features: A number of "system programming" features of the x86 architecture are not used in modern operating systems and are not available on AMD64 in long (64-bit and compatibility) mode. These include segmented addressing (although the FS and GS segments were retained in vestigial form for compatibility with Windows code)[3], the task state switch mechanism, and Virtual-8086 mode. These features do of course remain fully implemented in "legacy mode," thus permitting these processors to run 32-bit and 16-bit operating systems without modification.

that VM86 feature was introduced with the 386, it virtualizes 16bit CPUs at native speed and is used by windows from version 2.1/386 to 9x for DOS support, and on NT for DOS and 16biy windows support.
see http://en.wikipedia.org/wiki/Virtual_DOS_machine

 

[Quitch]

Personally I'm loving Vista, and it's certainly the most interesting (and biggest change in the) OS that Microsoft have made since 3.1 to 95. It's pretty shocking though how utterly unprepared the driver writing world seems to have been though. I recall XP being bad for scanners and some printers, but damn, with Vista everyone just seemed to be pointing at their "Vista ready!" kit, though in the case of Creative their "Vista ready!" kit didn't seem to have a non-beta driver at the time...

Vista sucks0rz... ram Hog, useless features, etc... evil.

or

Pff Vista ultimate 64bit rockxz0rs on my QXZ6969Ghz core, i had to nail it on the desk so it wont run away with so mutch powah!

ok, now forget what you read, and look at how the market behaved since Vista came out. Whatever M$ says its obvious Vista is failing to meet M$ expectations, they had to extended XP support due to demand! That speaks for itself in general how the consumer and enterprises see Vista.

And obviously Windows XP was a failure because they had to extend Windows 98 support... oh no, wait, backwards compatibility and long-term support is one of the reasons Windows is a good OS.

PSK uses a user defined password to initialize the TKIP, temporal key integrity protocol. There is a password and the user is involved, for the most part that means it is flawed. The TKIP is not really crackable as it is a per-packet key but upon the initialization of the TKIP, like during an authentication, we get the password (well the PMK anyways). A robust dictionary attack will take care of a lot of consumer passwords.

Yes, because rewriting from scratch worked out really well for Netscape. Now multiply the complexity by a million... oh yeah, I'm sure MS will be rewriting from scratch any day now... WTF? What possible benefit is there to dumping your 3D accelerated desktop? Your memory manager? Your caching system? Backwards compatibility is a STRENGTH, not a weakness.

Not to mention that each revision does dump certain elements of legacy support. To start from scratch... they'd have to be retarded or suicidal.

 

[swaaye]

To start from scratch... they'd have to be retarded or suicidal.

Or you get the equivalent of G80. Obviously a hugely different angle, but there are some similarities. It would be risky and a huge undertaking. I guess it all depends on whether there are potential gains. Maybe there aren't?

 

[Tokelil]

I thought a lot of legacy stuff was dropped when running 64-bit programs already. (It requiring a recompile at a minimum anyway.) At least the ABI is redesigned etc.