Industry


Ads by TechWords

See your link here


Subscribe to our e-mail newsletters
For more info on a specific newsletter, click the title. Details will be displayed in a new window.
Computerworld Daily News (First Look and Wrap-Up)
Computerworld Blogs Newsletter
The Weekly Top 10
More E-Mail Newsletters 

Microsoft has serious plans to do away with Windows

It's one thing to have a skunkworks operating system project, Midori, that could conceivably replace Windows. It's another to actually have plans on how to switch users from Windows to Midori. Guess what? Microsoft actually does have such plans.

According to David Worthington, Microsoft isn't only building a Windows replacement operating system, its "carefully conceptualizing a way to move millions of users away from the existing Windows codebase and onto Midori, a legacy-free operating system."

The plans, which are far from being finalized, indicate the Microsoft is really running scared of Mac OS and Linux on the desktop. "Midori's legacy-free objective [is] a preemptive strike against non-Microsoft operating systems, enabling the company to compete head-on by enticing customers to replace Windows with Midori instead of a non-Microsoft OS," wrote Worthington.

I know many of you still have trouble with the idea that the Mac or the Linux desktop could possibly challenge Windows. Microsoft disagrees. Look at the numbers.

Context, a European business and consumer technology research firm has just reported that pre-installed Linux was sold on 2.8% of all PCs bought in the UK in June 2008. Doesn't sound like much? In January 2007, desktop Linux's marketshare was 0.1%. As for Apple, in its last quarter the company sold more Macs than in any previous quarter and its sales marketshare in the States is now up to 14%.

This isn't the 90s anymore. In the aftermath of the Vista sales flop, Windows is no longer an automatic choice.

Microsoft knows that and seems to be seriously considering a future without it. Specifically, Microsoft seems to be considering three different paths. In the first, Midori would run under Windows. This doesn't sound terribly practical and Microsoft appears to realize this. The second plan is "the development of an executive for Midori that is based on and would run in parallel with the Windows executive." This would amount to little more than an augmentation to Windows.

The final plan, and the one that strikes me as the most interesting and easiest to develop, is to actually create Midori as a legacy-free new operating system and use a built-in hypervisor to run Windows in a virtual machine. This could work. Virtualization is quickly becoming mainstream as people realize that not only can you really run multiple operating systems efficiently on a single machine, but it can really save you money.

For Microsoft, it would also have the advantage of ensuring near 100% fidelity of the Windows experience for users and their applications. Underneath the hood, the engine would be running Midori, but end-users would still see their familiar Windows applications.

Of course, you can already do that today. For example, I use VirtualBox to run Windows XP SP3 on my Linux systems. Windows, I might add, runs with a lot more stability on top of Linux and VirtualBox than it does on native hardware. Red Hat, with its KVM plans, foresees running everything and anything on top of Linux-based virtual machines.

Will Microsoft actually make the jump? As I said before, I think a lot depends on the management. Ballmer doesn't inspire me with any confidence. If he were a GM executive I can see him insisting that the company invest more in the Hummer because of its profit margins while ignoring that with the price of gas no one is buying them anymore. For those who don't follow the car industry, GM is now trying desperately to sell its Hummer division.

Even if Microsoft does pull the trigger, I can't see Midori showing up until 2013. By that time, it may be too late to stem Microsoft's decline. Still, I find it interesting that Microsoft is even considering a radical operating system shift. This really isn't Bill Gates' Microsoft.

What People Are Saying

Ego.

Ego.
Could this be Microsoft's weakness?

If I recall "Back in the day", when MS was playing with Xenix, they chose to employ some VAX guy who gave them what later generations would lovingly term "that piece of sh*t Windows NT".
12 years after its introduction MS has done a creditable job of getting it sort-of stable but the big question is: "should they have?"

If I were in charge (and lord knows I am painfully aware I am not judging by the lack of chaufeur dirven limousines outside) I would have FreeBSD with a really polished front end. In short, I would have done an Apple but for the Business User.

FreeBSD has a completely free license, is solid as a rock, scales, resists malware and generally goes about its business like a Proper OS. So why didnt MS grab it with both hands?

Its Ego. Because to do that would mean tacitly admitting the Windows is what most of us have been saying for more than 10 years: sh*t.

So they keep on persevering with this astoundingly complex hairball when all they had to do was use a UNIX (est 1969) derivative with proven track record on the desktop and in the datacentre. Marketing and Apple's apparent predilection for style over function would have done the rest. At least in the lucrative business space.

It is interesting that of the three scenarios suggested for Midori not one is (for me) the simplest to do and most obvious to go for: a *BSD derivation. A new Windows running under and old Windows? A Windows running on a hypervisor... wtf? To this old engineer it sounds like more of the same. Ego needing to keep that Windows source code alive with another patch, another workaround, another layer of complexity. Will they never learn?

The OS is not "the PC"

"I know many of you still have trouble with the idea that the Mac or the Linux desktop could possibly challenge Windows."

This confuses the applications used with the OS as a whole. Put a "kid" in front of Windows, OS/X or Ubuntu Linux and the "kid" will be able to get around on any of them. A GUI hasn't been a problem for a Unix Desktop for a long time now.

An OS is as good as it protects you and runs your stuff well. I do believe that precludes all Microsoft OSes, including Vista. Ubuntu Linux looks better to me than anything Microsoft has done to date; and I don't care to familiarize myself with Vista, because the Operating System is not THE REASON that I run a computer, and I don't want to spend any time "getting into" the "new desktop methodology" that Microsoft has DECIDED I should use like it or not - on Ubuntu if I don't like the GUI I can use a different one, never the case on Windows.

I fear the majority of people reading here still buy into following "what Microsoft does" ... I used to do that until I realized that my distaste for their methodologies was a deep truth my subconscious was waiting for me to understand. Every next wonderful development methodology they offered became more and more encrusted with "Microsoftisms" until I gave up worrying about buying the latest quarterly update on how to program through their stuff and switched to FreeBSD the moment I could browse Usenet with it. That was 1995, and I never looked back at Microsoft for development purposes again.

We need a paradigm shift: Microsoft <> Computing and it never did; that was useful propaganda for them. People learning computing nowadays should learn on open source systems, because that's what we'll all be using in times to come. On Linux, I have a choice of some 15 file managers, and if I wanted one better than any of them I could write it starting from the SOURCE CODE of one I liked.

Not so on Windows- right?

I'm afraid you're right...

Microsoft's biggest problem is Ballmer. He's presided over a succession of balls-ups, Vista and Zune to name but two. 7 looks good, but Microsoft's vision for the future is blurred, to say the least. Not something you could ever accuse the company of when Uncle Bill was in charge.

I run Vista on my main PC

I run Vista on my main PC and I really do see huge problems with it; they did a lot of the actual interface well (Explorer is the best file browser I've ever used), they really screwed the pooch on things like speed, compatibility, etc...the thing is, what I hear happened in MS, is that the original project manager dicked around for so long it just became a mess of failure, and the only reason it was even released when it was (which nobody believed was going to happened, hence the poor state of readiness from the software and hardware community) was a new manager.

So long story short, I really hold hope for a "new Microsoft" that does things right for once. I think they're maturing to the point of seriously understanding how people want computers to work.

Change or die...

it's not survival of the fittest... it never was... it's survival of the most adaptable!

Not the most adaptable...

It's not survival of the most adaptable, it's survival of the one with the biggest club and most money in the marketing department - VHS vs Betamax, Blueray vs HD-DVD, Word vs Word Perfect, Windows vs OS/2, Windows vs Netware, etc. were all won by inferior products but better marketing.

Vista is the biggest marketing advantage Linux has ever had but with all the variants out there, most with no corporate-level support infrastructure, it'll still struggle to break into the corporate desktop world (Apple's been trying feebly to do this for years and still hasn't succeeded).

corporate support

You are right, in fact I am running a SLED 10 SP1 box - it is amazingly stable, secure (as far as I can say...) and, yes, beautiful.
A couple of remarks, however, are, in my opinion, necessary:

a) SLED 10 makes it difficult to allow your computer to install newer third parties applications. This is something you always need to do in a corporate environment, where sometimes you need to launch a video DVD prepared by a client, work on multimedia files and similar operations. Unless you open a [virtual] Windows session you are likely to be unable to be 100% productive. This MUST be fixed by Novell - somehow, I do not know, but they MUST fix it. Computing corporate use is no longer a text, spreadsheet and presentation paradigm only.

b) Novell support is ok as far as you have your SLED 10 box untouched - i.e. if you run it with no third parties software. Again, if I do so I keep my productivity to a minimum, unless I have an internal support team (but then Novell would earn almost nothing...).

c) corporate use has certain requirements, which are, at least for me, key and basic. One for all: full hard drive encryption along with pre-boot authentication. With SLED this is a pain in the *ss, and this MUST be fixed.

In a few words, when it is about corporate use, it is not only staying with a hyper-tested and locked operating system, it is also about understanding that corporate use means a lot of [unpredictable] user behaviours and needs and this is when the stakes raise.

Now, I like Linux, I like my SLED box, but honestly, for this to be adequate to today's corporate use requirements, much more effort must be put in.

you're gonna hate this, but for the most part you are right

Vista has problems. Linux can't capitalize because the Open Source movement is so fractured with hundreds of operating system distributions, Corporate America (and the rest of the world) are scared off by the lack of stability. No one knows what the next hot distro will be, so stay with Windows. It's stable. Not stable from an OS perspecive, but from a marketing perspective.

By the way, I lived through the OS/2 - Windows wars. OS/2 was vastly superior. Microsoft knew that, but understood the market much better than IBM.

IBM charged $1000 for a developer kit. MS gave them away for free. Guess who had the most applications to go with the OS? MS beat IBM simply because the user applications were plentiful and functional. Note, I did not say wonderful, just functional. People bought the OS for the applications.

Big problem with Open Source is the peripherals. Printer drivers specifically. To the average user, they need to be able to put a disk in and install the driver. Not so with Linux. Very geeky solutions that may require compiling the driver for your Linux distro.

Even finding the driver can be a difficult path. I found one for my printer with a rating of "Works, Mostly". Now that instills confidence! Not good enough for the average person to adopt Open Source for that level of service.

Even Vista has printer drivers and video drivers, and other drivers that simply install from Disk. Not so with Linux. Getting better but still highly distro dependant.

Yes, vista has problems, but perceived problems with Open Source are much higher.

what drivel

QUOTE
Vista has problems. Linux can't capitalize because the Open Source movement is so fractured with hundreds of operating system distributions, Corporate America (and the rest of the world) are scared off by the lack of stability. No one knows what the next hot distro will be, so stay with Windows. It's stable. Not stable from an OS perspecive, but from a marketing perspective.

REBUTTAL
Untrue. There are actually about 3 main distros. The Redhat series, Debian Series and Gentoo series. Users are "scared" off because they are ignorant. Ignorance breeds fear and fear breeds inertia.

QUOTE
By the way, I lived through the OS/2 - Windows wars. OS/2 was vastly superior. Microsoft knew that, but understood the market much better than IBM.

OBSERVATION
Yep. I lived through it too. OS2 was technically superior but ran like a dog.

QUOTE
IBM charged $1000 for a developer kit. MS gave them away for free. Guess who had the most applications to go with the OS? MS beat IBM simply because the user applications were plentiful and functional. Note, I did not say wonderful, just functional. People bought the OS for the applications.

OBSERVATION
Yep. I agree. I was "there".

QUOTE
Big problem with Open Source is the peripherals. Printer drivers specifically. To the average user, they need to be able to put a disk in and install the driver. Not so with Linux. Very geeky solutions that may require compiling the driver for your Linux distro.
Even finding the driver can be a difficult path. I found one for my printer with a rating of "Works, Mostly". Now that instills confidence! Not good enough for the average person to adopt Open Source for that level of service.

REBUTTAL
Nonsense. This might have been true 5 to 10 years ago but any mainstream laser will most of the time plug via USB into a Linux box. The Linux box will see it, recognise it, load the right module and have a test page in your hand _without even needing an installation CD full of cruft_. I recommend Samsung.

QUOTE
Even Vista has printer drivers and video drivers, and other drivers that simply install from Disk. Not so with Linux. Getting better but still highly distro dependant.

REBUTTAL
Unmitigated garbage. Simply avoid such problems by being educated about what manufacturers make good peripherals.

QUOTE
Yes, vista has problems, but perceived problems with Open Source are much higher.

REBUTTAL
Gah where to begin. These perceptions are fuelled by posts like this.

May I follow this up with a few observations of my own. I own an IT company. Nothing big but we look after a few folks. Maybe 40 windows machines, 8 linux servers and 2 macs. The only ones that have no issues are the macs and linux machines. The windows machines always need hand-holding, clearing of cruft, a/v scanning, spyware checks and so on. Of all my customers 2 came to me with significant problems with their windows networks. After I was done they wondered out loud how I had transformed the reliability of their shops. Easy I said. You are now on Linux. Of the two Linux machines converted I havent had a single support call in a year. They run printers, fileservers, NT domains, backups, RAID, etc etc etc. One got the company aerospace ISO9002 certification. And none cost my customers a dime.

Once they were educated they ceased to be afraid. Once they were a fraid they realised the technology is simply there to allow them to do their work.

Judging by this post the task of education is far from over.

Drivers

"Big problem with Open Source is the peripherals. Printer drivers specifically. To the average user, they need to be able to put a disk in and install the driver. Not so with Linux. Very geeky solutions that may require compiling the driver for your Linux distro."

I'm willing to "nominate" Ubuntu Linux to be "the" supported distribution. If the various distributions do drivers differently, rather than wait for "which is best" let's tell them we need ONE we'll care for - assuming Ubuntu == Debian at the driver level, Debian is Good Enough To Use.

Then tell the manufacturers that if they do not PROVIDE ALTERNATE DRIVERS, YOU WON'T BUY THE PRODUCT. I have an HP 1600 Color Laser printer for which the Ubuntu drivers do fairly as opposed to the Windows drivers which do quite well. I see no reason to accept this state of affairs, insofar as HP could have written a Debian driver for that unit in a week and have had it online for 2 years now, if they gave a damn. We could download it as whatever equates to an RPM from HP and be done.

So make them give a damn. Complain.