What Comes After the Windows Era?

As a computer journalist for the last 25 years, I've received a lot of review copies of software. As something of an obsessive magpie, I've tended to keep most of it, “for reference”. Until yesterday, that is, when I finally threw out all those copies of OS/2, Lotus SmartSuite, and my entire collection of Microsoft software. This included Windows NT 3.5, Windows 2000, Microsoft Office and many, many more. What's makes this little spring-cleaning exercise particularly apt as well as cathartic is that all of us - and not just me - may finally be witnessing the end of the Windows era.

The most obvious manifestation of this has been the utter fiasco that is Vista. This is not so much things like the recent security hole in Vista's memory protection, which may or may not be a big deal – although if Bruce Schneier sees it as a real problem, then I think Microsoft's probably in trouble. No, it's the fact that however Microsoft tries to spin the numbers, Vista is simply not catching on. The company can try all the marketing tricks it likes, but customers are voting with their wallets by opting for the “downgrade” to XP:

An official at PC maker Hewlett-Packard said the majority of the company's business customers are downgrading new computers from Windows Vista to Windows XP -- raising questions about Microsoft's Vista sales figures.

Another tell-tale sign that the days of Windows as we know it are drawing to a close are the increasing number of stories about Midori:

Microsoft is incubating a componentized non-Windows operating system known as Midori, which is being architected from the ground up to tackle challenges that Redmond has determined cannot be met by simply evolving its existing technology.

SD Times has viewed internal Microsoft documents that outline Midori’s proposed design, which is Internet-centric and predicated on the prevalence of connected systems.

Midori is an offshoot of Microsoft Research’s Singularity operating system, the tools and libraries of which are completely managed code. Midori is designed to run directly on native hardware (x86, x64 and ARM), be hosted on the Windows Hyper-V hypervisor, or even be hosted by a Windows process.

That is, we may be looking at a complete dicontinuity in the Windows line – a recognition that the creaking software architecture that has been around for over a decade now is simply too broken to be fixed.

There are a number of straws in the wind indicating that Windows no longer provides the benchmark for desktop computing. The continuing rise in popularity of Apple's machines is a clear indication that more people are looking beyond Microsoft, something that was reflected in Mark Shuttleworth's recent exhortation to the free software community:

"The great task in front of us over the next two years is to lift the experience of the Linux desktop from something that is stable and robust and not so pretty, into something that is art," Shuttleworth said to applause from the audience. "Can we not only emulate, but can we blow right past Apple?"

Quite correctly, Shuttleworth sees Apple as the one to beat these days.

The traditional desktop experience is being further redefined through the move to cloud computing solutions, which are frequently browser-based, and the appearance of GNU/Linux-based ultraportables, which also draw on Net-based solutions for much of their convenience. Both make the underlying operating system moot.

And as if that weren't enough, there's yet another interesting approach that could have a substantial impact on people's perception of free software – and of Windows. The Splashtop technology uses GNU/Linux as a kind of pre-operating system that lets users boot up very quickly to gain Internet access and perform key tasks such as checking email or moving around the Web. Strikingly, some of Splashtop's key advantages are essentially reactions to two of Windows' problems – speed and security:

Be online seconds after you turn on your PC. Why wait for Windows to load when you could be surfing the web right away!

Surf the Web safely, immune from the malware that targets Windows.

What started out as a cool, but rather marginal, idea has now started gain momentum. First, it was announced that Asus motherboards would be offering the feature, and then Asus notebooks. In a sure sign that the approach is entering the mainstream, now Dell has joined the club:

Dell’s Latitude On works by bypassing the Windows operating system so that you get immediate access to things like your calendar, email, Internet, and contacts. It’s a fully integrated technology that will appear later this year on the Dell 4200 and 4300 Latitude series notebooks, and is powered by a Linux OS, sort of as a secondary operating system.

What this means is that users of these systems, even when rigorously Windows-based, will be experiencing the power and speed of GNU/Linux, albeit unwittingly, when they use this mode. Even if they don't know it's GNU/Linux, they will certainly be aware that it is not Windows. Moreover, if the idea proves popular, it's not hard to imagine most PC manufacturers following suit in order to match Dell's move, ensuring that many more people experience the fact that it is possible to use a PC without using Windows.

That is the key breakthrough. At the moment, there is a widespread if tacit assumption that desktop computing *is* Windows. This is manifest most clearly in schools, where teaching computing really means teaching children how to use Windows and Office. Once people realise they have a choice, then Windows' stranglehold on the desktop market will be fatally loosened.

For many years, people in the free software world have dreamed of a day when GNU/Linux would replace Windows on the desktop. Although the market share of GNU/Linux there is finally lifting – figures range from 3% to 11%, according to the sector – it seems unlikely that GNU/Linux will ever take over the desktop from Windows. But that does not mean that Windows will maintain its dominance there, simply that the future is more complex than the monoculture we have seen and suffered for nearly two decades.

I would like to suggest that the free software world should start looking at things from a different perspective – not how many percentage points GNU/Linux gains on the desktop, but how many Microsoft is losing to *all* of the alternatives to Windows. Free software has nothing to fear from a heterogeneous environment – indeed, mixing technologies almost forces open standards upon manufacturers if they want to provide full interoperability.

Moreover, as many have pointed out, the low number of GNU/Linux viruses is only in part down to the latter's superior design. If GNU/Linux became as dominant on the desktop as Windows, there would be a greater incentive to break that security – and a bigger knock-on effect when it was broken. Far better to promote heterogeneity, which is good for the software ecosystem, and thus good for *all* users.

Against this background, then, we should be celebrating what looks increasingly like the end of the Windows era on the desktop, but not in the naïve, and slightly selfish hope that it will usher in one purely based on GNU/Linux. One of the best proofs that free software is superior in all senses to proprietary software is having the maturity to recognise that the world of personal computing today is far more complicated, and far richer, than Bill Gates's original, monopolistic vision of “a (Microsoft-based) computer on every desktop”.

Glyn Moody writes about open source at opendotdotdot.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.


Saint Linux's picture

Nice article. Thank you. Matter is that I'm depressed about Linux for its unhappy and ill looking Graphical Interface especially like gnome, new KDE. There's something not spontaneous like Windows. Can anyone tell how can we make Linux more elegant, more friendly in both Textual and Graphical Interface..

By the way, sorry for out topics..

We're getting there

Glyn Moody's picture

Just compare now with a few years ago....

overlooked GNU/ prefix opportunity

mario's picture

Hey, Glyn. Nice article. I very much like your unobtrusive attention to details. There is however one occurence of the word "Linux" which you forgot to prefix with the GNU/ reminder.
For example when citing people, you can correct the wording if you place the fixes into square brackets. So with the Shuttleworth comment, you could have written: "...... lift the experience of the [GNU/]Linux desktop from ......"

I'm a bit torn

Glyn Moody's picture

This is what my inner RMS wanted me to do, but equally it would have been slightly disrepectful to Mr Shuttleworth to be putting in that implicit square-bracketed criticism.... What's a chap to do?

Microsoft Declining, Cloud to Remain Flat

JohnMc's picture


I think your observations are reasonable. Especially the note that the desktop space is MS's to lose which seems to be the case. I don't think it unreasonable over a 5-10 period to see Linux acceptance at the 25% mark as a portion of market share.

Where I might have a quibble is the expectation of the cloud being a bigger player. Couple of factors are in the way --

1) Tech is very democratic price wise. The price I pay at Frys for a hard drive off the shelf is not that much different from what Amazon pays for those drives by the box load. Run the numbers and you see 3-4 months of S3 for storage and you have paid for the cost of the drive onsite.

2) Broadband is not where it should be at a bandwidth/pricepoint to make the cloud price competive for a replacement of the OS. The only service of any significance is FIOS and its not sufficiently deployed nationally.

The cloud will continue to play a role but it will be limited to certain situations -

A) Offsite storage where the costs are less than say a physical service like Iron Mountain.
B) Certain services where the need is not consistent and the customer pays by the dip.
C) Company has capital constraints on IT acquisitions that can be differed as a expense stream using a similar service online.

That is about where we are right now in the online space. I would not be surprised to see that usage change till we see massive changes in price structures of data transport.

Linux still gains regardless.

Good Article!

FredR's picture

I wholeheartedly agree. I always like to read your writings. I find myself just nodding my head in agreement, as if you're putting into words many ideas a lot of us already know for fact.

I'd like to extend a bit on what you mentioned. The Windows Monoculture. I've noticed a paradigm out there and people are happy to live within the little bubble. They're intimidated by technology in general, and they believe Windows makes things easy enough for them to cope.

In fact, when I left my last job (an MS Gold Partner) part of it was due to the fact that I consider myself a "Linux Guy". The president actually said to me "well, I can't change the entire paradigm". I felt like telling him to look outside the box he's put himself in.

I didn't grow up entirely inside the Windows bubble. I grew up around it, with technology and cut my teeth on so many different things, Windows was one tool in the toolbox. I decided I preferred Linux and it's corresponding tools because they were just better. And they were there, available, with a low barrier to entry.

Don't get me wrong ... I own some Windows clients and am rather adept at using them. But I look forward to the day that the idea which takes majority is the best technical solution from the opinions of software developers and engineers, not marketing based on what the masses will find "easy" because of their comfort level with technology.

-- FLR or flrichar is a superfan of Linux Journal, and goofs around in the LJ IRC Channel


Glyn Moody's picture

...for the kind words. I'm glad you like the telepathic element.

What Comes After the Windows Era?

paul (the unverified)'s picture

Oh man! I didn't know there was gonna be a quiz! You know I've studied for this question for a long time. Ah, let me see...

(Insert expectant pause here)

What comes after the winDOHs earache ...er, urea ...ah, error ...ire?!

Improved computing security? Stable operating systems? Software applications that actually work with interfaces that make sense? Lower costs to business re software and hardware? Fewer computers going into landfills? A bunch of m$ trolls going on unemployment? Steve Ballmer opens a 'pre-owned furniture' store with chairs that have been thrown only once? Steve Ballmer becomes the spokesman for the Hair Club for Men? A sudden influx into the job market of one-time spammers?

Oh. I know. All those 'Life Coaches' suddenly have an increase of clientele from all those winDOHs people with MCSE certs.

future is linux ? - not with these software developers

wwwpirate's picture

keep dreaming boys
linux will NEVER move beyond linux type ppl unless linus exercises some power among developers - hippie mentality does not play with businesses and general population
linux software programs are too buggy, not much tested before releasing final versions and program version is just a joke - developer writes 2 line code that does not work at all and calls it version 1.1
my experience - tried to substitute ms word with open office (i write books) - does not work - oowriter is a joke for any serious job
linus needs to do some spring cleaning among developers (here i think mostly abt program version number) and allow some commercial opportunity otherwise there will be no significant move in linux market share - it is simply not good enough for general population and businesses
i'm talking here about linux applications not kernel

Wow... really?

esteban's picture

you write books OMG that must be where you get them mad punctuation skillz

im so impressed wth how claerly you express you're self wwwpirate u must b rich

linus needs to do some

Blackhole's picture

linus needs to do some spring cleaning among developers (here i think mostly abt program version number) and allow some commercial opportunity otherwise there will be no significant move in linux market share - it is simply not good enough for general population and businesses
i'm talking here about linux applications not kernel

If you are not talking about the kernel, why do you bring up Linus? He has nothing to do with GNU, GUIs, or applications. (OK, he did submit at least one patch to the Gnome project. Maybe a few other isolated submissions. But he has a full time job with the kernel.)

Perhaps you don't know what you are talking about and you are just trying to kick up some dust so others can't see clearly? If I'm wrong, please submit some cogent arguments that indicate you know what you are talking about.

Word is better? ...ahh, no.

paul (the unverified)'s picture

This has not been my experience. I've been using OOo on winDOHs and Linux for the last 6+ years to process, edit, create all sorts of MS docs and spreadsheets for a client that is winDOHs only. All of my work is an improvement on theirs since I'm more familiar with formating and layout. I've also been using NeoOffice on the Mac for the last 3 or 4 years. I've always had an excellent experience with the software.

OTOH, I have had many frustrations with MS Word. I'm also disappointed with the deterioration of the UI in Word. Back in '97 and '98, I used it for docs that consisted of several sections. The UI at that time was becoming bizarre and inconsistent, but I was able to get my work done. But for some reason, MS thinks it knows better and really began to screw with the UI not to mention bloating the daylights out of the software.

The writers in my department all used PageMaker or FrameMaker. NOBODY wanted to use MS Word. NOBODY. Because it sucked. I was the only one (I being the manager of the group) that worked with it.

So when a product manager came to us with a doc they had started in Word, rather than go through a messy convoluted conversion and since prod managers didn't really understand how to do much formatting to begin with, I finished it off.

I was not about to expose the writers to an enormous amount of frustration and wasted time.

I tried Word from time to time up until OOo came out and it still sucked. That's when I switched. I knew OOo would get better and MS Word would only get more bloated and more clunky.

I'm glad you're able to work with it, but from my experience as a manager of a group of technical writers for 12+ years, MS Word does not cut it. It's fine for business letters, but if you have to create a larger doc, it's just no good. Even if you can keep it or winDOHs running.

Over the years, I've almost completely migrated to Linux and OSS. The only proprietary software that I still use (occasionally) is Acrobat Pro, BBedit, and Bryce for my 3D stuff. I'm in the process of replacing BBedit with Bluefish and Acrobat Pro with several smaller light-weight apps. Bryce will be tough to replace, but I know there are good options available.

So no. MS's software is the weak product. OSS on Linux is fine. I use a great deal of OSS a great deal of the time. I'm not alone. There are many people doing this. And we're all saving lots of money and avoiding a great deal of irritation because of better stability and security.

here is reply

wwwpirate's picture

pls answer these questions
i shall really appreciate
here is my opinion about oowriter
this is 1.0 beta not 2.4.1 version
i really appreciate an effort but result is not there
do not like ms same as you do but buddy they have something that works well with minumum effort to learn product

winDOHs??? Really?

Anonymous's picture

You can make any claim you want. But after "winDOHs", "M$" and all other derivatives it's impossible to take you seriously. I don't believe a single word you say.

At least an entertaining read...

Kim's picture

... but besides that not more than a confirmation that some folks get stuck in the past, or even worse an imaginary illusion of a past. By the way what personifies "Linux people" from users in general?

Furthermore, many of us involved in what you call business already use Linux and BSD the way the article describes as a heterogeneous environment, and the Microsoft slice of the mix is declining. A decision made solely on the ground of technical merits. I actually don't care whether the developer looks like a hippy or dressed in a tux(edo), or what mentality he or she has, or as long as the code is effective and secure. Some things work well in Windows, and others are a nightmare, hence it's more pragmatic to be open to alternatives.

A couple of notes concerning your choice of word-processor: many writers would even argue that a word-processor is a poor choice, and that document processor like LaTeX is far better suited for the task. So opinions differ, and your "truth" isn't the absolute truth about computing. Anyway, it's still a good idea to have OpenOffice included in the software mix, because at times you actually come across Microsoft Office produced documents that can't be opened by current version of Microsoft Office (some updates have even broken fairly newly made documents). I've use OpenOffice as a rescue in many such situations.

Not Wrong

Ken Sarkies's picture

When I was teaching Engineering all the postgrad students used LaTeX for their theses. One guy decided Word was the way forward. When he got about 150 pages done, Word corrupted his document. He went back to the last good backup and tried re-entering his day's work. Again Word corrupted it. In the end it cost him a huge amount of time, frustration and effort splitting up the document into parts that Word would handle. This is where prejudices start.

Interesting suggestions

Glyn Moody's picture

I'll keep that list handy for future reference....

Awesome article

Anonymous's picture

Congratulations Glyn, that's one of the best articles I've read in a long time. I loved your comments about having an heterogeneous environment in particular. Even as a fellow LJ author, I have to admit that I'm not necessarily a Linux advocate, but an open source one. It's not Linux itself that captures our imagination, it's open source surely?

There are some very cool looking alternative OSes being designed, and I certainly wouldn't want to steer anyone away from them into Linux for some weird reason. I think Linux has made for itself the biggest name in OSS, but Linux itself isn't really the issue. The more diversity in this environment that is picked up by the mainstream, the better.

John Knight


Glyn Moody's picture

...for the kind words - much appreciated.