Information Weeks Editor in Chief Rob Preston really needed to do massive editing of the 6-page article by Alexander Wolfe on 7 Reasons Why Linux Wont Succeed On The Desktop. This article purports to examine the issue of Linux on the Desktop thoroughly and bias free. Let us cut to the chase and see what Alexander Wolfe has to say about the future of Linux on the Desktop in his conclusion:
“What now? Perhaps 2010 will finally be the year of desktop Linux. Unfortunately, I believe that that the seven arguments Ive set forth in this piece make an airtight case that this isnt going to happen.”
Well, Mr. Preston this is a very strong conclusion so lets take a look at the AirTightness of your Wolfe cub-reporters case.
1. Prohibitive application porting costs.
One of the strategic advantages that Microsoft enjoyed throughout the 1990s was that its developer tool set, Visual Studio, was at least two steps ahead of any available cross platform solution (IBI Focus, Borlands Builders and plethora of 4GLs or generation tools) which lacked either performance, GUI ease of development, equivalent functionality, or some other critical capability. So it should be no surprise that there are many small to medium tier software vendors that committed to Windows only and MicrosoftsVisual Studio tools with one of its highly proprietary languages.This is the so called Developers, Developers, Developers advantage of Windows.
However, since the rise of Open Source, Java, and Eclipse among other tools that situation has changed dramatically, and Alex completely misses this critical point. For 10 years and running the top priority among CEOs and their IT Staffs is Application Interoperability => make systems work not only anytime, anywhere, on any device or platform for any authorized users – but well and easily with each other. It is the still largely unfulfilled, Information at your Fingertips promise.
But of great importance is the fact that prohibitive application porting costs do not have to be absorbed. There are many cross platform software apps for both desktop and server. And there is a fast growing array of tools from such industrial strength vendors as Trolltech for C/C++, OpenLaszlo for RIA and Media, InterSystems Cache+Ensemble for extreme scale OLTP, sophisticated modeling and simulations from MathWorks plus the phalanx of cross platform Java tools from JetBrains Intellij through Eclipse Frameworks to Suns Netbeans among others. Being Linux, cross platform and GUI savvy has never been easier.
So if the software developer or vendor has not fallen for the Microsoft Siren Song “to use our proprietary features to run best in Windows” then the options for developing into other OS platforms (while enabling a Windows version as well)are not only quite rich but also fully feature equivalent. Thousands of top-rated tools from Adobes Photoshop through Oracles JDeveloper to Wolframs Mathematica are testament to the fact that cross platform and technical excellence can very well coincide.
But of even more impact is that some of Microsofts key software partners have chosen to layoff Linux in the hopes of non-entrancy by Redmond in their software niches. Well kiss that notion goodbuy for all three software vendors that Wolfe mentions. Adobe has seen every one of its major products: Acrobat, Photoshop, Illustrator, Dreamweaver, Flash, and Flash Player targeted with massive new-investment technology from Redmond. So why not introduce a Linux Photoshop or Dreamweaver- they have some natural constituencies in the Linux space. Ditto for AutoDesk and AutoCAD and its full high-end graphic support tools. They crave reliable, cheap computing power and some are already over to Linux – why not a AutoCad Linux mini-suite?
And as for Intuits Quicken line – the only way it will survive the onslaught of Money, the huge MS Business Accounting and Server line, plus new Redmond Live Business Services is if Intuit has an Apple or Linux based box solution to go with its Windows counterpart. Because Redmond is the ultimate player of the “double-down-to-zero” pricing in markets that it wants to dominate – and the DOJ barely has an Antitrust heartbeat, this a huge risk to Intuit and its Quicken product line.
In sum, Wolfe has ignored one of the key trends of the decade – the drive toward open, and interoperable – and the ease at which this can be achieved through Web 2.0, SaaS and ever easier cross platform development tools. In sum, score Wolfe 4 on 15 marks for this point for at least getting the major issue but utterly failing to bring easily accessible and stronger analysis to bear.
2. Linux Nuisance Fanboys
This is one of the most specious arguments in the whole article. If you have had to endure an imperious Apple priest or a goshgolly Microsoft MVP or the Linux Slashdots you have unfortunately seen some of computings various Fanboys. Still if you stay away from Poison Ivy – no harm done. They are a dying breed until rescued by an article like this. Serious demarks to Editor Preston for letting this one get through editing and of course score Wolfe 0 on 14 marks for trying to raise the issue at all.
3.You cant make money on the Linux operating system
Like point 1, this has all the earmarks of being a very relevant point on the issue of Linux Desktop Success. But unfortunately again, Wolfe does not discuss the issue in sufficient detail. Wolfe rightly raises the issue of Linux revenues (through nominal one-time distribution fee plus any ongoing support services) and sharing of those among say 50 active and twelve leading Linux distributors. Even at say 300-500 million Linux boxes worldwide,Linux revenue streams can quickly get divided up. But everyone knows that the top tier players like RedHat, Novell/Suse, and Oracle take the lion shares of the Linux revenue streams and do well by it.
Wolfe raises the other salient issue elsewhere(in the spokesman section) that a substantial portion of Linux users, led by the likes of Richard Stallman believe all software should be free. This has created a tension and some fruitless disputes in the Linux and Open Source community as to how to co-exist beneficially with commercial software+hardware vendors. The classic has been the senseless suppression of Java in Linux distributions(shame on Stallman, BEA, IBM, Oracle and a host of other “Java” vendors and beneficiaries for not coming to the aid of Sun and saying in no uncertain terms to Microsoft if you do not support the latest Java JVM in your browser- then your IE browser is unacceptable in our and our clients PCs and systems. This being done – then Java goes Open Source a hell of a lot sooner).
So this is a critical point for Linux Success on the Desktop – indeed how do vendors make money given the tension in parts of the Linux community against commercial software. Let me posit two forces that are working towards resolving these issues. First, commercially-oriiented Open Source Vendors (think BSD, MySQL, OpenLaszlo, OpenSolaris, Trolltech)with their diverse Open Source models are working with commercial vendors with fairly high degrees of success. Over the past 2-3 years Open Source models have gotten a real wringing in the marketplace- and Enterprise Web 2.0 start-ups will be pushing what I call the Pentaho frontier.
Second, remember that Open Source is partly a reaction to the excesses of Microsofts EULA. Microsofts EULA is the anti-warranty. In effect, Microsoft says once you our customers recognize that we, Redmond, absolutely have no legal liability for the effective working of our software, for supporting our software other than as we so choose and for developing our software in the future other than we so choose – only now are we prepared to consider what kinds of support and at what fees we are prepared to offer you, our customers.
Linux and Open Source tends to turn this around and says to customers – since we make only a nominal distribution fee, we make our money on supporting our software working for you. Lets see what support services and future directions in our software will be possible. Gartner is currently calling this one of the 5 Major Forces on the IT market. I agree – we shall see how the emergence of Open Source and Linux plays out in the years ahead. But the Microsoft EULA Model with its Zero-sum Monopolistic overlay (we reserve the right to enter and takeover any Windows app market using variations on monopolistic pricing) is just not working. Just ask any VC what they demand of IT startup firms looking for new capital – how are you going to cope with Microsoft. So currently IT innovations have to skirt around the Microsoft Monopoly Sinkhole. Linux and OpenSource are market reactions to this problem.
Give Wolfe 7 on 15 marks for raising the right issues (if not in the right place); but still a tendency to be way too lean on current trends and possible breakthroughs.
4. Resistance from average users
So why does the above Janco chart of IE market share over the same period apply? IE managed to reach and retain a nearly 90% market share from Sep 2001 through May 2004 a period when the differential advantage of IE6 was most rapidly waning. Yet the rival browsers were at that time under less security worm attacks, were notably smaller and were easier to download and install. Only the last 3 years, even with IE7s improved security and addition of tabs, has seen a major erosion of market share. Perhaps small tipping points can explain the situation.
Users have little slack time and willingness to explore. For most users a PC is a hammer – a tool the less they have to understand, the better. So the inertia against Linux will be large and will require a significant performance or obvious functionality advantage to offset that inertia and the availability of so many more apps on Windows.
True as the cost of Windows +Apps becomes a significant fraction of the total PC (versus minimum charge of free Linux plus Open Source apps) pressure will build towards adopting Linux. But the relearning-to-use-my-PC cost may still be too great for switching to Linux . No wonder Microsoft engaged in the People Ready ad campaign – they were changing the interface of the OS + Office(even more so) + IE7 browser substantially – so why not promote the myth that there was really no change at all because Vista+Office 2007+IE7 were from Microsoft and it was all just so people ready.
Given that no Linux or PC vendor has capitalized on this great vulnerability in the transition to Vista and Office 2007, the only likely next opportunity for Linux to grab market share quickly will come if OLPC-One Laptop Per Child and Linux Playtoys match technical success with large markets. Also Linux mobiles and savvy embeddeds could contribute. But this party puts a substantial probability on the fact that inertia will keep Linux market share well below 10% for the foreseeable future.
So give Wolfe 12 on 14 marks for this point, full marks if Alex has an OLPC.
5. Linux is “simple”; Windows “just works”
After reading this I am hopelessly confused. Rather than supplying the reader with measurable performance tests for the concepts of “simple” and “just works” , readers are treated to a sociology argument. Meanwhile I and tens of millions of Vista users have very, very serious problems with “Windows Vista Just Works”. Vista is overpriced, over-marketed, bloated and slow on start-up; has serious Windows application compatibility problems; experiences lots of subtle, perplexing networking problems; and let users contend with huge hardware and peripheral support problems…. This is “just works”?????
So give Wolfe 0 – 15 marks for this point and for being Vista “Just Works” blind. Like a lot of the IT Trade Press, Alex has fallen for Vista Blindness – WYSNINWRIJWASTITFAWBMG – What You See Now Is Not What Redmond Intended Just Wait And Some Time In The Future All Will Be Made Good. Currently that date is set for March 2008 … and counting because Vista SP1 simply will not fix the problem – SP2 was required for Microsoft to get it right on XP, a less complex OS.
6.There are way too many Linux distros
Damn – we have been here before. This is really a derivative of point 3, repeated. Nice try Wolfe, score Wolfe 0 on 15. Note to editor Preston, perhaps 6 not 7 “Airtight” points would have worked better.
7. No powerful evangelist for Linux comparable to Bill Gates or Steve Jobs
I must give full credit here. Linux is missing a natural desktop spokesperson. And having a good spokesperson certainly does help. I would beg to differ that Bill Gates has lost a lot of his developer and IT creds with the Antitrust Case, the onslaught of IE virus attacks despite promises of Trustworthy computing, “Cutting off the Oxygen to Nescape” remarks, and cold calculated thwarting of all Web development in IE for 7 years.
But make no mistake, Steve Jobs puts the A-1 Spokesmans Sizzle on all Apple products. And in a smaller way Adobes Cold Fusion continues to thrive through Techno-Evangelist Supreme, Ben Forta. So Linux on the desktop could use a technical evangelist. For example, on the server side, Linux has had that eminence gris, IBMs Irving Wladawsky-Berger guiding near faultlessly IBMs embrace of Linux despite its other OS including AIX/Unix. IBMs support has been critical for Linux and Open Source to make the robust server side blossoming that it is currently enjoying.
But on the desktop side, say Steven J. Vaughan-Nichols of editor of Desktop-Linux has the background and credibility to pull off such a role. But he may not be so inclined. In any case, a Linux evangelist also needs a sponsor and be honest – have you seen a common Linux desktop spokesperson at the Fedora, Suse, Ubuntu, or Xandros meetings (I cant comment as I have not attended nearly enough). Clearly a gap that needs to be filled.
Score Wolfe 15 on 15 for this point.
So we have examined all of Alexander Wolfes points and the score, and like the article, it is not pretty – 38 on 100. Now remember we are assessing this against Wolfes own standard – an airtight case against Linux on the Desktop ever prospering. As I said, Editor Preston was remiss on his job.
But hold on a second. Havent we conceded in our analysis of points 3 and 4 that “But this party puts a substantial probability on the fact that inertia will keep Linux market share well below 10% for the foreseeable future.” What goes on here?
Yes I think that Vista Blindness – WYSNINWRIJWASTITFAWBMG – What You See Now Is Not What Redmond Intended Just Wait And Some Time In The Future All Will Be Made Good, sponsored in no small part by a docile IT trade press, has squandered an opportunity for Linux to gain market share. But remember Vista still will likely retain many of its Bloat, Clot, and Rot characteristics well beyond SP1 due in Spring 2008. It really took until SP2 for Microsoft to get XP right. Vista is bigger, more complex with many fundamental changes.
So the opportunity Window for Linux on the Desktop making big in-roads is still open. Here are the 4 opportunities I see for Linux to gain big market share chunks on the desktop.
1) Linux in OLPC. The OLPC, goes on sale in the US for $400 – 2 copies, one you get to keep, 1 for overseas distribution. I am going to buy one of these – the technology is just too interesting to pass-up. Its like being at the Microsoft PDC conference 5 years back when the new Compaq handheld came out on promo for $400. They got eaten up at the show. I suspect the same will happen with OLPC – given the $200 tax rebate, a lot of gadget and techphiles are going to grab this if they have their distribution right. Can you imagine if suddenly 200,000 or 800,000 techphiles saying the real “Wow” – this Linux handheld is amazing!
2)Linux VMs. The VMs are coming! The VMs are coming! The VMs are coming! if you believe the IT Trade Press. Well, I do on this one and the comparison of a Linux VM running like a bat out of hell on Vista box is going to raise same eyebrows, some corporate eyebrows.
3)Linux Embedded. Mobile phones, play time toys, kiosks, cheap PCs everywhere in the home and at work is going to be an equally important eye-opener. Linux as Open Source and lowest total cost for delivery is being embraced by entrepreneurs and developers alike. This presence everywhere else will lead to inroads on the desktop.
4)Linux is the lowest cost OS on the desktop. If you select a common portfolio of Office, Mail, browser, basic game and one more app that PC users use, Linux will cost out as the much lower desktop OS. And Microsoft has thrown in the towel and closed down the GetTheFacts website that alleged otherwise. Innovation is currently equally strong roughly in both OS desktops, so eventually lowest cost will prevail.
Given these factors let me go out on a limb and put a gut-feel probability to all that has been argued here. I expect with 70% confidence that by 2010 Linux OS of various flavors will have 40% or more of all desktop and/or household PCs and computing devices worldwide. Or equivalently Microsoft will have bought a Linux distro (very likely Novell) and have gone cross-platform bigtime.
This is my story and I am sticking with it.
(c)JBSurveyer 2007 If you liked this, let others know:
Slashdot Digg del.icio.us reddit newsvine Y! MyWeb