What doesn't get mentioned is that Open Source is only just beginning - it will be going from strength to strength as Microsoft hobble from weakness to weakness. People who care about technology use Open Source. Microsoft only people are just Power Users. The winners will be where the intellectual capital goes, which isn't a trough that the pigs can easily stick their snouts into.

>> The city of Munich is spending $4,500 per user to migrate to an open source stack that will probably never offer the functionality or ease of use of the Microsoft desktop.

Many think that today the Vista desktop and Microsoft's newest server offerings really are inferior to some of products available on Linux. I would have to day that your use of the word "probably" is probably too strong a term to use here.

>> In other words, it's an illusion to think that by getting rid of Microsoft you can get rid of vendor lock-in.

The vendor lock-in techniques open source companies provide are different than what Microsoft achieves. It's a matter of degree, of which lock-in is more costly to leave.

>> Unless that development organization is subsidized by someone else with a strategic motive (in the way that, say, Apache is subsidized by the Googles and Yahoos and HPs of the world), it is going to have collect fees from at least some of its users. And it is going to have to collect these fees for as long as anyone wants it to stay in business, which means that it will inevitably develop all sorts of ingenious ways to incent its users to keep paying these fees.

You seem to fail to consider that a large portion of FOSS work (looking at smaller but essential projects and at the many volunteering ideas, solution patches, and any form of feedback) is currently done by volunteers as well as by users that improve their own primary business as a result of these contributions. These end users absorb development costs so that IBM and others do not have to.

Again, it's a matter of degree. Who is claiming that FOSS is not business friendly? FOSS is not *monopoly* friendly. Companies will always be trying some sort of lock-in technique, but FOSS takes away their biggest weapon, the weapon that left Microsoft atop the heap in control in various markets for so many years.

>> But it's not really feasible to put together a complete desktop application stack using 100% non-commercial software. OpenOffice doesn't qualify as strictly non-commercial given the behind-the-scenes support from IBM and Sun.

Between 100% and 5% there is a wide gulf. I would be happy to see 95% market share. I don't expect 100%. I think 95% of the software on a machine being FOSS would make life easier for many an end user.

In what way are commercial players cooperatively working with each other and with the community to a limited extent that similar to a closed source monopolist keeping control and pulling in most revenues (for the monopolist products)? You can mix FOSS and commercial support. You can make a living on FOSS, but it is a competitive living. With a monopoly market, you basically have an all or nothing proposition. That is why FOSS has increasing support from commercial players that were bumped out of their markets or that are really feeling the heat. Microsoft has simply gotten too big and stepped on too many toes. Those embracing FOSS early are in line for a bigger piece of the pie Microsoft will not be able to keep.

>> And in any event, the consultants a large organization will need to hire in order to implement and maintain this stuff aren't going to work for free.

Hence FOSS is business friendly. But what does this have to do with monopoly control? Not all forms of lock-in are the same. There is a reason monopolists are singled out in US laws.

The larger the percentage of the infrastructure and apps that are open and based on standards, the easier it will be to migrate away from an abusive relationship with that vendor in the future. There are many types of lock-in, not all of which are created equal.

>> OpenOffice and Debian can't offer the same system management tools, the same end user ease-of-use or the same guarantees of timely integration with third party hardware and software as the Windows desktop.

A lot of people do not agree with this opinion. For example, many unsophisticated end users consider Debian package/app installation to be superior in many ways to anything out there. As for management, many administrators prefer the tools and flexibility that *NIX and FOSS offer.

But putting this issue aside, clearly there is a way to add value on top of what is already available for free. Some will be willing to pay for that price and others will not. That choice, the ability to control and upgrade according to your needs, is much more difficult to get from

>> I recognize that Debian and Ubuntu and the other desktop distros serve the extremely valuable purpose of holding Microsoft's feet to the competitive fire.

I agree and one has to wonder for how long Microsoft can stand the still increasing pressure. For some reason they took it easy in the past. It's because it is ... easier. If Microsoft fails, what would the spell for businesses that are not hedged, perhaps significantly, with other vendors or noncommercial FOSS?

>> But for all of them the name of the game is to keep customers writing a never ending stream of checks in exchange for a never-ending stream of new product features, at least some of which must actually be useful. This software strategy is as old as the hills, and is practiced by every developer who is not operating as a pure charity.

This is why *competition* among suppliers is recognized to be a powerful tool for consumers to have. Take competition away and the "name of the game" becomes a lot more achievable for the player now in control.

>> economics teaches us that "near monopolists" like Microsoft, if they are rational, have an incentive to price their products just low enough to discourage entry from competitors. After all, they aren't blind to the fact that despite their dominant market share their customers still have choices. Since they have a huge installed base over which they can amortize any price cuts, they will segment their offerings into cheaper products for price-sensitive customers (consumers, governments, developing countries) and premium-featured products for price-insensitive customers (blue chip corporations, Hollywood stars, hedge fund managers). Then, on top of this rational price discrimination, the near-monopolist will throw in generous volume discounts for really big buyers, because that's the efficient thing to do. And this approach is pretty much exactly what you see with Vista with its multiple editions and price points, and even with Windows Server.

Some actually believe that, in light of the problems with monopolies in general and considering the particular abuses in which Microsoft has engaged, what you state is proof that the government needs to get tough with Microsoft.

But let's look at this a little closer.

How many FOSS groups do you know that have been discouraged from offering their FOSS products or engaging the market in some other way on account of Microsoft lowering their prices? Is the theory wrong? Is Microsoft failing to find a price point they can live with? Is Microsoft irrational? Some might answer "yes" to all three. Clearly though something is missing from this simplified picture when FOSS continues to make gains on Microsoft's territory. In many cases, that gains don't even show up on some radars because you have consumers that are trying out FOSS while still using the Microsoft product. Perhaps we simply haven't gotten to the equilibrium price point yet.

As far as Microsoft expanding their offerings, this is one way they can fight the pressure. It's what they have to do to meet the same goals they could more easily meat in the past. The question is how long can they manage this so as to continue to grow or even hold their revenues? Will their installed base continue to be huge? The simple fact is that what you describe is not a recipe for success. They are tools to manage competition, but there is no guarantee that the pressure will not be too great.

As for economic theories in general, be careful about what you apply to the software market. FOSS continues to make a mockery of those that apply old rules to a system where distribution costs are near zero and development costs are distributed widely.

>> Bottom line, Microsoft is going to close any long-term price gap that emerges between Windows and Linux for as many customers as it can -- not for all, certainly, but probably for the majority.

I don't share your confidence in this conclusion. I would say that the numbers for which Microsoft can close price gaps will continue to diminish as FOSS continues to improve at a quicker pace than does Microsoft products and as Microsoft exhausts their ability to remove "fat" from pricing and operations.

>> So even though I'm still persuaded that switching to desktop Linux is pretty much a waste of money for big legacy Windows organizations

I suppose you won't be surprised that I disagree with this conclusion as a generality. Still, hedging a little is probably better than hedging nothing at all.

[I wanted to post now but I have to run. When I get back, I'll fix remaining mistakes I notice.]

I have to agree with you that Microsoft just needs to coast this side of some imaginary line that keeps people from migrating. They seem to have done quite well till now, the Linux desktop has been a good replacement for 95% of what any office worker is doing for a long time now.

I am always amazed that CIOs and IT Managers in general have not been hauled over the coals for the situation they are currently in. The city of Munich migration is one, that cost is not actually the cost of migration to Linux. It should rather be seen as the expense of unbundling from lock-in. They have been locked in good and proper by their own decisions and by their vendor. The fact that no other department in any organisation is likely to have allowed or been allowed to continue with such an unbalanced situation says much. How can people who have setup processes that bleed money be allowed to continue in their jobs? Perhaps its time for shareholders to pay more attention to the decisions made in IT that have cost them a packet over the years.

It is interesting that you say everyone has benefited from those who migrate. I agree. In fact anyone using IE7 needs to thank Firefox that they're not still using IE6. Clearly encouraging healthy competition encourages innovation that you benefit from.

I would go further to say the any CIO who does not have some Open Source in the works and in production doesn't deserve their job. Why? On the simplest level although they benefit from the price pressure the general Open Source adoption outside of their company they still do not have a strong negotiating position. If they say they doing something it gets better, and if they actually have deployed solutions then they're in the strongest position.

What they will of course discover is that they built some internal skills in Open Source and are in a better position to decide which way to go. In fact anyone looking at a proprietary solution who doesn't pilot OSS needs to be examined again. If you do OSS you get to understand the domain so when the sales people arrive you actually know what you want and can compare their offering against what you've learnt. That way you by proprietary knowing it has what you need, not because the sales people convinced you. And you choose OSS because you know it does what you want and is equal to or better then the proprietary offering.

One aspect that is clearly missing in your economic debate is the issue of macroeconimics. It might be worth considering that government don't like being on the end of a monopoly that they don't control. That this is money that bleeds out that they are quite willing to spend locally. Even with no cost saving a country benefits by creating local skills, new businesses, competition, etc. These are worth quite a bit. In South Africa we spend about R3billion $600million or Microsoft every year. That is a lot of money that could be channelled into local skills creation.

As the previous poster stated its the different type of lockin. I'd prefer to be locked into Canonical knowing that if they really mess up I can go to Novell. Or better yet I spread my bets at a national level between vendors. And my choice in the current climate? No choice, I can choose the colour of the supplier but they're all just Microsoft solution providers, they'll package it like I want, but they have not internal high level skills. But know I'm locked people that I choose, that I can move from and that compete with other providers. Long term prices are going down down down. Or at least I get exactly what I pay for.

The only way to escape this "customer pays forever" model is to use software that will literally never have to change or be touched again.

Untrue. You could also use Free Software, which is maintained and is free to upgrade. That's what my small business does; I estimate we've saved something like $100,000 in software licensing fees in the last 5 years.

Why don't you write stories on the successful migration from Windows to Linux...I am sure they are out there. If I remembered correctly, the French parliament were supposed to migrate by June of 2007 and I believe that the City of Largo in Florida and a host of Libraries in the US did the same.......

My first point in response to your article is that the deployment cost of open source desktops can be lowered just as it has been with MS products. This is just a software issue. As distributions gain parity in terms of user facing functionality there is no reason to suppose that open source programmers will not see the deployment problems and produce solutions for them. It would be an interesting task to produce open standards for desktop deployment and then open source code to do the job. So as Microsoft tries to walk the pricing tightrope open source can continue to reduce costs and make the task that much more difficult for MS. MS does not entirely control the cost of entry for its competitors.

Next, I believe you are overstating the amount of lock in inherent in open source. It is true that without programmers you can't fork your vendor's products but with them you can if the vendor becomes too obnoxious.

I guess I really have three points. The last one is that I expect the competition from open source to be part of the value delivered by open source. MS tends to be lazy without competition as do most monopolists. MS can produce some decent stuff when the heat is turned up.

I'm assuming this article should be labeled "satire". If not, then Mr. Gould may very well be one of the most ill-informed bloggers on the net. His bio states that he's a CEO... now listen to me people.. shhh... if you work for him, run... run for the hills...

I definitely had more "bugs" to get out of the comment I wrote earlier, but since most of it is probably understandable, I won't bother to fix it (not getting paid you know.. and writing comments, while a habit of late, is not a passion). In this second comment, I want to give more attention to a few items I mostly ignored before.

I think you may have missed the big picture if you think that the existence and sustainability of quality FOSS products depends on a few commercial companies sacrificing contributions to the community for the community's sake and in order to keep the community alive and competitive. The depth of FOSS is that most of the work, with a lot of it invisible, is done by people that are not getting paid to do that specifically or who would nevertheless do a significant portion of what they do were they to stop getting paid. There is a lot of depth. The noncommercial community, at a minimum, is likely to tip the balance from one commercial competitor to another. And I am talking about FOSS commercial entities, so that even when there are proprietary add-ons or support contracts, there is usually a solid fall-back that is 100% open.

Why do some get paid today to write code? Because companies have found that paying for some FOSS here or there brings back dividends. Maybe at any one time there is a limited capacity of paid work-on-FOSS-only developers, but it does exist and whatever business rises to the challenge stands to benefit. Would Red Hat be were it is today if it was nothing but a bunch of managers, lawyers, accountants, etc, and no paid developers? Don't top closed source companies spend loads of money on projects that never make it out of the labs? They do it because all it takes is one winning formula, multiplied by a very large volume of customer contracts, to make up for the R&D costs. Or at least that is the case most of the time [companies do fall from grace now and then]. Anyway, the view that commercial interests will stop spending on R&D or stop putting those R&D dollars into FOSS is extremely unrealistic so long as FOSS gives them such a high leverage. Even if Red Hat goofed for a while and make bad decisions, they would still have a growing and improving body of code to use [Don't get me wrong; they can lose their high flying status, but they would not be blown out of the water without an awful lot of goofing off].

I mention this to get to the real point which is that it is the end user that matters. Free and (let's assume simply) good enough really is good enough for most users. Certainly it is a good backup to have. Here I am looking at end users that don't appreciate source code. Let's assume that no one can make a dime tinkering with source code. It's a lot to assume, as it goes completely against reality, but let's consider that scenario. Well, while the fun lasted, many end users would be getting much better value and companies like Microsoft would take a financial hit (ie, their spare cookies would be flying off the accounting shelves). But then wouldn't the fun end and the software would start to get moldy and stale? Not if a significant number of developers and end users would continue to value that deal, of getting a bunch of software for free and having a say in what it becomes. That is the case, I believe. It is how against bigger obstacles that exist today, FOSS made it as far as it made it, even in a time when few could make a living exclusively coding FOSS. Microsoft cannot outlast an ecosystem that runs on a currency that is generated on the spot by whomever is willing to put in the time. We are not talking about dollars. Microsoft does run on dollars though, and they can't continue forever to raise prices over there to give a discount over here.

A fuller coverage of the business of FOSS (not something I am about to do here or soon) than what was considered above would show that many can make the same sort of living they do in the Windows ecosystem than they do in the Linux ecosystem.. at least once they adjust their knowledge. The only threatened species here is the monopolist or monopolist wannabe that lives off license fees for closed source software. Closed source software is just a bad deal most of the time. That you have to pay for it is rather insulting unless it really stands out. When Microsoft loses enough market share to lose their monopoly hold, the slide will be painful to watch, as they seem very unable to adapt to a FOSS business plan. Maybe their execs will roll up their sleeves to form the volunteer community that will slow down Microsoft's fall, as I don't know many that would make that sacrifice.

You spoke of lock-in by FOSS companies. Well, a simple type of lock-in is that there is a cost simply to changing vendors. There are relationships, trust, habits, very particular details in many processes (contracts, glue source code, various human and nonhuman protocols, etc), etc that make established relationships a bit costly to get out of. The more such details involved, the greater the lock-in, but this is very different from the added costs resulting from the closeness of a monopolist's software products.

Reverse engineering has brought the FOSS world fairly close to Microsoft currently (a window of opportunity), but Microsoft's new products are well on their way to raising the bar to escape once again. Those continuing with Microsoft's new software are voluntarily making any future decision to migrate away considerably more painful than it would be now, at least until FOSS once again manages to reverse engineer enough of the Microsoft locks.

No matter how much a company is stuck with Microsoft, unless they want to risk everything, they should start creating a Plan B that includes FOSS. FOSS, for example, is what allows a small business to survive while business is slow. During bad years, when you have extra time time on your hands, it helps tremendously to be able to trade some convenience for zero licensing fees. You don't have that option if Microsoft owns the entire tightly closed platform that runs every facet of your operations. During those bad times, you will be bleeding to death with unavoidable expenses, or you will then have to make a very costly one-time rushed migration away from Microsoft (to pay your debt for past sins, if you will). With Linux, you can adjust how much support you pay for vs. how much of the free offerings you take at face value.

I think some people really do do a better job under pressure. Some would prefer to wait for a very difficult moment in order to rid themselves of the chains that currently have them tied down.

You have missed it. Yes, you still pay for software, but with Open software, the lock in is gone.

You still pay, BUT you choose who you pay.

You control your own destiny. If you want to upgrade to the new version, you can, if you are better served fixing the old one, since you have full source, you can.

Tell me how to handle an application, to which you have no source, but your business depends on it, when the vendor went out of business, and the application won't run on anything newer than Windows 98.

Now what? You can try to find another vendor, buy all new hardware, and retrain all of your people.

If it had been an open source application, you could weigh whether it is more economical to get a new application because the old one is unsupported, or pay someone to support it.

You still pay, but you are in control.

If you cannot modify the code, and you cannot control what the machine does (Vista reduced functionality mode, Windows Update updating things you told it not to, etc.) is it really your computer, or are you just renting it?