Outsourcing open source? Bad idea

I'm on the phone with a WebEx technical support person. Not sure why tech support is necessary for taking my money (I'm just trying to set up a WebEx account for Alfresco - signed up online, gave my credit card number, and have received nothing in return), but whatever.

The thing that frustrates me to no end is that I went from talking with a local, US salesperson to a remote, India-based support person who gave me infinitely less support. He didn't understand the urgency of my requirements, though he tried hard to care. I eventually hung up in desperation and called back to the US sales team, telling them they were not allowed to transfer me outside their building. Result? Successful account setup.

Which got me thinking. I'm not a big fan of outsourcing in any situation - as I feel that support and development is best done as close to the problem one is trying to solve as possible - but in the open source world it seems insane.

Two big reasons:

1. Development. In the open source world, source of code matters as much or more than the source code itself. It matters that JBoss employs the vast majority of developers that contribute to it. Ditto for MySQL, SugarCRM, JasperSoft, and others, including Red Hat.
   
   In the open source world, enterprise buyers want to buy from the source of the source code - diluting this through outsourcing is a bad idea. (Yes, as Thomas Friedman writes, some outsourcing is a foregone conclusion. But not all. And in open source, I think Friedman's rules are less applicable. Open, it matters who writes the code in open source, and where they sit. Coders matter in open source.)
   
   
2. Support. In many companies, support is not considered core. (Clayton Christensen talks insightfully and humorously about companies continuously outsourcing "non-core" functions until they have nothing left.) This is foolish at the best of times, but this thinking in open source is suicidal. At its foundation, whatever the business model around it, open source is always about superior support. There are two cores to any open source business: excellent code and superior service. Outsourcing one or the other is sheer stupidity.

If you're an open source company, you need to keep development and support close to home. If not, we'll be outsourcing your jobs next.

Open source VC: Now a good time to invest?

LinuxWorld Magazine (written by Paul Sterne and Nicholas Herring) has an interesting piece on open source venture capital investments. The report focuses on the question: is now the right time to invest in open source? The answer to some is "Of course!!!" But the data might suggest we're in an inflationary period, with smart money waiting for valuations to come down.

From the article:

Starting in the mid '90s, a few brave pioneers like Benchmark invested in an open alternative to proprietary software and made a fortune. By the end of the decade, everyone wanted a piece of the action. A second wave of VCs rushed in at ridiculous valuations and got their clocks cleaned. In 1999 and 2000, over-capitalized, over-valued open source companies burnt through hundreds of millions of dollars. Shame on the dumb money that gives efficient markets a bad name. Then, the bubble burst, and we entered nuclear winter followed by the trauma of 9/11. Valuations plummeted to the ground. Investors panicked and ran for the exits. Things hit bottom when the vulture capitalists waited until the CEO of SuSE Linux was in a cab on the way to file bankruptcy before agreeing to invest at a valuation set at 1% of SuSE's peak value in 2000.

In retrospect, the SuSE recap was the turning point.

In 2002 and 2003, more smart money like Benchmark, Accel, and AdAstra fueled the open source industry with much needed capital. With the sale of SuSE Linux to Novell in November 2003, the industry reached a new inflection point. Valued at 50% of its peak, the SuSE exit excited a new rush to invest in open source in 2004 (no one wanted to miss the boat). After a small dip in relative value in 2005, when Novell dropped the open source ball, venture capital investment is on the rise and valuations are healthy.

In short, venture investing is a roller coaster. Whether you "lose your shirt" or make a mint largely depends on when you enter and exit the market (and not merely when you enter and exit a specific investment).

For example, take a look at the chart below. It shows the stock prices of Novell and Red Hat over the past 9 years. As an investor in either, you could be bankrupt or filthy rich depending on when you joined the party.



Funny enough, I helped to build that "smart money" period in Novell's history, and made nothing from it. Matt = stupid.

So, is this a good time to invest? It's certainly a great time to be in the business of open source. My decision to go to Alfresco has been the best career move I have ever made, and I suspect those at MySQL, Red Hat, JBoss, SugarCRM, etc. would concur. The problem for VCs, however, is that valuations are on the rise, and the amount of money in the market has moved well past the $1.3 billion I blogged earlier this year, making it harder and harder to make money, even on good exits.

Btw, if you want to know where to get the smart venture money, look to experience:

Way off-topic: May 17, Paris

And I wasn't even wearing my lucky shirt. Bienvenue a Barcelona a Paris le 17 mai pour le final de Champion's League.

Tranlsation: Arsenal just won. Going to the finals for the first time ever in its 100-year history.

Competing with closed, competing with open (Hint: Open up!)

I've been watching the ruminations surrounding Red Hat's acquisition of JBoss with lively interest. Most of the past 5-10 years have been spent wondering how to beat Microsoft. Now, it seems, the struggle is to beat Red Hat...

...and it hasn't even won yet. :-)

It is testament to Red Hat's success that the Financial Times reports that Oracle has considered buying Novell. Stephe rightly points out that Oracle is not a middleware provider, per se, and so should not feel overly threatened by Red Hat's acquisition. Yet it clearly does.

Why? Because there's nothing more terrifying to a closed-source vendor, no matter where it sits in the software stack, than an aggressive, hungry, and successful open source player. Success with JBoss today should lead Red Hat to open up other areas of the stack tomorrow (or, I should say, tightly knit together other open components of the software stack into its own certified offering). Red Hat is not an applications company today, but why not tomorrow?

This is why Oracle, Novell, and others need to quickly learn how to compete in an open world. As I blogged recently, hybrid models are a short-term salve, but they don't effectively beat either closed or open source competitors.

So, it's great that Novell has announced an open source collaboration suite. However, to compete effectively against Microsoft's own Sharepoint + Exchange/Outlook + Office/etc. suite, Novell needs to engage in asymmetric competition: disrupt its proprietary competition by offering a truly open product. Novell's "open" collaboration suite should be...open. Of course, it's easier to say "open" than to actually open up GroupWise/etc. after years of their being closed. But third-party components tied into the suite should be open source, or there's simply no compelling reason to choose Novell over proprietary (or open source) competitors.

The software market will not be won in a day. Companies that take a long-term, open source approach to winning will dominate the next few decades of software. To Oracle, Novell, and others who want to put together compelling "stack" alternatives to Microsoft et al., I have one word:

Open.

If you don't, it's only a matter of time before the open source ecosystem combines in various guises to crush you. Today it's just Red Hat + JBoss. Tomorrow it's Red Hat + JBoss + MySQL, or MySQL + Google, or Alfresco + Compiere, or...you get the picture. The future is open, and your business is closed unless you are, too.

Open source and the rise of asymmetric competition

I'm putting together an Executive Radar event with Tim O'Reilly for this year's OSCON in Portland, Oregon. One of the sessions actually has me somewhat sleepless (literally - I'm typing this at 5:17 AM...):

Open Source and the Future of Asymmetric Competition
For years the software industry has largely competed on the basis of symmetry: Oracle versus IBM in databases; BEA versus IBM in application servers; etc. Feature wars, price wars, but not true competition wars. That is, competing by playing a different game, with different rules. Open source enables an alternative battleground upon which to compete, with community, code, and culture the new competitive tools. This session brings together the top open source executives deploying these tactics of asymmetric competition, to learn from their experience.

The topic has me excited because it promises to help unlock some of the secrets of Red Hat's, JBoss', and MySQL's success.

A few hints in advance (based on my own experience "selling free"):

1. To succeed with open source one must truly embrace open source. In the past, I have advanced a 'both source' strategy wherein a vendor provides both open and closed code to take on proprietary software vendors. I think this was a mistake. It may have been the right initial move for the company I was with at the time, because it was the most dramatic open source move possible at the time (for cultural and business model reasons), but it is not a winning strategy over the long run. Open source must be open to disrupt a market. Hybrid models are a decent short-term fix, but they don't provide long-term advantage. (Note: Putting our money (and code) where our mouths are, Alfresco is moving to a 100% open source model. So yes, I'm walking the walk, not just talking a good game.)
   
   
2. Open source is a marathon, not a sprint, and requires marathon-minded investors and management. The best companies (Red Hat, Google, and others) recognize that a software empire - a Rome - is not built in a quarter. Open source requires patience - the best open source companies take years to build properly, because they cannot subsist on big, up-front license fees. The subscription business model slowly strangles competitors - they might spend months or years laughing, but eventually they suffocate, asphyxiated by their own lame, top-heavy business models. But it doesn't happen overnight.
   
   
3. Open source companies should aggressively exploit their open advantage. It used to nettle me to no end to hear Matthew Szulik (CEO, Red Hat) decry "open source pretenders" and hammer on the point, again and again and again, that source code matters. I hated the message because it was winning against me in the market. Red Hat, more than any other company, has used source code - its quality and its openness - as a wedge to drive into enterprise buyers. Free to try (lower the bar to tasting the Red Hat product), with an open model (pay for superior support and software certification, not massive license fees) that makes it easy to use an excellent product. Red Hat has - just as JBoss, MySQL, Alfresco, and other do - made a fetish of openness: open roadmap, open pricing, open code. The proof of its success is in its income statement.

These are just a few of the lessons I've learned from my years in open source (starting in 1998). I'm looking forward to hearing what Michael Tiemann (Red Hat), Marten Mickos (MySQL), and others say in this OSCON session. Hope to see you there.

Four books diverged in a wood....

While in Costa Rica, I had lots of time for reading. I plowed through Thomas Hardy's Return of the Native, Charles Dickens' A Tale of Two Cities, Willa Cather's My Antonia, and (in desperation because I still had four more hours of flight time and my kids were using my laptop to watch Batman cartoons) re-read J.K. Rowling's Harry Potter and the Half-Blood Prince. Four very different books, with some random ideas emerging from them....

1. There's much more to life than software. I could be very happy with no computer and with lots of Charles Dickens.
   
   
2. We spend a lot of time coveting others' things simply because others have them. This is the tragedy of Native, where a man and woman take turns casting each other aside in the hopes of latching onto a better mate, only to have everything end in calamitous tragedy in the end. In the software world, the best companies are not those that ape the successes of others, or seek to one-up the competition. They are those who completely reshape or re-create industries. This is why Microsoft struggles to compete with Red Hat and Google - in different ways, they significantly alter the rules of engagement. This is what open source promises for many vendors, and yet we spend far too much time competing on features. We don't, in short, engage in assymmetric competition. (This is actually a session I'm helping to put together at OSCON this year - should be cool.)
   
   
3. Open source is sometimes governed by a mob mentality. This was the lesson from Two Cities. In it Charles Darnay foolishly thinks he can abandon his noble heritage in the face of a rabid mob. Madame Defarge sends him to la guillotine, refusing to believe that anything good can come from noble blood (except, well, blood). In the open source world, the free sourcerors are akin to the French revolutionaries - demanding an extreme breed of unthinking patriotism that brooks no opposition and allows for no reasonable debate. In this fixation on free (as in liberty), the free source movement tends to overly focus on software freedom at the expense of user of software freedom, it seems to me. BSD is not "free" because it allows people to freely decide what they'll do with software, whereas GPL-like licenses are free because...they force people to use software in a certain way. Don't get me wrong - I actually prefer the GPL precisely because of this control it allows me to exercise over my code and all that use it. But this is because I'm a shameless capitalist, and not because I'm a freedom-loving code fascist. :-) We need less ideology in open source, and more pragmatism.
   
   
4. Harry Potter teaches me that the best code (authorship) doesn't always win. Often, it's just the group that markets the best story. This is as true in open source as it is in the proprietary world. JK Rowling is not a great author - she doesn't hold a candle to Dickens, Hardy, Dostoevsky, etc. But I've read each of her books 3-4 times for a reason - they're very entertaining (and my kids love them, so I've read each book to my older kids 1-2 times). In like manner, my software choices - and yours - are not usually governed by the best code, but rather by brands, convenience, cost, etc.
   
   
5. From Antonia I was reminded of a similar theme - the greatest community and nation-builders have tended to be immigrants. Cather captures the influence Czech, German, Norweigan, etc. settlers had on the United States' plains, working the ground despite exhaustion until it yielded prosperity. In similar manner, it has been developers and companies from outside the United States that have created the world's most significant open source projects. Thoughts as to why?

There are many more themes that struck me as I read, but I'll leave it at these for now. (In the meantime, I've ordered more Dickens so I can re-read all of his novels, more Hardy, and a range of others. Watch for more Asay moralizing in the near future.... :-)

Religion and open source

Hunkered down trying to meet and exceed my quarter at Alfresco, I had forgotten that some people still treat open source as religion. I'm used to dealing with CIOs who appreciate open source for its high quality code, for the communities that undergird it, and for the low cost.

While in Costa Rica this past week, however, news reports started trickling in from the FISL conference down in Brazil. Because I'm now on the OSI board (effective April 1), I also see a wide range of emails from those who see open source in religious hues.

I don't understand it.

I like software. But I like it in the way I like my house, my car, etc. It's an inanimate object that provides utlity for me. That utility is not something or someone to worship - I go to church for that. Rather, the utility is something that makes my job easier, saves me time, etc. Software is a productivity tool for me.

It's therefore funny to me to watch battles rage over whether BSD offers more freedom than the GPL (of course it does - freedom to do what the user wants, not what I, the copyright owner, wants, just as liberal democracy affords people to speak, exercise religion, assemble, etc. according to what they prefer, not what the government prefers (within bounds)), and to see so many wringing their hands over open source licensing. Fixating on licenses leaves the vast majority of open source's power untapped.

Open source is powerful because open source encourages communities to develop around a given software project. Open source is powerful because it offers a way for customers to benefit from code as it is and extend it as necessary to what it needs to be (yes, enterprises make short-term code changes all the time). Open source licensing is a means to these ends, but it is not the end in and of itself.

Off-topic: Costa Rica is amazing

My wife and I have been to Hawaii several times on vacation, but we have a new "warm weather" vacation spot: Costa Rica. Hawaii is the vanilla of American vacation spots: safe and easy (though by no means inexpensive). Costa Rica is what I think Hawaii should be: tropical (we just got back from touring through a rain forest, complete with poison frogs, toucans, etc.), great beaches, beautiful people, and enough spoken English so a gringo like me can manage. (I speak French fluently, so I'm not a complete American, but French hasn't yet helped me on our vacations....)

Anyway, this blog isn't meant to be a travel-log. But if you're looking for a good vacation spot, I'd highly recommend Costa Rica. The food is fantastic (like the tres leches show here - I've gained 50 pounds on it since we arrived on Friday). The weather is impeccable. The roads leave something to be desired, but that's actually part of the charm. In Hawaii, I find myself getting bored because there's nothing to do. Here, I have no shortage of crocodile watching, rain forest hiking, etc. to keep me busy.

One more thing: it's much, much cheaper than Hawaii.

Infrastructure: You Get What You Pay For

A few months ago I wrote from Caracas, Venezuela on the importance of infrastructure. I'm on vacation in Costa Rica now with my family, and the same lesson is hitting home.

In software, we're largely coming to take infrastructure for granted, thanks to exceptional middleware from JBoss, databases from PosgreSQL and MySQL, etc. We think, in other words, that 280 (California), I-80 (CA to UT and beyond), I-95 (Massachusetts), M25 (London), and other roads just happen. They're free, like much of our best software. Infrastructure is FREE!!!!

Except that it's not. I drove on pseudo-free infrastructure Sunday as my family and I headed to church in Puntarenas, Costa Rica. The drive is roughly 60 kilometers, and took us 1.5 hours. Why did it take so long? Because try as I might, I couldn't jump our little Toyota over the massive potholes (often spanning the entire highway) safely, and so had to slow down to negotiate the craters in the road.

This pseudo-free infrastructure would start to look a lot more like the "free" infrastructure that Americans and Europeans take for granted if more of us were paying taxes in Costa Rica. Because, of course, our infrastructure is not free. We pay for it. Heavily.

For those who can afford to pay for great software infrastructure (like JBoss and MySQL), yet opt to free-ride, over time these decisions will be revealed in the quality of that "free" infrastructure we "drive" our applications on. Big potholes in which a minor planet could rest comfortably.

Whatever the open source project you use - from Mambo to Handbrake to Adium to Tomcat - keep in mind that someone, somewhere has to write that code, and however much they may like writing software for free, at some point they need to get paid. It might as well be by you. And me. After all, we're the ones driving on it.

Just when you thought life couldn't get any worse...

...someone buys you a "PoshAir." Fortunately, I've never seen anyone wearing one of these bodybags on a flight, but apparently people do buy them. Like this happy PoshMan.

Fabrizio: Our next bet is going to involve this ridiculous bodybag. Whomever loses will have to wear this suit on their next 10 flights. With the hood on.

Think I'm joking about this outfit? Here's what MSNBC had to say about it:

"We're sure the poshAir will be the envy of fellow passengers trying to catch some shut-eye with those flimsy/tiny airline-issued blankets..."

MSNBC Business Editor Brian Tracey

Sorry, Brian. I don't think anyone entombed in a PoshAir is going to attract much envy. Try again.

The costs of implementing open source

The cost of implementing open source

At last week's LinuxWorld Boston I moderated a panel entitled "The Real World of Open Source Application Implementations: Case Studies from the Front Line." Contrary to what Fabrizio says, it is precisely because of sessions like these that Linuxworld (the conference) and OSBC still matter. Everyone knows Linux, true, but not everyone knows what it (and other open source technologies) actually costs, how and where best to implement it, which business models have been successful in the market, etc. What is needed, rather, is more candid information exchange, not less. OSBC is an unparalleled place to house that sort of exchange.

But I digress....

In my session, we brought together two SIs (CorraTech and Optaros) and three IT buyers (Putnam Investments, Athena Health, and The Christian Science Monitor) to figure out if open source is actually the bargain that people like me regularly spout off about. (Btw, Seth has a good summary of the session.)

I asked the IT buyers how they find new technology they buy, open source or proprietary. Answer?

1. Talking with competitors/industry peers.
   
2. Web (News sites, Google, etc.)
   
3. Venture capitalists

The first and last sources surprised me a little. Geoffrey Moore talks about how references provide the "bridge" to get buyers over the chasm. I get it. But I was surprised by how pervasive inter-company communication is, based on the responses of the panelists.

It's clear that few want to be any technology's guinea pig. Hence, it becomes important for open source technologies, in particular, to establish visible customer credibility: case studies, press releases, Sourceforge download statistics, media mentions, etc.

As for venture capitalists, it makes sense that they'd be a trusted source: if they're willing to put down a few million on a company, surely they must believe in it. VCs also act as connectors to link together different IT buyers.

In the course of the session, I also asked both SIs and IT buyers whether open source was cheaper to implement, on the theory that source code access might facilitate development and integration. The answer? The cost of any implementation (at least, in CRM and ECM) is always 2-4x the license cost. The real value in open source is in how you spend the license cost savings: extra customization? More IT staff? Training? Documentation? etc. You choose.

Ultimately, then, you can choose whether you want to save money, or whether you'd like to invest those cost savings in achieving a tailored solution that is simply not possible in a proprietary world (at least, not without paying a massive premium).

Larry Augustin illustrated this well in his OSBC2004 presentation:




Where do you want to spend your money? You choose. Open source = more choice.

France: Many rights, few jobs

France's problem with basic economics

Not sure if you've noticed, but the French are outraged (again). When I lived in France, it was the farmers and truck drivers constantly blockading cities with leeks, potatoes, and, well, trucks.

This time? It's the youth.

What are they protesting? Work.

French President Jacques Chirac (or, rather, Prime Minster Villepin) last year introduced the CPE, which basically makes it easier for companies to fire younger workers. After the youth rampaged, the government capitulated and introduced the CNE, which allows for companies with fewer than 20 employees to fire employees without reason (during a 2-year trial period).

Sounds heartless, right? Letting big companies fire little workers?

The problem with the reasoning is that by not allowing companies to fire employees, France essentially makes it impossible to hire them. Expanding into France is a big risk because the costs of getting out (should anything go wrong) are too high. Why take that kind of risk when other, equal options are readily available with none of the same risk?

Employment is not a right. It's a privilege. To earn good work, you take a calculated risk: education, training, networking, etc. Perhaps the risk is wasted, and you don't get a good job. But the risk is worth the return when you do.

Markets don't work when they're heavily controlled, as in France. France's unemployment woes are caused and exacerbated by foolish state policies like over-protection of workers. French youth, trying to help themselves by desperately avoiding the possibility of failure, are guaranteeing the very failure (unemployment) they fear.

What does this have to do with open source? Nothing. I just get annoyed by inanity.

Open source has reached the "Laugh at you" phase

First they ignore you, then they laugh at you, then they fight you, then you win.

Mohandas Gandhi

Linux is in the "win" phase. Apache webserver is in the "won" phase. MySQL and JBoss are in the stage where the laughter turns bitter and the pushing starts. Big wins are on the horizon.

Open source applications? We're at the point the ignorance is breeding laughter. SugarCRM, Alfresco, JasperSoft, Plone, Compiere, etc. These are all applications that used to be ignored, but ignorance is no longer serving proprietary competitors well.

As a case in point, InformationWeek just ran a story on Boise Cascade's use of Alfresco for invoice management. Big customer, big need, big value.

Documentum's response? Completely off-base, ill-founded commentary ("Boise's need to cobble together links between Alfresco and MySQL is one reason some parties turn to commercial document management systems") that open source solutions require development, and off-the-shelf proprietary software requires none. Not only is this not true, but it also cleverly hides a ball that every IT buyer already knows: EVERY ERP/CRM/ECM solution requires customization/development.

Documentum not require heavy customization to make it even remotely relevant for a company? Of course it does. The difference is that the entry-level cost for one of these proprietary systems is in the six figure range (server costs), which only is exacerbated by per user costs.

Run SAP or Siebel out of the box? Of course you won't. The fact that they have more features (in some cases) does not mean that any buyer immediately, out-of-the-box benefits from those features. Most proprietary systems in the ERP/ECM/CRM/etc. worlds are cobbled together "suites" from years of acquisition. Much of the integration work on any given IT project is likely due to getting their own products to work together, nevermind getting them to work within the business processes and infrastructure of the customer.

Let's not try to obscure the issue for IT buyers. Customization is standard. The question is how much a buyer needs to pay for a vendor to deliver the 20-80% of a products features they won't actually use. Open source (well done) offers a granular way to tailor software to an enterprise's needs. Do creative, development-minded enterprises benefit more from open source today than more passive consumers of technology? Probably.

But that is a momentary blip, if it's a blip at all. Laugh while you can. Better get ready to fight. Because we're planning to win.

Red Hat's JiHAT

Red Hat's got a jihad going on. (JIhad, of course, does not denote "armed struggle," but rather a striving. In this case, Red Hat has just taken a major move up the software stack, with the threat reverberating from Redmond to Somers (IBM). Importantly for the rising stack business, as well, Red Hat has clearly decided to go it alone in building and deploying enterprise-grade software stacks.

Despite my earlier analysis on why Novell should acquire JBoss, it's clear that the synergy between the top Linux distributor and the top open source application server company is clear. Integration between the two companies will not be seamless, but it's arguably a better fit than with Novell:

And no man putteth new wine into old bottles; else the new wine will burst the bottles, and be spilled, and the bottles shall perish.

But new wine must be put into new bottles; and both are preserved. Luke 5:37-38

I witnessed new wine (SUSE) being poured into old bottles (Novell) - it finally worked, but it was a difficult transition (cultural, business model, license model, etc.). Red Hat + JBoss doesn't have the same issue. Both use GPL licenses (GPL and LGPL). Both are used to service-based models.

What Red Hat adds to JBoss, however, is its ingenious licensing strategy, and both contribute to each other by feeding leads. Companies that use one open source technology are more likely to use another. In this way, the two companies mutually reinforce each other. And Red Hat offers JBoss additional teeth to its business model - a way to help users become paying customers.

This is a great step forward for open source, and for these two companies. Let the JiHAT begin!

The rise of RedBoss

Today, Red Hat announced that it will acquire JBoss, and RedBoss is born. (My name, not theirs.) I was obviously wrong on suggesting Novell was the suitor, and I'm not yet sure if they were right to consummate the deal (not that they asked me ;-).

The price? $350M upfront (40% cash, 60% stock) and another $70M for hitting performance targets. Not bad for Matthew, who just acquired a very different organization, whatever its business model.

It will be interesting to see how long (and how well) it takes for Red Hat to fully digest JBoss. I'm not a believer in acquisitions - I've never actually seen one work well - but Matthew is a strong CEO and could pull it off, if anyone can. That said, he's also acquired a company that is 100% different from how he runs Red Hat. About the only similarity is open source and two strong, vocal CEOs.

JBoss: The Next (Novell?) Chapter

JBoss is going to get bought.

It's no secret that Oracle wasn't the only one sniffing after JBoss. Red Hat has talked about buying JBoss (though it's hard to see Matthew and Marc getting along well :-), as has IBM.

But Novell is the best fit.

Disagree? You think JBoss + Novell = NoBoss? Think it doesn't make sense, or won't happen? I think you're wrong.

Why Novell? Let me count the ways...

1. Novell has lots of cash. Too much cash, comparatively. (See right.) Marc Fleury wants cash. (Who doesn't?) Perfect match.
   
   
2. Novell has struggled to convince enterprises to move from NetWare to Open Enterprise Server/SLES, though results have been improving under Ron Hovsepian. Still, the needs something else to attract and hook would-be customers. Ximian? Nah. There's no Linux desktop market, at least not now. JBoss could be the answer....
   
   JBoss has a large (and growing) user base. The biggest thing it lacks is a solid business model to monetize all its demand. The company did less than $20M last year, and will probably ink $50-60M this year, if things continue as they have. Where's the problem, you say? A lot more money could be made if the company moved away from a pure support model toward something like what MySQL, Red Hat, or others use. Something that makes it easier for JBoss to convince customers to pay for software they already love.
   
   
3. JBoss and Novell already have a strong, profitable partnership. Novell distributes JBoss' application server, having dumped its Extend app server. (This is exceptional irony, btw. Novell's SilverStream/Extend app server came from David Skok, who founded SilverStream, which Novell bought. David was the lead investor in JBoss' venture round...which company's product replaced SilverStream at Novell...and which product Novell is going to acquire into the fold. David wins every way he turns. :-)
   
   
4. Marc could use some toning down. Novell could use some livening up. (Having said this, Marc, you shouldn't overlook the fate of every other CEO Novell has acquired into the company. They don't last, and I'm not laying the blame at either Novell's or their feet. They just don't last, for whatever reason.)
   
   
5. Oh, and Novell has cash, and Marc would like to have cash. Did I mention that one? :-)
   
   
6. Lastly, it's unlikely that Marc Fleury would want to be the CEO of a publicly-traded company. I'm not saying he couldn't do it. I'm saying he wouldn't want to. Marc is a free spirit. The public markets tend not to like free spirits. Much better for JBoss to make Mark gobs of cash without the hassle that comes with public accountability.

Convinced? Better prepare for it, anyway. It's going to happen.

The real question is what a move like this would mean for JBoss' employees. Marc owns 50%+ of the company. I believe VCs (Matrix and Accel) own another 30% or so. That leaves very little for the employees, who likely won't see much from a deal. Will they stay? Do they have the incentive to do so?

(Note: In the Novell-Ximian deal, no one but 3-5 top executives made much money in the deal, and yet most people stayed with Novell. So perhaps this stock incentive-thing is overblown, and it will have no effect on JBoss employees staying through the acquisition.)

P.S. Why did Bob Bickel leave JBoss?

Hamiltonian open source

So, if you haven't heard (Dave talks about it here and here), the sky is falling and open source may be destroyed by its very success. Or so opines Michael Goulde in a recent Forrester report. Lest you waste money to fuel anxiety on the matter, let me put your minds at rest. (Dave has already written a good counter, and Michael Tiemann offered up this gem:

"I believe that the effect of open source on the proprietary vendors is a force 1,000 times more powerful than the force of proprietary principles on the open source community." From time to time we do see some charlatans claiming to be committed to open source when it is really the success of open source that they covet, and from time to time we see them fail and blame open source instead of themselves.

"But looking for the effect of proprietary software on the open source community is like trying to measure the gravitational effect of the earth on the orbit of the Sun." [From an OSI email to LinuxInsider.

Amen.

Open source is not a mask that one can don and pretend to be someone/thing else. Open source is a methodology for developing, distributing, and supporting software. If you sham it, you fail. If you fail, you go bankrupt. If you go bankrupt, no one cares to write about you anymore. Etc.

The risk in all this open source furor is that the idea behind open source will get tainted in people's minds, not that open source software, itself, will become ruined. If this happens, I suspect the good companies will keep chugging away, and eventually the cream will rise to the top (as happened in the dot-com fall-out).

In the meantime, let's not fall into a Jeffersonian trap of pining for the good old days of an agrarian economy - that idyllic pre-modern world of open source when everyone hugged each other and lovingly shared code. We don't need that. I much prefer Hamilton's vision: capitalism unleashed. In this case, open source capitalism unleashed.

Will we get chaff with the wheat? Of course. But the market has a funny way of obliterating dumb ideas and charlatans. The open source market, Hamilton style, will be no different.

The changing face of software sales

I had breakfast this morning with a sales director for a large, successful software company. He heads up several of the company's major accounts, and had fascinating insight into the software industry, generally.

Hint: It's being destroyed.

Well, "destroyed" in the Schumpeterian sense. Destroyed in the sense of a massive transfer of wealth from those who depend on upfront license fees to those that can thrive on much lower deal sizes that primarily consist of downstream maintenance revenues.

He mentioned that his particular branch of software (EAI) is undergoing commoditization, which surprised me: CRM, ECM, operating systems, etc. These areas are clearly being commoditized, but EAI? What could be causing this, as there aren't currently dominant open source contenders, and there hasn't been much standardization of EAI technology and product? He listed a few reasons (which mostly equate commoditization with a significant lowering of prices):

1. Vendors like IBM, giving away the software to drive hardware and services revenues.
   
   
2. Second-tier vendors trying to build market share by offering lower prices.
   
   
3. "It's plumbing - why should I pay for it?" (This sounds like an open source opportunity waiting to happen.)
   
   
4. Accenture and other large SIs driving down license costs in favor of higher services costs.

These were just a few of the factors he cited. Again, it was very intriguing to me given that EAI isn't one of those areas that I'd normally consider in the commoditization phase.

But he should know. Sales people feel pricing pressure first. He told me, in fact, that across the board, CIOs are rejecting large, enterprise-wide deals, preferring to pay "by the drink" instead. So I asked him, "What does this mean for sales people? How are they feeding their children, given the smaller deal sizes?"

The answer? Some are getting out of enterprise software sales altogether. Companies, for their parts, are shifting to compensate their sales forces based on maintenance deals (or, rather, weighting compensation toward the full value of the deal to get sales people to not discount maintenance to try to raise license revenues), rather than on upfront license fees. It's a rapidly changing world.

It's a world, frankly, that most software companies are ill-prepared to meet. They're either in denial (The market will turn around. Open source can't touch our market. Etc.), or they're in retreat. Very few (IBM perhaps among them) are actually meeting the threat head on with a hearty embrace. These few are succeeding. The rest are withering.

In short, regardless of your thoughts on open source, it's clear that the general fundamentals of open source apply to the changing world of software:

1. The future of sales is on the web. (I've yet to visit a single customer at Alfresco, yet sales are booming. The software must sell itself in the new software economy....)
   
   
2. The future of sales is technical. My director of solutions engineering (part SE, part architect, part sales guy) speaks the customer's language - his job is to round out the edges of Alfresco's code, to help the customer see how to implement it, etc. He's exceptional, and exceptionally productive for us. I wonder if the sales model of the future doesn't look like this: downloads -> inside sales -> sales engineer/architect -> sale. It's a lot cheaper, and a lot closer to what the customer actually wants.
   
   
3. Sales compensation must be tied to maintenance and support. Maybe this means the sales people must be as much about support as they are about driving upfront fees...perhaps more so? (See above.)

Food for thought. In the meantime, I'm waiting for the United States to allow me to donate to my favorite corporate charities.

Can I get a write-off for buying Microsoft Office? :-)

Where IT doesn't matter, use proprietary software

It's become common to think of open source as a commoditizing force - a great way to level an existing industry and capitalize on low education costs (i.e., the market category is well understood, making it possible to buy over the 'Net, etc.). But in an string of articles today from InfoWorld, the notion of open source as the good enough and cheap alternative to superior and expensive is dashed to pieces.

One article that I found particularly interesting is the case study on The Christian Science Monitor. I know Curtis Edge (CIO) and Terry Barbounis (CTO) pretty well, as they use my employer's software. But I didn't realize just how much open source was having an effect on how the organization hires:

"“I need to hire inventors, and I want to be partnered with groups that want to do the same thing. We need a culture that does that,"” Edge says. That'’s why Edge has pushed the use of open source technology at the church, including adopting the Alfresco content management system for the Monitor. He's also made changes in his IT staff to ensure that the church's in-house IT culture fits the open source culture. In the past 18 months, he'’s had 60 percent turnover in his IT staff as he changes the culture to a more flexible, entrepreneurial -- even creative -- approach to technology development.

If I'm an IT administrator or an architect at an enterprise, this is revelatory (pun intended). Instead of hiring people to be mindless drones that point-and-click Microsoft technology, Curtis is looking for bright, creative people to actually mold IT to his organization's advantage.

Maybe IT does matter, after all....

Further in the article, Curtis provides more insight into their reasons for using open source:

Although Edge is happy to use commercial software for established business practices, such as Oracle Financials for accounting, he does not want to be tied to a specific vendor's approach for an emerging, core technology like content management. "“When it comes to the Web, we're still sorting out best practices,"” he says.

Instead, Edge wants the broader collaboration fostered in an open source community. He feels that the open source methodology allows problems and opportunities alike to be identified quickly by people who can do something about them.

"“I like the open source community because anyone at any level can talk,"” Edge says. "I want my developers to talk with other developers, rather than with a help desk or a support organization. There'’s a lot of benefit from a grassroots community figuring out what's good for everyone."”

So, reading between the lines somewhat, for the technology that is well-understood and mainstream (meaning, it's a common need that isn't going to be much improved by dramatically tailoring it to one's organization), go with proprietary software. But for those areas where IT really does matter, use open source, because you can tailor it to your advantage.

In short, for IT that doesn't matter, use Microsoft. For that which does, use open source. :-)

Another way to say this? Open source software is as much about people as it is about software. With proprietary software, you get bits. With open source, you get people and bits. I'll take the latter any day over the former.