Ode to Tiger

William Blake is one of my favorite painters. He's also an excellent poet. And a prophet, of sorts: he saw Mac OS X Tiger long before Apple released it:

Tiger, tiger, burning bright
In the forests of the night,
What immortal hand or eye
Could frame thy fearful symmetry?

In what distant deeps or skies
Burnt the fire of thine eyes?
On what wings dare he aspire?
What the hand dare seize the fire?

And what shoulder and what art
Could twist the sinews of thy heart?
And when thy heart began to beat,
What dread hand and what dread feet?

What the hammer? what the chain?
In what furnace was thy brain?
What the anvil? What dread grasp
Dare its deadly terrors clasp?

When the stars threw down their spears,
And water'd heaven with their tears,
Did He smile His work to see?
Did He who made the lamb make thee?

Tiger, tiger, burning bright
In the forests of the night,
What immortal hand or eye
Dare frame thy fearful symmetry?

I've been playing with Tiger (rev 10.4 of OS X) since FedEx delivered it late yesterday afternoon. I had been tracking the package all day long, waiting for the FedEx guy to bring me something. (He laughed when he dropped it off: "Looks like I finally have something for you, and not your wife.") One hour later, I was in heaven.

It is, quite simply, the best desktop operating system on the planet. Speaking objectively, it is heads-and-shoulders above anything else out there (though the Linux desktop has momentum to move beyond it at some point, it's just not near parity now). Windows XP? No, it doesn't crash as much anymore, but it's just a lame operating system. OS X Tiger is beautiful. The widgets are simple yet powerful (what an amazing thought, to clutter a second "desktop" with dictionaries, train schedules, calendars, and all those other things we tend to go to a browser for, desipte it being 99% more software than we need for such tasks). Spotlight is impressive (though it has a hard time searching my Entourage mail and my OpenOffice files - I'm sure both will be remedied soon). And I haven't used several of the other new features.

This is what computing should be like. The things that should be mindless, are. (Everything is just where it should be.) The things that should require effort (video editing, etc.) do, but you're given heavy-duty, easy-to-use tools to accomplish these difficult tasks. And so on.

If I sound like I'm gushing, it's because I am. I'm so lucky that Dan Montierth (my then boss at Novell) pushed me to get a Mac instead of the back-ordered ThinkPad back in late 2002. The Mac gives Linux desktop developers something to shoot for: both founded in Unix/a Unix-like operating system. Both can be equally powerful and easy to use. Windows has the market share, yes, but that won't last forever. Monopoly never does.

Open Wallets for Open Source (NY Times Article)

Gary Rivlin attended OSBC, and was gracious enough to moderate our VC panel. From that experience and a fair amount of post-show due diligence, this article from today's New York Times emerged.

I was gratified to see that Gary got a few "contrarian" views. Not that these people (Mike Olson of Sleepycat, Danny Rimer of Index Ventures, etc.) are negative on open source, generally. Quite the opposite. Rather, they see, as I do, that a lot of companies are getting funded that don't necessarily deserve it. Wearing the open source label doesn't make you a good company - open source needs to be used intelligently to be a sound business strategy.

Ask Christof Wittig, CEO of db4objects. Christof is one of the most pragmatic, rational thinkers/practitioners in the open source business community today. Not the giddy dot-com type that is increasingly pervading open source. I'm just glad that there are others like him.

OSBC: Not your father's vendor show

So, for any still mistakenly believing that OSBC is a vendor show (ISVs/IHVs represented less than half of this year's attendees, down from OSBC2004), consider some data from the conference. (I've thrown in a few data points, as well.)

• OSBC2005 San Francisco drew a record number of attendees (700+);
  
  
• Over 75% of OSBC attendees were industry executives (Loosely defined as director-level and above - CxO-level attendees accounted for 34%+ of attendees; another 24% were VP-level and above);
  
  
• OSBC2005 San Francisco experienced a five-fold increase in IT executive attendance (~10% of the attendee base, and more senior on average than Linuxworld IT attendees), covering a wide range of the industry’s premier players;
  
  
• 78% of OSBC2005 San Francisco attendees are directly involved in IT spending within their organizations. Over half authorize/influence IT spend of $100,000 or more;
  
  
• This year’s IT executive group included SVPs and CIOs of Thomas Weisel Partners, JP Morgan, Fidelity, Bank of America, PayPal, Commonwealth of Massachusetts, Motorola, and a range of others. To put this in perspective, we expect that OSBC soon draw more SVP-level and CIO buyers than Linuxworld;
  
  
• Based on a post-conference survey, 96% of attendees were either "Satisfied" or "Very Satisfied" with the show's content;
  
  
• Media. OSBC2005 San Francisco received almost 50% more news coverage than the debut event last year. With 83 leading business and technical reporters in attendance (including the New York Times, Wall Street Journal, Bloomberg, Forbes, Businessweek, Business 2.0, CNET, ZDNet, eWeek, Financial Times, and a range of others), OSBC2005 San Francisco has generated over 510 online web stores alone (and counting), with nearly 40 million average daily visitors in total.

OSBC is not a vendor conference. It's not an IT manager conference, either. And it's not a VC/Attorney/whatever conference. It's an industry conference that brings together the best and brightest from all parts of the industry to determine where to take open source next. The demographic shift – from vendors to buyers – will continue at OSBC2005 East and OSBC2006 West. More and more CIOs are asking about the event, looking for help in thinking strategically about how to leverage open source. Where is your money best spent? :-) (Sorry, IDG.)


Here are a few quotes from attendees to support our general excitement:

“OSBC gives me great ideas of how to leverage open source.” Andrew Black, CIO, Jostens

“OSBC is the Mecca for open source businesses.” Doron Gerstel, CEO, Zend

“Really outstanding job. OSBC’s exceptional speakers, relevant topics, and collaborative format, make it the very best open source conference.” Peter Fenton, General Partner, Accel Partners

“Flat out AWESOME job (again) this year.” Julie Farris, Founder and Chief Strategy Officer, Scalix

"The only conference I've ever regretted leaving early." Daniel Frye, VP, Linux Technology Center, IBM

OSBC2005 Boston: Call for papers now live

Interested in speaking at OSBC2005 Boston? Then visit the site to submit your proposal. Absolutely no vendor pitches will be considered and, should you manage to slip past our vendor pitch defenses, we can safely guarantee that you will never speak at OSBC again. Or maybe we'll sic Gollum on you. We want your intelligence, not your salesmanship.

Startup Focus: Funambol

In the past year, I've met several great open source startups. I've been wanting to highlight them somewhere, and figured here is as good a place as any. I'm going to periodically showcase promising startups here, and invite you to comment on them (or contact them, if you're a VC or customer with cash). Also, if you think your startup (or one you know) deserves special mention, please send it along. (Please also let me know if you'd like different questions asked.)

On Funambol, the first startup to be showcased here, I'll admit to initial skepticism. I mean, Fabrizio (CEO) is a Juventus supporter! :-)

But I've been impressed with the company as I've taken a close look. It solves a real problem (data synchronization, among other things), and is already profitable. Two hallmarks of a solid company. Anyway, here goes.

Company Overview

Q. What is your background with open source?

Internet Graffiti, the company I started in 1995, was the first Italian web company. We were selling websites and I ran everything on a Linux box (setup with floppies, remember N1-N2-...?): a web server with dynamically generated pages from a mSQL database and Tcl, an access point for modem connections, a fax server with fax delivered to emails as .pdf. Open source is in my blood.

Q. When, why, and how was your company founded?

Funambol was started in 2002. We realized that our open source project, Sync4j, was going to become the standard to build and manage mobile applications. We were open source lovers and we figured we could make a living around it, and build a great product with our community. A product that would become ubiquitous, since there will be 9 times more mobile
applications then all the applications out there today on desktops (we agreed with that part of Jonathan Schwartz's keynote at OSBC).

We chose the name Funambol because it starts with fun and it comes from the Latin words funis (rope) and ambulare (walking), meaning tight-rope walker. Being an open source company means walking on a rope, every day. We must always keep a balance between the open source community and the business. Every step we take, we remember to keep our balance. Or we will fall. And we find it amusing to be a wireless company (ever tried to be a wireless tight-rope walker?).

Q. How big is your company? Where are your offices?

We are headquartered in Silicon Valley, with a development center in Europe (as every good open source company). We also have a sales office in Milan. We have 12 employees and healthy revenues that allowed us to be profitable within the first year of starting up.

Q. What stage of venture funding are you at? Are you looking for venture money today?

We are currently raising Series A.

Q. What does your company do? What human/business problem does it solve?

We solve a basic need: getting your data and applications on mobile devices. With more then 1 billion devices out there and 750 million sold every year, we better find a way to use them. They are getting smarter. People sometimes forget that today's smart phones pack more processing power than desktops just a few years ago. They can do a lot of stuff to make our life easier.

With Sync4j, you can get your data from any source of information (DB, CRM, ERP and so on) to any device (Java phone, BlackBerry, Windows Mobile, Palm, iPod, laptop and so on). You can build applications based on the sometimes-connected paradigm (that is, you are not going to be always-on). Applications that work. Always. With the data available at your fingertips when you need it. Applications that start up immediately. Sync4j also takes care of updating the applications on the device and lets you manage the device from a remote location.

Q. Who are your top competitors? Today? Three years from now?

Microsoft with ActiveSync, RIM, Good, Visto, ... Three years from now it is going to be just us and some other open source projects... This is a market that is going to become a commodity. Open source always wins in these cases. Maybe Microsoft will still be around, though ;-)

Q. Why did you choose to use open source with your company, instead of a pure proprietary model?

Open source gives customers better software. Period. As an open source company, we save costs in sales, marketing and QA (which is fundamental in our market, since there are millions of different devices with different OSes, firmware versions and gateways to talk to). You can focus all your revenues on improving your products and adding the features your customers need. For infrastructure and middleware, open source is the only way to go. It is just better software.


Business Model/Strategy

Q. What open source license(s) do you use?

GPL.

Q. What is your revenue model? (i.e., How do you make money?)

Dual licensing. If you want to build a wireless app with Sync4j and open source it, you are free to do so. If you do not want to open source it, you can buy a license from Funambol.

Q. What is your business model? (i.e., How do you take your products to market? Who are your target customers?)

Funambol's target market is the development community. The product is aimed at developers who need to extend an existing product to the mobile space or who are looking for an easy way to create and manage a mobile application. In particular, we partner with System Integrators for deployments of the platform at the Enterprise and Wireless Carrier data centers; we propose our solution to ISVs, who want to embed it inside their product; we target Device Manufacturers who need a SyncML client solution.


Miscellaneous

Q. Based on your experience, what are the primary drivers/inhibitors to open source in your market(s)?

Drivers: quality, source code accessibility, no vendor lock-in, great support
Inhibitors: Questions like: Is this real? How can it be that good? Where is the catch? I will stick to what I know.

Q. Why do you think open source is well-suited to your chosen market(s)?

Quality, need for open architectures and protocols, commoditization.

In fact, in this market customers will consider Funambol BECAUSE we are open source while they would likely ignore us if we were proprietary:

• Our cost structure and business model is such that we can sell our products for less than proprietary vendors and yet be much more profitable. This gives us longevity in the market. Customers like their infrastructure suppliers to be viable economically, to be around for a long time.
  
• Carriers historically don't like to work with small companies as suppliers. If they don't develop a solution in-house for infrastructure, they prefer to rely on big third party vendors who will be around to support the software for years. The fact that we are open source makes us bigger than any third party supplier! We have more than 13,000 downloads a month. An incredible amount of people use our software. Not only is there no vendor lock in, but if everyone in our company was hit by a bus the software would continue to be developed by the larger community.
  
• Open source license risk. Some ISVs and even carriers are afraid of the GPL. Our dual license eliminates that objection. If a customer wants to use our software in a proprietary service or application, they can.

Q. Is Microsoft a viable, long-term threat, generally? How about in your particular market?

They are the king. They are the only closed source company with a mobile development community that can compete with ours. They are a short-term, long-term, forever-there threat for anybody. They are really good at what they do (excluding software). I love Microsoft.

Q. Anything else you'd like to add?

I love open source. For infrastructure software, it's the best deal for customers.

Jason's question: ownership or availability?

Let me hasten to answer the question: Ownership matters, not availability.

So, what is the question? Whether ownership is required to drive development/innovation, or will mere availability do? Jason never poses it directly like this, but I think this is what he's getting at. (Frankly, I suspect jet lag caught him before typing up the post. :-)

Economic opportunity is driven by scarcity. If I have a unique product my economic opportunity is much greater. The more successful OSS software companies today are doing so by building unique value (which is not shared in most cases) into their software.

Availability has the potential to increase adoption. Ownership creates economic opportunity.

Actually, I think economic opportunity is driven by lots of people using one's product, and has little to do with the relative scarcity of the product. As Michael Moritz said in an April 2004 MIT Technology Review interview:

TR: It also says you "prefer a simple product with lots of prospective customers over patent-protected devices." Not exactly encouraging words for someone in a top university lab.

MM: That's a bit of hyperbole. Obviously, if you can have a product that you can sell to lots of people that you can have technology barriers around, that's a wonderful thing. But lots and lots of customers probably afford your business better protection than a few patents.

In short, utility drives purchases - whether something actually fills a need for customers. Whether the technology is copyrighted or patented really doesn't enter the equation. I would be willing to bet that exactly zero IT buyers sit around after a vendor has left the room, debating whether they should buy a product or not because, "Man, did you see the IP on that thing? Whew! I sure would like to have some of that IP!"

No, we buy things because they're useful (or because we're mindless sheep that can't say 'no' to a well-marketed brand - I do think some of this goes on in IT, generally, and certainly with open source projects, as well). Open source happens to succeed because it's available (in the sense that it's easier for me to download and try it out than a proprietary product - not because of source code availability, but because of the licensing/distribution model), and because quality undergirds that availability.

If it were merely cheap and the source code was available, who would care? I'm spending (way too much) money on U2 tickets because I want to take my wife to see an amazing band - I'm not tempted to see Dave Rosenberg pound out death metal on his drums, despite the fact that he'd surely impose it on me for free. (Sorry, Dave.)

Anyway, after bashing on Jason for the last few paragraphs, I think it only fair to say that I agree with him. Ownership matters. It doesn't matter for the reason he noted, but it matters. The "availability" argument is the same lame argument that Real, Napster, and others are using for their (inane) music rental businesses. It's a bad argument. People like to own things. They improve things that they own, like homes. They let "rental property" go to waste.

Open source is more about ownership than proprietary software, though, and this is something that I think Jason still doesn't quite accept. People contribute to "true" open source projects because they feel themselves part of those projects. It's their neighborhood, as it were, and they communally keep it looking nice. Companies like MySQL, SugarCRM, and others that effectively are their own open source development communities also own both their code and their respective communities - the fact that they leave their code open for outside use is immaterial (or, rather, it is material in their strategy of maximizing usage of their code).

Ownership matters to open source development, as much or more so than in proprietary "communities" (where developers are wage slaves rather than full participants in the development process). Availability? It's nice, but it drives neither innovation nor development contributions.

[Reading the above, I can see that my own jet lag contributed as much to the post as my brain. Hopefully, it makes sense to someone, somewhere. With enough acid, anything is possible. :-) ]

How open source spreads (Steve Howe conversation)

Arsenal v Chelsea - 20Apr05

I stayed over a day in London to see mighty Arsenal (my club) wimp out against Chelsea. Aside from the first twenty minutes, Arsenal didn't really challenge them. But then, they were lacking Henry and Ljungberg (the one in the underwear ads all over London).

Arsenal - at the pub - 20Apr05

Oh, well. The best thing about going to the match was hanging out with Steve Howe (VP and Head, Open Source Initiatives, Information Technology, Dresdner Kleinwort Wasserstein) before and during the match. He was nice enough to accompany me, especially since we went to a Chelsea pub before the match, and I was the only one there drinking water. I wasn't parading my team sympathies (I'm not stupid), but the water probably made me ten times worse to that crowd, anyway. :-) Here's Steve ordering another pint (he's half-blocked by the column).

As I tried to pretend to like Chelsea, Steve explained to me how his project, OpenAdaptor spread beyond DrKW, where it was hatched. Not by speaking at conferences. Not with little articles in CIO Magazine. Nope. At the pub, as he and fellow IT people from other firms gathered at pubs after work.

I think this is probably how open source spreads, generally. Through personal relations. Granted, the commercial open source models have greatly amplified this, but I still think the decision to use an open source project is heavily influenced by the day-to-day, personal interactions developers have with other developers, at their own employers and beyond. With this in mind, I suspect that open source would have taken over the enterprise, anyway, but its rise would have been more gradual without commercialization propelling it along.

OSBC2005 presentations online

If you want the .pdfs of the OSBC2005 presentations, please visit the OSBC.com website, browse to find the speaker you want, and look for the "Download Presentation" link. We currently have them for Jonathan Schwartz, Larry Augustin, Kim Polese, Jason Matusow, Irving Wladawsky-Berger, Edward Screven, Marten Mickos, Geoffrey Moore, and r0ml Lefkowitz.

So, for example, you'd find Larry Augustin's here. You'll need to go to the abstract page, and then click on "Additional session information," which will take you to the full abstract/download page.

We'll be adding others as they come (mostly the legal tracks), though some will never make it to the site (e.g., Lessig's - but what would you do with his slides, anyway? They're pretty much useless without him delivering the presentation).

Speaking of which, thanks to the goodness of Doug Kaye's heart, we're posting audio streams of the keynotes to IT Conversations. Check them out here. So far, only Jonathan Schwartz's keynote is up there, but Doug will be adding more each week. You can subscribe to an RSS feed so that you don't have to keep checking back on the site, if you prefer.

Prison terms for pirates

In case you needed more proof that the US Congress is as stupid as it is Hollywood-led by the nose, it has just approved the Family Entertainment and Copyright Act, which roughly translates into other languages as "Where can we get more of that crack to smoke?"

As CNET reports:

The bill, approved by Congress on Tuesday, is written so broadly it could make a federal felon of anyone who has even one copy of a film, software program or music file in a shared folder and should have known the copyrighted work had not been commercially released. Stiff fines of up to $250,000 can also be levied. Penalties would apply regardless of whether any downloading took place.

If signed into law, as expected, the bill would dramatically lower the bar for online copyright prosecutions. Current law sanctions criminal penalties of up to three years in prison for "the reproduction or distribution of 10 or more copies or phonorecords of one or more copyrighted works, which have a total retail value of $2,500 or more."

Are these people for real?

It's not a question of whether piracy is right or wrong: it's wrong. (Actually, I tend to differentiate theft from piracy, because I think it's possible to illegally download something but still pay for it when the distribution system finally joins the 21st century to allow you to pay for things as easily as you can steal them. But I've talked about this before.) Rather, it's a question of having the punishment fit the crime, and of not clogging US prisons with...downloaders? Do we really think it's worthwhile to fill prisons with people who download Britney Spears (OK, maybe that's a poor choice of examples), such that they can't accommodate rapists, murderers, etc.?

Overstating the argument, you think? Nope. US prisons tend to have plenty of these seriously bad people, but they also are unable to accommodate many of them because of overcapacity (drug offenders are one group that tends to bloat the system). Adding downloaders only exacerbates the problem, and for a non-lethal demographic. At its very worst, P2P piracy keeps James Hetfield (lead singer for Metallica) and his label from affording a limo ride to the airport - no one physically suffers.

Seems like a poor use of prison space to me. The US Congress has once again proven that there is precious little intelligence in the Senate and Congress Just lots of lobbyists. If Alexander Hamilton was right about the masses ("The masses are asses"), what would he say about US legislators...?

My kingdom for some service

It's 7:31 (19:31) and I'm finally able to connect to the Internet from my hotel room in London. I say "finally" because I've spent the last 2.5 hours working with the hotel's ISP (STSN) to get my computer on the network. Actually, I guess it was three. But who's counting?

I walked through every detail of my configuration with customer support "technician" after technician. I did this, telling them each time that I knew very well how to reset my IP address, work with DHCP, etc. etc. But me knowing something about networking was not in their support procedures, so they forced me through the agony of their tedious list over and over again. I kept repeating, "Nothing is coming out of the jack - it is NOT live." "But sir, you are the only one in the hotel experiencing this problem."

I finally called down to the front desk and nearly screamed at them to give me proof of intelligent life. They sent in Moe, their IT guy. Moe came up, started to ask me to do what I'd done a few zillion times before, and then noticed that I was transforming a coat hanger into a noose for a homicide and then a suicide, and he suddenly started treating me like an adult. He went to the hotel's network closet and after a half-hour, figured out that my room was not, in fact, connected to anything. Nor was I alone in this - my neighbor had also spent an hour with him trying to get it fixed. (STSN was either ignorant of this or they were lying - I'm not in a forgiving mood right now....) Moe plugged in my connection to the switch, and I'm now live. I've gone prematurely gray in my hair, but I'm alive.

I'm told that enterprise support is an equally harrowing experience. At OSBC this year, at the SpikeSource town hall, Phil Moore (formerly Executive Director of UNIX Engineering at Morgan Stanley) decried the state of enterprise software support. He didn't pick out one or two vendors - he lambasted every vendor. (You can imagine that Morgan Stanley would have had a few of those....) Much like my experience today, he talked about the frustration of knowing more about the problem than the person at the vendor's customer support desk. Often, he said, his team would figure out a resolution to the problem before the vendor did, but would still need to wait on the vendor to make the change.

If this is true (Phil's comments were echoed by the CIOs in attendance - it was not a vendor lovefest), then open source is immediately better supported out of the box. Google for your answer, and fix it yourself. Open source affords that opportunity.

But maybe only to the largest IT shops? Perhaps. Perhaps these are the only ones with adequate resources/expertise to support themselves. However, even if we assume this, it begs the question why they bother paying their open source vendors for support at all (unless, as is the case with Red Hat, they're forced to by the contract)? Why pay for something you don't need? I'm sure the answer is as simple as, "Because it makes the CIO feel better." That's a pretty weak argument, though, if that $1-5M (or whatever the number is) could instead be spent hiring quality people to support the technology. Wouldn't you much rather have a local Moe than a remote STSN? I would. I did.

Selling integrity

As mentioned in a previous post, I've been sitting in on a friend's company as they walk through a three-day strategy session, defining the company's future. Today, the company had to settle on target markets (having grown opportunistically for the past four to five years). On the table, among others, was the porn industry. (Not to be mis-labeled as "Adult," which is an unfortunate euphemism for juvenile "entertainment." It's porn.)

I held my breath (well, not quite - I interjected my opinion which, if you couldn't tell, is diametrically opposed to porn), waiting to see if the company would sell out. People expressed distaste for the market, but would hedge their morality with "Still, there is a lot of money in enabling the 'adult' industry." Ultimately, enough people stood up to say that they didn't care how much money was in it, they felt uncomfortable with it and wanted it off the table. It died.

I silently cheered the decision. More important than the implications for the company's support of the porn industry was a principle: money should not buy integrity. For those who think there's nothing wrong with porn (I don't know any of these people, but I'm sure they exist), this would not be an issue. Their integrity would require that they support porn as a focus market if they felt it was also an important revenue generator. But that wasn't the case in the room - these people felt discomfort with the industry, but were initially loathe to speak out and close porn off as a corporate focus because there was money involved. Ultimately, integrity to their ideals won out. Very refreshing.

Taking this off my moralistic pedestal, doesn't the same hold true for open source? Many have allowed their ideals to wane in the face of near-term revenue pressure. They have sold their birthright for a mess of pottage, to use the biblical phraseology. For those without open source religion, this is not an issue. But I routinely talk with open source proponents who cut corners in the interest of pleasing investors/making money.

Clearly, open source offers a viable revenue model. Actually, several. But it is often a different, more gradual revenue model. (That is, the kind that most industries since this world began have deployed. This hyper-growth of the software world has been somewhat anomalous in historical terms.) Sticking to it requires fortitude.

More importantly, it requires integrity.

Should we hasten standardization/commodification?

Jonathan Schwartz declared at OSBC that standardization drives market growth. I think this is correct. It rings true, and history bears it out (over and over again). Jonathan's more insightful point is that a correct view on standardization is that it raises all boats - so, one should take a proactive view toward standardizing (ergo, in a sense, commodifying) one's own industry.

That said, it's difficult to embrace standardization unless you're on the losing end of a technology battle. It's hard for a winner to see the value in standardization, because it enables competitors to compete more effectively, or appears to. In the office productivity suite space, of course everyone but Microsoft wants standardization (meaning, neutral, industry standards) of office file formats. Microsoft has no incentive to open up their dominant file formats. And, speaking bluntly, why should they? To benefit humanity? They can persuasively argue that the industry did not standardize without them swinging the monopoly club, so why should they rely on the unguided market now?

Why am I thinking about this? Because I've been sitting in a 3-day strategy session for a friend's company, acting as an outside observer/agitator to the process, and standardization is (to my mind) at the core of their business. They're in an emerging market, one that is highly fragmented. No surprise, it's in the embedded/mobile space. The company seems to have a technology lead over most competitors, and offers strong proprietary value to its customer base.

My perspective (still largely uninformed, to be sure) is that the company wins more by attempting to standardize their corner of the mobile world, even at the expense of their proprietary technology. It strikes me that they're trying to lock up what is still a very small market. Let's say they succeed: they're proprietors of very little, and perhaps it keeps their market from expanding to its full potential for many years. (One analog is the PC - Apple owned a very small market, and even its 2% share of a completely standardized computer world today is much bigger than its dominant share of the nascent market. Only the opening up the IBM PC set the market on fire, allowing much more money to be made.)

Highly correlated to the idea that standardization creates growth is the idea (again, proven again and again) that low cost expands a market.

Effect of cost on PC market
Originally uploaded by mjasay.

Larry Augustin encapsulated this well in his OSBC2004 presentation. I've included here a slide that I use in my company presentations which includes Larry's slide. Unfortunately, I mis-label the slide somewhat - the real message to get from the graphic is that standardization has dramatically lowered hardware costs in the PC business, which has both expanded the market and freed up a great deal of resources for users to spend on applications to make the hardware power more usability and productivity.

In this startup's case, their technology will not lower hardware costs, standardized or not. Is it still in the company's interest to standardize, or to make a Microsoftesque land grab to try to make its technology the de facto standard, the way Windows/Office have been? I want to say that the company should pursue standardization to grow its overall market size, and then find ways to monetize that market, rather than squeezing existing, proprietary value from a nascent market. But that's a hard argument to make when the company is in a position to milk existing interest in proprietary value, AND needs to deliver a return on its investment in that technology NOW and not in the distant future.

It's a hard decision. The open source bigot in me wants standardization. But the innovator in me loves the company's technology, and sees no reason that it should free up that value for others' use until it absolutely must. Thoughts?

"Ones to Watch" from OSBC

As part of the Startup Showcase at OSBC2005 West, we had a panel of judges (all of them VCs with leading firms) judge the participating startups. From our group of 25+, they culled out "The Ones to Watch." Here they are:

SugarCRM
Groundwork Open Source Solutions
Realm Systems
BitTorrent

Congratulations to each of these promising companies. Take a look at what they do - very diverse, with open source at their cores.

Next decade of software (Presentation)

Attached here is my latest slide deck. It's in .pdf, but if you want it in its native format, just email me.

More data on wireless billing

So, as is often the case, my mouth (or, rather, fingers) was out in front of reality. I received some interesting feedback to my thoughts on the mobile billing world. I've pasted in one particularly insightful comment below. (No attribution, however, because I'm not sure he wants it.)

Not sure where you got this, but ....

The tier 1 wireless telcos in Europe want a very large number of apps, written by third parties, typically in some form of XML. Unlike the i-mode model, which this to an extent emulates, a number of layers are far more chaotic, notably the handset space but also the decentralisation of decisions made by networks of affiliates and subsidiaries across national borders for stuff like billing systems and location services.

The tier 1s also want (unlike the i-mode model) to take a very large share of transaction revenue, as well as data revenue, which acts as a barrier to the model's success (Amazon can't sell books on 60% revenue share). The other issue for content providers is that the tier 1s want to wrap the content in their own branding, some of which is highly restrictive (not just striping red or orange logos round everything, but enforcing conformance to style guides). As a result strong brand owners are looking at off-portal solutions.

[Matt Comment: This strikes me as equally true from a consumer point of view or, at least, a certain segment of consumer. I can't remember the last time I used Verizon's or Cingular's portal on my Treo (I'm now with Verizon, but used AT&T/Cingular for 1.5 years). I go straight to the "apps" I want, unmediated by the carrier, because I don't want the carriers' imprint on my content. For the same reason, I've never understood AOL or, frankly, Yahoo!. I know there's a large population of people who appreciate such intermediaries, because they "dumb down" the browsing experience. But I'm not in that group.]

As a result, to try and grab back the revenue share, the tier 1s got together and founded Simpay. Simpay is a way of getting off-portal purchases to be paid from your phone bill. If you use it, you're back in the 60% revenue share game. The other aspect to this is that only certain payments can go through this - you can't buy a $40k car on your phone bill. The upper threshold is about $10. Most of the telcos also have a "macro billing" solution in place for this kind of issue, but have no services which can consume it.

Your AT&T contact [Matt Comment: r0ml Lefkowitz, no longer with AT&T Wireless] is right, billing is a very interesting issue, because it is currently (after technology churn) one of the biggest barriers to new service uptake. AT&T, however, paid a massive amount of money to QPass. Nobody in Europe is paying anywhere near that kind of money for a billing framework (compare QPass with Valista). I think the billing software/service market in the US is anomalous.

Very good stuff. Mobile/embedded is the new frontier - we just need to resolve the complexity that still plagues it, both on the front-end and on the back-end.

Some pictures from OSBC

The audio feeds will come later (through IT Conversations). Many, many thanks to Doug for attending the show and agreeing to make the proceedings available beyond the Westin St. Francis Hotel, where it was held.

For now, here's a growing store of pictures taken at the conference.

The insidious bribery continues (PSP)

So, the OSBC organizers are scheming the next gadget to give away to (early-bird registrant) attendees. For OSBC West 2004, we gave away the

iPod Mini.

For OSBC West 2005, it was those cool

Bose Quiet Comfort 2 headphones.

For OSBC East 2005 (November 1-2, Boston), it's looking like we're going for the

Sony PlayStation Portable.

Here's another look. It's beautiful. It's cool. It's that last little nudge we hope you needed to register early and join us at the conference this fall.
:-)

The new Internet phone

Schwartz - cell vs pc adoption - OSBC2005
Originally uploaded by mjasay.

Jonathan Schwartz made a provocative statement at OSBC2005 on Tuesday that I doubt most people caught (being enterprise software focused). He talked about the growing divide between PCs and mobile devices (the latter having much, much more volume, as shown in this slide), and then stated that much of the developing world would likely experience the Internet first (and then primarily) through mobile devices. Not laptops. Not desktops. Phones.

To Americans, this should be a shocking statement. The Internet on a phone?!? To Europeans, Japanese, Koreans, and others in developed mobile markets, perhaps an eyebrow-raiser. But to Latin America, China, and other such companies, it's excellent "news."

Yes, there's work to be done on UI to make phones more usable as Internet-accessing devices. But it's very doable, and with (by Jonathan's numbers) over 1 billion people online today, a "problem" worth solving. After all, that number leaves several billion more who cannot or prefer not to afford a clunky laptop or desktop - they already carry a phone.

SugarCRM: I knew them when....

So, a year or so ago I spoke at an SDForum event on open source. It was their inaugural open source event, and so fairly small. But a great group.

Three of the members of the audience were John Roberts, Clint Oram, and Jacob Taylor, the three founders of SugarCRM. I remember John excitedly telling me what they were planning to do ("planning," I stress, because I don't know that a single line of code had yet been written). I also remember not really knowing how seriously to take him.

Not seriously enough. Next thing I knew, they were on their Series B and my firm at the time, Thomas Weisel Venture Partners, could not get into the round. Today, SugarCRM walked away with one of OSBC's awards for the most promising open source startups, and got a mention (or five) in just about everyone's presentation/hallway chatter. These guys are pulling in serious deals (quickly moving into the enterprise space). They've done everything right in marketing themselves and raising downloads for their product. (It helps that the product is excellent.)

Best of luck to the team. It just goes to show you: Even people who supposedly have a clue about open source like me can often not see the forest for the trees.

Value in the billing....

Had a great lunch with Nat Torkington of O'Reilly on Tuesday. We talked about a range of things, including mobile phones. Nat talked me down off the make-money-on-data ledge I was on.

His passion (and O'Reilly's)? Get wireless carriers to open up phones to hacks, add-ons, and innovation. This flies in the face of their current business models, but need not. As we discussed, why can't carriers enable a thousand services to bloom on their devices (services they neither create nor necessarily condone), and simply offer a convenient way to bill for them? As r0ml presented at OSBC2004, billing systems (for which he was responsible at AT&T Wireless) can be "kewl." They're also a nightmare to build and adminster. As such, they're a real competitive barrier, and a strong value.

Carriers need not be in the business of trying to mastermind applications for their customers. The market will do that. They simply need to get (mostly) out of the way, and provide service/app providers an easy way for subscribers to pay those vendors money. They can take a slice of every transaction.

Am I missing something?

In memoriam: OSBC West 2005

What a fantastic show. I started the morning of this second day with Tony Wasserman of Carnegie Mellon (and the Center for Open Source Innovation) telling me that Tuesday had been "the single best day [he] ha[d] ever spent at a conference." I heard similar comments throughout the day, which makes me very, very happy: if people feel that they have had more than their money's worth, we've done well.

That attendee satisfaction allowed me to sit back and enjoy many of the conference sessions, whereas last year I was so stressed out (along with my fellow co-founders) that I didn't actually enjoy the conference. This year, I heard some of the most interesting, insightful analyses of the open source business world that I have ever heard. The most we can say is that we helped to get it started - all of the value came from the intelligence of the speakers and the active audience.

Some examples are in order.

At OSBC, we have tried to balance out the two days with equally excellent speakers. Jason Matusow gave a pretty frank presentation on the costs and benefits of open and closed source: he didn't offer up a proprietary software dogma, nor did he soft-pedal his concerns with open source software (in some contexts). He was pretty clear: the world is big enough for both, and there are good uses for both.

I'm actually optimistic that Microsoft will increasingly bake open source development into their methodologies, though Linux will continue to be a threat. Frankly, it behooves us to separate the two: open source is not necessarily a threat to Microsoft. Linux is. Linux as it is being adopted by enterprises is a product sold by my company, Novell, and Red Hat (and others). OpenOffice is a competitor to Microsoft - the development process that creates it is not.

This brings me to Larry Lessig's fantastic keynote. The man is a rock star. If you've read Code and Other Laws of Cyberspace, you'll appreciate Larry's arguments that the Internet works because the core is free/open, allowing for innovation (both proprietary and open) at the edge of the network.

In like manner, I see this happening in the software stack. Increasingly, the "core" of the stack (Operating system, database, etc.) is being opened up. Some of the "edge" (CRM, ERP, office productivity, etc.) is, as well, but need not be. So long as the foundation is free/open, vendors and developers should be given a long tether to create value on top. We're seeing this all around us: Google, Friendster, and a myriad of other startups are building "proprietary" services on top of an open "network" core: open source software.

Moving on, I heard Geoffrey Moore's talk was great. I wouldn't know - I was out trying to fix the wireless. I failed.

SpikeSource held a private reception/town hall on the evening of the first day, which was mostly instructive in just how much CIOs despise the treatment they receive from vendors. The panel, moderated by Ray Lane (former president, Oracle, and general partner, Kleiner Perkins Caufield & Byers) included Tim O'Reilly (President, O'Reilly & Associates), Kim Polese (CEO, SpikeSource), Peter Quinn (CIO, Commonwealth of Massachusetts), Tim Golden (SVP, Linux Design & Engineering, Bank of America), r0ml Lefkowitz (VP, Research and Executive Education, Optaros), and Charlie Brenner (SVP, Fidelity Center for Applied Technology, Fidelity Investments). Great line-up, and extremely interesting dialogue, especially as inflamed by audience members (which included a packed-room full of CIOs and other IT executives), who clearly don't feel that they're getting their money's worth from their various vendors. Is enterprise software really that bad? Apparently so. SpikeSource (and SourceLabs) certainly seems to have something these companies want, but only if it/they can provide the superior service that is apparently lacking from enterprise software firms.

Earlier in the day, Kim delivered a thought-provoking keynote about the opportunities afforded by commoditization. As Jonathan Schwartz also argued (see below), commodities allow us to shift our competitive focus. In Kim's world, open source offers vendors the chance to spend more time integrating useful technology and tailoring it to meet customer-specific requirements. SpikeSource has built a framework to manage the complexity of the various open source projects that combine to solve customer problems. The more I hear her speak, the more I believe the fundamental premise underlying the company.

On that, Larry Augustin gave a superb keynote on Tuesday, highlighting the odd software business model that spends 76% of license revenues on sales and marketing. That means, as Larry drove home, that software makers generally spend far more on trying to sell their software than they do in actually making it. Open source, on the other hand, provides a more efficient sales/distribution mechanism, allowing vendors to invest more in product excellence, and less in trying to get customers to look past their products' weaknesses.

Jonathan Schwartz gave an excellent keynote on "The Participation Age." More money is made in standards-based industries where value migrates up the stack, as it were, rather than in building a better railroad tie, or whatever. Services that run on the commodities - that's where the money is. Not in attempting to hold back the commodity urge to protect profits on yesterday's innovations. An exceptional presentation.

There were a range of others (Marten Mickos, r0ml (though I'm still a little lost on what he said - I was an English major, after all), and others). What a great group of people, both speakers and audience.

OSBC West (February 14-15, 2006)

For those preferring the sunny climate of the Bay Area, we're announcing the dates of OSBC West this morning.

February 14-15, 2006, The Argent Hotel, San Francisco. Please plan to join us.

Also, we're kicking off a series of open source legal events in Seattle this September (September 7, 2005, to be precise). If you or someone you know would like to learn more about the nuances of open source licensing (as well as the effects of patents and such on open source from a legal perspective), come spend a day with us. We deliberately planned our inaugural legal event in Seattle so that the Redmond folks could join us. Word on the street is that they need this sort of training. And software. :-)

OSBC East (pre-announcement)

OSBC will be heading East this Fall...to Boston. We're going to officially announce it tomorrow, but you might as well know now.

November 1-2, 2005. Marriott Newton (along the 128/95 corridor in Boston).

We're looking for the most innovative startups to join us there, so please tell us about what you do. Go to http://www.osbc.com/live/13/startups to fill out the submission form. Startups are welcome from any geography. Come be part of the $150M (and growing) bonanza.

$150 million in open source venture capital since March 2004

I'm an English major, so any math I do is somewhat suspect, but my rough tally of venture capital that went into open source companies since the last OSBC is (drum roll please)...

$150 million

That's a lot of money, and only the tip of the iceberg.

(OSBC can claim some of the credit for the 2004 investments - a fair number of contacts between funders and funded happened at the March 2004 OSBC. /patting selves on back. :-)

Money is flowing into open source startups at a torrid pace, which begs the question:

Why aren't you starting one? (For that matter, why am I not doing so? ;-)

I haven't added up total venture funding (which would include MySQL, Red Hat, VA Software/Linux, Linuxcare, JBoss, etc.), but it would take the amount well over $200M. As for returns on those investments, Red Hat's market cap, alone, justifies the investments. All by itself, Red Hat puts us at a 10x return on all venture money spent. Not bad.

Looks like the open source investment hoopla is founded in intelligence, after all.

Red Hat's results...

So, Red Hat posted its results, and they look good. I work for Novell, but I'm happy to see any company proving open source as a viable business strategy. Even a direct competitor. (Especially one that employs several of my friends.)

It's interesting to see how focused Red Hat has become. Red Hat is an enterprise server OS vendor. Period. It has never done anything else very well (just as Google struggles to do more than search), and it's a net positive for Red Hat investors that the company is no longer pretending to do other things.

Like embedded. The company has made ~$5M from its embedded business since time began, and has never done much to improve this number. The number is now in decline, despite its partnership with Wind River. While the rest of the world is linking embedded to the enterprise (this is Wind River's explicit strategy, and nearly every other enterprise vendor is at least looking at ways to extend its software to mobile devices), Red Hat has turned away from the market.

In the short term (by which I mean the next 2-3 years), this is probably the right thing to do. There is just too much low-hanging Unix "fruit" to gather, and Red Hat lacks the resources to effectively fight on two fronts. Or, rather, it lacks the ambition to do so, in the sense that Gary Hamel and C.K. Prahalad write in their classic Strategy as Stretch and Leverage" article in Harvard Business Review. Resources are always available when there is a will.

But in the longer term (3-5 years), I think Red Hat will reconsider. By that time, MontaVista (and Novell?) will be far ahead of Red Hat in embedded, and the game will be up. Worth watching.

New OSI board announced

Danese just announced the new OSI board. I think it's less noteworthy for its non-American demographics as it is for its pragmatism. Open source advocates, all, but no open source bigots. (My bias is for pragmatists.) The times, they are a'changin'. :-)

Intel disarms (its open source license)

Danese has a great post on the behind-the-scenes efforts of OSI to limit license proliferation, as well as a success story on Intel disarmament.

"Moisture" and open standards

I live in Utah. There are lots of Mormons here. This is normally not a problem (I'm LDS/Mormon), but we're in the middle of a 7-year drought, and I think Mormons may be the reason.

You see, someone, somewhere in Mormondom decided that "moisture" was the term de rigeur for importuning divinity for rain. Other people/religions call it "water" or "torrential rain" or "best snow on earth" - not us. It's "moisture" that we routinely (daily) pray for. And that's about all we get. (Till this year, mind you - the snow is DUMPING this year. It's moist snow, no doubt, but snow all the same. Not moisture. :-)

Now, I don't know if you knew this, but the connotation for "moisture" is not "lots of water." Its denotation is not much better: According to one dictionary (and echoed by virtually all others), moisture is "wetness, usually felt as a vapor in the atmosphere or as a condensed liquid on the surface of an object." Millions of Mormons every day exercise their faith that divine powers will bring us...dew. Hardly the thing to end a drought (unless it's a lot of dew. ;-)

In like manner, we often conflate open standards with open source - we want open standards, but we "pray" for "open source." Let me explain (though I doubt I can do it as well as Jonathan Schwartz, who argues a similar point constantly (and very well here. I think he'll be addressing it on Tuesday at his OSBC keynote, as well.))

Open source and open standards are not the same thing, though they often go together. Open standards ensure industry-wide interoperability (if they are adopted) - open source does not. Open source simply means that one has access to source code.

In practice, Linux and other prominent open source projects are becoming de facto open standards, because they invite broad participation in single projects, which leaves vendors and buyers free to build on them (without lock-in). But at inception, open source is not, by definition or really even by implication, open standards.

If I had to choose between the two, I'd take open standards. Like rain over moisture, open standards are what will make it so that my data are not stranded in a vendor. Open standards provide me, as an IT buyer, the bridge from one vendor to the next. Open source, as seen in Red Hat's licensing and development model (locking buyers into paying for support they may not need, and preventing easy interoperability with other Linux vendors by closing binaries - it's proprietary open source, as Ian Murdock has blogged), does not achieve the same thing. At least, not automatically.

So, while open source is a Very Good Thing, it's not the only thing one should consider when looking at software. If you're an IT buyer, open standards should be the overriding concern, with open source a close third. Your second consideration should be whether the product meets your needs. Why standards before needs? Because your needs tomorrow may change - if you're locked into a product because you bought into a closed standard, you're bien fichu, no matter how good it was at the beginning.

In short, pray for "open standards." Not moisture.