SDForum's Open Source Executive Workshop

Sometimes I don't know whether to consider Laura Merling, Executive Director of SDForum, a friend or a foe. She and the SDForum team (in alliance with Andrew Aitken and the Olliance team) are putting on what looks to be an excellent open source event in January. Confound her for staging a quality event right before OSBC! :-)

SDForum's Open Source Think Tank
Seriously, I like the idea, and it's something to which my recent blog entry speaks: we need more intelligent discourse between members of the open source business community. SDForum's event is designed to be just that: a think tank for leading open source executives.

I've been fortunate to be friends with the executive teams at MySQL, SugarCRM, JasperSoft, and others. Talking with them keeps me informed...and humble. Just when I think I know a lot about open source, I talk to one of these and find out that I know very little.

Laura and Andrew are bringing this insight (and humility :-) to the growing body of open source executives, and I can't help but think we'll all be the better for it.

Here are some details on the event.

Who?
By invitation only. Will include ~75 industry influencers including CEOs from the top open source companies (meaning Alfresco, of course ;-), executives from large technology vendors, and leading venture capitalists.

When?
January 19.

What?
An open discussion moderated by John Markoff, Senior Writer, New York Times with Tim OReilly, President, O'Reilly Media; Jonathan Schwartz, President & COO, Sun Microsystems; Rod Smith, Corporate Vice President, IBM; and many others. SDForum has arranged a professional facilitator to then manage a range of break-out sessions to generate intelligent discussion (and debate) of where open source is going, and how to get there.


It should be an exceptional opportunity for those invited to attend, and I'm hopeful that its effects, like OSBC's, will reverberate well beyond the small group assembled. Open source, contrary to popular mythology, hasn't magically taken off in the enterprise. It took some extremely intelligent people at Red Hat to figure out its innovative license model (helping them get funding, then an IPO, and then customer POs); it took similar neural activity at MySQL, SugarCRM, and other successful open source companies to get them where they are today.

The good news for the rest of us is that we have this SDForum event, OSBC, and a few other select events to help us figure out and craft the future of open source.

Blogs I read (Please point me to others)

A few people have asked me which blogs I follow. (OK, two of the three inquirers were my children, but....) Here's a list. (Dave Rosenberg and I have complained before that there isn't much intelligent discussion of open source in the blogosphere. If you know of something worthwhile, please share, either directly (masay @ osbc . com) or in the comments. My primary objective in posting these is to elicit feedback on others I should be reading.)

The List

• InfoWorld's Open Resource by Dave Rosenberg...and me. I suppose this is a bit self-serving, but I was reading Dave's blog long before I joined it. Dave's blog is the best place to get unvarnished opinions on open source (plus web services and other random stuff). Dave doesn't bother to try to put a veneer on rubbish.
  
  
• Nick Carr's Rough Type. I've been getting to know Nick through OSBC. In terms of general technology commentary, no one is better than Nick. Period. Again, I appreciate Nick's perspective because he tells it straight. He's not trying to polish his opinions to be sexy with this or that crowd. You may not agree with him, but at least you always know where he stands.
  
  
• Jonathan Schwartz's blog. Jonathan's blog is often self- (or, rather, Sun-) serving. But it's also one of the few corners of the 'Net where intelligence meets technology. Also, like the others above, Jonathan generally tells it straight. (Notice the theme? I like the news or commentary to be candid and forthright.) No varnish, so long as you discount the pro-Sun cheerleading.
  
  
• Infectious Greed by Paul Kedrosky. I resisted this one for some time (though I can't for the life of me remember why), but eventually capitulated. Paul writes about everything (though usually technology or VC-related), and has a fascinating, quantitative way of looking at things.
  
  
• Zack Urlocker's Open Force. Zack is the VP of Marketing for MySQL, which fact occasionally trickles into his blog (not that I can complain - I promote Alfresco here at times, as well). But when Zack is on, he is really on, and ends up giving away some excellent insight into how one of the world's most successful open source companies operates, and where the market is going.
  
  
• O'Reilly Radar. I subscribed to get access to Tim, but I've found that Marc Hedlund's writings tend to hit hardest (and most often). I don't really get into the geeky "Gee whiz! Look at this new scripting language trick!" that infects many of the other posts, but this is a fault of mine, not of Rael's or Nat's. There are lots of people (maybe you?) that love that stuff. It's just not my thing.
  
  
• Stephen Walli's blog ("Once more into the breach.") I check in with Stephe to get his take on standards, since few to none have played that game as long as he has.

There are a few more that I've either just started to track (like Mark de Visser's blog focusing on open source marketing and Stephen O'Grady's Tecosystems), or only sporadically track (like ZDNet's Between the Lines, which is filled with a lot of stuff that I don't care about), but that's pretty much the core list.

As far as news feeds, I subscribe to Linux Today, The Register (Hilarious and cynical), BusinessWeek, eWeek Linux, and a few others. But I can get the news anywhere - I like the commentary-style RSS feeds the most.

Used to read

• Jason Matusow's blog. The 2-3 times he actually posted something. It's a shame that one of the smartest people in open source has moved on, and never shared much anyway. Typical Microsoft !%!%!%. :-)
  
  
• r0ml Lefkowitz's blog But then he went silent....

Oh, and I also follow a large number of Arsenal-related blogs and news sites. For Arsenal, however, I prefer the news sites, and don't care much for the commentary.

Lessons 2005 (#2): Support is not a viable business model

There has been a pervasive myth in open source since its inception. The myth? That IT buyers will pay for support. A small percentage will, of course, but not enough to sustain a venture-backable company.

Think you know of some counterexamples? There are none. There are companies that appear to make a viable business off support (like Red Hat), but they're not really selling support. They're selling bits and bytes, disguised under the banner of "support." This isn't a matter of such companies being sneaky - it's a matter of them recognizing what customers will actually pay for.

Also of note, the closer vendors come to overtly selling software (instead of the support thereof), the easier the time they have. I think Red Hat is an incredible company and think its business/licensing model is absolutely brilliant: make source code available (though with difficulty) but lock down the compiled product, and lock it down further with a contract that keeps customers away from the certified version without payment. But word on the street is that Red Hat is having difficulty getting its customers to renew their "support" subscriptions, and is being forced to seriously discount in order to keep customers renewing. 40% or more.

In an ideal world, it wouldn't be this way. Customers would pay for value and wouldn't try to get something for nothing. But we don't live in that world, and if there's a way to use something for free, enterprises will.

I've written on this before. It's unfortunate, but it's true (and there's not a single significant open source company out there right now that hasn't had to grapple with this). Many enterprises either use open source software without support, or discontinue support after a year or two. They're rational beings (or collections of them), and won't pay if it's not a requirement to do so. I'm not talking theory here - I'm talking the day-to-day reality of advising open source startups and working for them (Alfresco now, Novell and Lineo before).

So what's the Right model? I don't know. I don't think anyone does. But it's not support alone. That much I know for absolute certain.

What I learned in 2005 (Lesson #1: Open source is not a fix for weak code)

Dave Rosenberg had a good idea the other day. Rather than come up with our predictions for 2006, we decided to share some things that we learned in 2005 (which will impact technology in 2006).

So, since my wife kicked me off FIFA on the Nintendo, I'll start.

Lesson #1: Open source is a poor substitute for a weak product.

There's a feeling out there that the panacea for weak marketing/engineering/management/fluoridated water is open source. Have a dying business? Open source it! No one wants to buy your product? Open source it! And so on. I would have thought that this idea was thoroughly discredited years ago, but based on the things I see VCs funding, I have to believe that people still believe in the open source fairy.

I've said this before. But apparently the message isn't sinking in. Good open source projects are founded on good technology. Of course, successful open source projects require more than this, but great code provides the foundation for a great community.

So, a lesson for closed-source companies hoping to be relevant again by opening up some or all of your code: don't bother. It won't help. Instead, work to develop exceptional code and then follow that up with a strong licensing and revenue model to ensure you can monetize that code. Because support alone doth not a successful business make. But that's a lesson for another day....

Open source and Nessus: Paradise lost? (Free as in free loading?)

Linux Insider has a great story on Nessus' move to closed source with its 3.0 product. The article has some wringing their hands about the decision, acting as if open source were going to shrivel and die because the code's primary developer decided he couldn't make a living off free-loaders.

And so Nessus is facing complaints from users like Alan Shimel, Chief Strategy Officer for StillSecure:

"Here's the danger we are running into," he said. "People contribute resources to these communities, whether it be time, money, or code. When they see everything they give converted for the commercial success of an individual rather than as a community as a whole, how long do you think they are going to want to keep giving?"

Don't you believe it. I'm willing to bet that the vast majority (80-90%+) of Nessus development is done by the Tenable team. Most of those complaining about Nessus being closed off are complaining not about any loss of community (though they couch in that language), but rather at the loss of a free (as in cost) piece of infrastructure they enjoyed using.

Just as they complained when Red Hat split its Linux product line into Fedora and RHEL. And when MySQL had its dual-license scheme, and then moved to its Network product. And Sugar when it gave away most of its product, but kept some proprietary.

That's the nub of the issue, isn't it? Open source allows, even encourages, a certain amount of free-loading/free riding in the quest for the few who will return to contribute cash and/or resources to fuel further development. Free riders talk about support models and the like, but few to none of them actually pay for it. That's the dirty little secret of open source. Pure support models don't work except for smallish companies. Because most buyers really don't value support as highly as they do the software itself.

And so Red Hat, MySQL, SugarCRM, Groundwork, Alfresco, etc. come up with innovative licensing schemes to get people to pay for software...while calling it support. It really shouldn't be this hard. The problem, however, is that a generation has grown up believing, whatever RMS may say, that "free software" is free as in beer, not as in freedom.

Hence, you get silly angst like this one:

...Shimel said users are now forced to make a decision, with three options available: use Nessus v3.0 for free but with a seven-day delay in updates; pay Tenable fees required to obtain a direct feed for updates; or transition to a commercial vulnerability management system.

Regardless of the long-term implications for the open-source community, the move to Nessus 3.0.0 has short-term implications for security software vendors and users. What do individuals and corporations do?

I've got a wild and crazy answer to that question:

They should pay.

HP leaving LinuxWorld?

Word on the street is that HP has pulled out of LinuxWorld...completely. If so, it marks the continued exodus of the BigCos from the BigCons (Big conventions).

I talked with the events person at Large Database Vendor X a few months back, who told me they were moving away from the big exhibitions because they found the smaller, more tailored events more effective. Maybe this is HP's rationale, as well (assuming the news I heard was true - very good source, however, so I'll assume it is until I hear otherwise). At any rate, we're saving a slot for you at the Open Source Business Conference, HP. Give us a call. :-)

I wonder what this means for Martin Fink's advisory board position with LinuxWorld...?

And, lastly, it's a bit unfortunate since IDG finally filled the gap left by Dave Rosenberg's departure with John Mark Walker, formerly of Groundwork. John Mark is a sharp guy, and should do a good job of transitioning LinuxWorld into a broader Open Source Solutions World tent. No one can replace Dave, but IDG did well to get John Mark.

Google bans Kozoru for mashing up with it

Hopefully, this will be the last Google-related post for some time. But when I read Dave's post on Google's banning Kozoru, I couldn't help it. I consider myself a fairly religious, spiritual person, and thought I knew what evil was. But I guess my definition and Google's are very different. This bang-up on Kozoru's mash-up isn't evil, but it certainly calls Google into question as the proverbial tech Good Guy.

It also begs the question of whether it's wise to build a business on someone else's technology, as Dave calls out. I'm talking about product-based businesses. Frankly, like Google and the others who ride on top of others' networks. BusinessWeek has a short but interesting (and very Lessig-esque) article on the possibility that the broadband networks might start charging those that freeride on them.

Just ask AT&T's CEO Edward Whitacre Jr., as BusinessWeek did in its November 7 issue:

"What [Google, Vonage, and others] would like to do is to use my pipes free. But I ain't going to let them do that."

(Btw, I loved the WSJ's piece on Google's SVP of Sales (Omid Kordestani) behind the AOL deal. He sounds like a great person, and actually gets respect at a company that tends to disdain anyone that isn't a hard-core engineer.)

Google gets glitzy...for a fee

First, Google did no evil. Then they decided to do a little because the money was just too good. (For the Google fetishists out there, I'm defining "evil" on Google's terms, not mine. I don't personally think it's evil to do deals with big companies and give them preferred treatment. Last I checked, pride was the most egregious of the Seven Deadly Sins. Google has that one in spades. :-)

Now they're deciding to look a bit more like Vegas. Again, because the money is good. It's something they've never done, chiding others for doing it and insisting that they'd never be so shallow. But when $300 million (and counting) is at stake....

According to the New York Times:

Users of Google's search engine will soon see something they are not used to on the notoriously spare site: advertising with logos and graphics. And the advertisers will not be limited to America Online, whose talks with Google prompted the change in policy, according to two executives close to the companies' negotiations.

As part of their deal, which is expected to be announced this afternoon, Google is providing AOL with $300 million in advertising on Google's Web sites, intended to use to draw Google search users to related content on AOL's sites, the executives said. That sum is on top of the $1 billion in cash that Google is to invest to buy a 5 percent stake in AOL.

Funny thing, principles. They become so malleable for some when money is on the table.

Am I saying that Google has sold out on its principles? No. Not yet, anyway. What I am saying is that the company appears to be making small concessions (which, of course, invariably lead to large ones - the road to banner-ad-splattered websites and bought-and-sold search is paved with good intentions). Google is the best search company out there. Period.

But Google needs to be careful that it doesn't become just like everyone else. This AOL deal puts them on that track.

Yes, I am an old man

Do not go gentle into that good night,
Old age should burn and rave at close of day;
Rage, rage against the dying of the light.
Dylan Thomas, "Do Not Go Gentle into that Good Night"

Last night I was reminded by just how old I am. My wife and I attended the U2 concert in Salt Lake City, and the only ones older there than we were Bono, The Edge, Adam Clayton, and Larry Mullen, Jr.

Well, that's not completely true - we were surrounded by old fogies like ourselves. In fact, we kept bumping into old junior high school friends and, oddly, old Novell friends. U2 concerts must be where old friendships and work acquaintances go to die....

Anyway, given that Bono et al. could hardly move - tapping their respective feet was nearly all they could muster - I guess I should be glad that my own geriatric feet could at least keep time and I occasionally clapped. I had been wanting to attend a U2 concert for nearly 20 years, and finally had the chance. When it came, however, I found that I was more interested in getting home to go to sleep than in reliving my youth. (In fact, I left in - what turned out to be - the middle of the second encore, because they kept playing songs that I didn't like the first time around like "Stuck in a Moment (that You Can't Get Out of)." Turns out, the next song was "Yahweh," the song I'd be waiting all night to hear. But I was in a taxi heading home, and I wasn't terribly disappointed when I read the playlist this morning. I'm old.)

Btw, the concert, just like open source, got better the more the crowd got into it. The crowd - just like at a good football (soccer) match - was a crucial part of the experience. (On that note, I watched Lyon play Lille the other day as part of the Club World Championship, and it was almost dull because very few people were in attendance. Soccer just isn't soccer without 50,000+ louts screaming obscenities at the ref.)

In open source, community is everything. Otherwise it's just a group of aging rockers hoping to milk money from an aging fan base. That's why this February's OSBC in SF will deal heavily with community issues. How to tap into open source development communities (in the wild, as it were), and how to build your own.

Now I've got to go rearrange my dentures....

Google does no evil...until it pays well

So, Google got the Time Warner deal. $1B and its soul, according to the New York Times:

Google, which prides itself on the purity of its search results, agreed to give favored placement to content from AOL throughout its site, something it has never done before.

The lesson? Never, ever trust a capitalist who pretends to be otherwise. "Do no evil" was a catchy slogan for Google, but one that it was willing to sell for a few bucks.

For all those who thought Google was somehow different, I'm sorry. I can hear your idealism dying from here.

Seriously off-topic: Looking for a song

"Pete's Dad" by Pornos for Pyros (Percy's gig after Jane's Addiction). It's one of my favorite Christmas songs, and I can't buy it on iTunes or steal it from LimeWire....Can I borrow it from you? Please scour your hard drives. I'll happily send a check for $.99 to Columbia or whomever owns the song....

The differences between Red Hat and Novell

I've been meaning to write about this for some time, but couldn't. Firstly, because I couldn't touch on the subject while I was still employed by Novell. Secondly, because i didn't want to create problems for Novell while it was going through its road bumps a few weeks back.

But I thought now was a good time to talk about the differences I perceive in the two companies, having worked at the one and talked extensively with the other. In no particular order....

Customers

Red Hat has long dominated the Linux market. In part this has resulted from serendipity - the company raised gobs of cash in a boom-time IPO and so was the first big player to market - but it also results from the company's rabid focus on customers. Importantly, Red Hat has never wavered from a core understanding that the low-hanging fruit is Unix. A friend there recently speculated that only 5-10% of Wall Street is Linux right now, and Wall Street is an early adopter. This means there's still truckloads of money to be made in converting Unix to Linux, with fewer barriers to doing so (skillsets transfer, dramatically lower cost of hardware, etc.).

Novell, for its part, has had to play catch-up, just as SUSE before it did. The integration of SUSE into Novell's corporate and technology infrastructure took time, and Red Hat extended its lead during that time. However, Novell brings some serious value to the market, including superior support. Customers, like the Swiss government and the UK's National Health Service are leading indicators of Novell's customer resurgence.

It's not the size of Novell's support staff that matters most, I don't think. That matters, but the larger issue is Novell's history and accumulated expertise in supporting operating systems and software, generally. Novell has been doing this for two decades, and they really are leaders in support, certification, etc.

I think Novell still does itself a disservice by focusing more on Microsoft and Windows than the Unix market, but this is changing. It's just hard for the company to give up on its eons-old battle with Redmond.

In short, both companies are improving their customer focus - Red Hat is adding employees (though still at an intelligent pace) to be able to better service customers who spend with them, and Novell is tightening its focus on Linux (and, frankly, shed some jobs to accomplish this) to better meet customers' requirements. Both are well-positioned. I think we're finally going to see some competition in the commercial Linux market.

Culture

Red Hat has a hard-charging, take-no-prisoners approach to the market. If you're not making them money, you're not going to get their ear. At times, because of how tightly Szulik runs the ship, they simply didn't have enough employees to be able to service all the demand, causing people outside the company to view Red Hat as aloof and arrogant.

This has led the growing open source ecosystem to Novell, which is partner-centric and easy-going almost to a fault. Ron Hovsepian is changing this, and Novell is starting to become much more choosy about opportunities (customer and partnering) that come its way. The company's culture is changing for the better along with this shift in opportunity mindset. Novell is becoming less concerned with popularity and more concerned with dollars.

Here, again, I see a convergence between the two. Red Hat is loosening up and Novell is tightening up.

Partners

I already addressed this a little above, but I think the companies are converging here. Red Hat has been historically difficult to work with, in large part because they simply lacked bandwidth to service all incoming requests. So, you were either SAP and Oracle (and a few select others), or you were no one. This was good for Red Hat in that it tightened the company's focus on revenue-generating partnerships, but it was bad in that it now has to play catch-up in being a central part of the growing open source ecosystem.

In that world, Novell is the first choice because it's easier to work with and more generous with terms. Novell is becoming choosier as its clout grows, but I don't sense it's doing so out of arrogance. I think Ron is just instilling discipline. From my perspective as a prospective Novell partner, this is a good thing. I'd much rather work harder to partner with a company that has its act together than to slap together an easy, meaningless partnership that is good for a press release and little else.

Red Hat is also improving its partner programs to accommodate more. Again, this is more a matter of adding employees to cover things than it is a real shift in the company's mindset - Red Hat has always cared about partners. Now it has the people to put meat on the bones of those good intentions.

Conclusion

Novell is growing into its role as a major Linux vendor, and Red Hat is growing into its role as a major company, period. I really like the changes I see in both companies. They're still very different companies with different mindsets, but the differences are narrowing.

Linus on the Linux desktop, importance of source code, and developer craziness

What follows is a fascinating interchange between Linus Torvalds and Chris Blizzard on the Desktop Linux (public) mailing list. Linus' comments on the value of source code, the importance of letting developers run amok, and other sundry things is great fodder for those looking to better understand how to make open source projects thrive.

(Chris' comments are indented below. All bolding for emphasis is mine.)


From: Linus Torvalds
Date: December 13, 2005 1:44:23 PM PST
To: Christopher Blizzard
Cc: xxx@lists.osdl.org
Subject: Re: [Desktop_architects] Printing dialog and GNOME


On Tue, 13 Dec 2005, Christopher Blizzard wrote:

I suspect that what you see as a raging hatred for user configuration is instead just a symptom of what we consider important in GNOME.  We are willing to prioritize "working well out of the box" and "consistent and easy to use" over user configuration.  So that particular set of features just never bubbles up to the top.  As near as I can tell it just is a question of priorities.

[Linus]I don't understand why you and Havoc seem to be of the opinion that working well out of the box and having good defaults is in any way something I argue against.

I absolutely don't. I think it's very important that defaults should be sane, and the "out of the box" experience should be what a user can be expected to, well, expect. I was kidding about "focus-follows-mouse" being the only true window focus method: click-to-focus does actually make sense as a default, because it's what a lot of users are brought up to expect.

Similarly, I do actually agree with people who say that KDE is cluttered. The KDE menu system is overwhelming. But that's a totally separate issue from the notion of _capabilities_. You can decide to have a uncluttered desktop that is still _capable_. I believe we've seen that with some of the distros that do use KDE, eg Linspire. 

So you can have your cake and eat it too. It's not an either-or thing.

That doesn't mean that we don't need to fix printing, and the printing dialog - we still do.  But it means that fixing printing is probably going to take a higher priority than adding new features to the window manager. :)

I really don't care about the mouse thing. It's the reason _I_ hate using gnome, but hey, I've got alternatives.

The reason I exploded is that (a) I've been watching this desktop thing because I got added to the mailing list by mistake and (b) the gnome disregard for flexibility has been grating me for a _long_ time.

To me, open source _is_ about flexibility. And no, I'm not talking about people re-compiling their applications and making changes to them. The fact that the source code is open is in some ways both the least important and the most important part: it's the least important in the sense that in practice, very few people actually change source code, and even those that do tend to be very _focused_ on one particular project (or even just a small _part_ of a project).

So the source being open is - on average - not important to people directly. Even major developers only work on a small part of the whole stack at a time, they don't go around changing all the programs they use to suit them. 

But _indirectly_, the thing that open source really excels at, is the flexibility it offers thanks to having lots of users, and lots of users whose needs get _heard_. THAT is the core of open source. You've got different kinds of people that get attached to a project. It's _not_ a corporate mono-culture, because people from different backgrounds can get together and work on it _without_ going through the corporate mind-wash.

And to me, gnome is killing itself as an open source project, because it ends up dismissing exactly that thing. Having strict UI rules ("The HID says so-and-so") that are really a religion that you're not allowed to question. The whole notion that things are supposed to be done just one way is antithetical to what makes open source successful in the first place.

I think the KDE development process has been a lot more "lively", and I think a _lot_ of the reason for that has been that they haven't allowed the "interface nazi" kind of stifling of what people feel they need to do. Read the recent KDE-3.5 release announcement with the "visual guide to new features", and you can _feel_ the energy. Sure, they have three different kinds of desktop choosers. So what? You don't have to use them. But the capabilities are there if you want to.

And I think that's important. It's important, because that developer energy, in the end, is what get things done. And as a side effect, you will automatically end up with a system that understands that defaults may be good, but that different people have different needs and views. Because you had a very diverse group of people that worked on it.

So developers are more energized, and I think users are also automatically happier. They may not even realize why, but I believe it's to a large part because their needs are taken care of - not necessarily because they ask for it, but because the developers themselves are more varied and thus tends to have more different needs, and often took care of the different 
needs of the user base to some degree automatically.

This, btw, is also why a "enterprise desktop" should never be allowed to drive development. It is, by definition, boring and same-old, same-old.

And if you don't see the parallels with "enterprise UNIX" and "Linux" here, I think you're blind. The thing is, Linux (the kernel) got better than just about any enterprise Unix kernel _not_ by trying to develop itself for the enterprise, but by allowing and encouraging different kinds of people to all scratch their own itch.

Yeah, the whole development process is a bit more chaotic, and maybe a bit more "cluttered" and even scary, but the end result is BETTER. And yes, Linux (the kernel) has a million drivers that the "serious guys" don't care for. But that wild and crazy thing is exactly what made Linux a success in the first place.

In any case, this is a different question that hasn't been asked here, and one that I think people are stumbling over and that is "what are the effects of design decisions on the size of an open source-based community where choice is more important than design focus?"  I suspect that given that question I would expect that the KDE community would be larger, but less focused on a single vision.

I agree - I think this is part of it. But see above. I think it's a small part of a much bigger issue.

But the fact that users and developers don't know does NOT mean that customization is bad. Quite the reverse. It means that defaults make sense, but since you don't know what they'll be doing, you should always strive to have ways to let _them_ make the choice when they have some reason the default doesn't agree with them.

I kind of agree with this statement, but I think it's overused to justify all kinds of nonsense and avoid good system integration.

The thing is, developers shouldn't see themselves as system integrators. If they do, they just limit the end result.

This is a hierarchy. You don't put the developers at the top of the heap, the same way you never put developers face-to-face with your customers. That would just scare the customers away.

You let developers be developers. Encourage a bit of crazyness, because the best developers _are_ a bit crazy. It's ok. Let them do things that you don't think necessarily always make sense for the user, because that not only makes them happy, it's also how sometimes you get the really great things that _do_ make sense after all.

And encourage them to make things configurable, so that the system integrators and distributors can then make it all come together as a more unified whole. Maybe it won't be _totally_ unified, but what you win from allowing people to be people _more_ than makes up for it.

Thinking that developers should also have to be aware or care about the crazy UI HID notion of the day is just stupid. It just alienates developers. And don't tell me that Gnome as a project hasn't alienated a lot of developers, because some of them have been emailing me privately as a result of this flamewar.

Linus
_______________________________________________
Desktop_architects mailing list
Desktop_architects@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/desktop_architects

BusinessWeek's ActiveGrid infomercial

Updated

How does Peter do it? Peter Yared, CEO of ActiveGrid (great company), already has his own infomercial, courtesy of Steve Hamm, in BusinessWeek. The gist? That Java is dead and LAMP is manna from heaven.

Now, don't get me wrong. I'm a LAMP fan and flogged that horse for years at Novell, and continue to do so at Alfresco. We're Java-based, but have an interface that makes it easy to extend Alfresco with scripting languages like the uber-trendy Ruby and P-languages (Perl, Python, PHP). So, I'm fine with wherever the market wants to go. And I think the world of Peter and ActiveGrid.

The problem with Steve's analysis, however, is that the market hasn't gone to the scripting languages. At least, not to Java's detriment. They're the sexiest right now but sex appeal isn't a market. I've yet to talk with a BigCo (and I talk to lots of them now, both with my OSBC hat on and my Alfresco hat on) that is junking Java (if they're a J2EE shop) for the sexy upstarts. I'm sure they exist, but not in the way Steve paints the landscape in his article. Here are a few excerpts:

Peter Yared, CEO of software maker ActiveGrid, spent a critical chapter of his career steeped in Java, the programming language developed by Sun Microsystems (SUNW). In the late 1990s, Yared was chief technology officer of NetDynamics, which pioneered an application server designed to boost the performance of Web sites. It was based squarely on then wildly popular Java. He went on to spend five years as an executive at Sun. So it's especially surprising that Yared holds this view: "Java is a dinosaur."

But Yared has good reason for thinking that way. His two-year-old company sells what he calls a "next generation" application server, used to build Web sites and corporate applications, that doesn't rely on Java. Instead, it's tied to open-source software packages, including the Linux operating system, the Apache Web server, the MySQL database, and a collection of so-called scripting languages that all start with the letter P -- Perl, Python, and PHP. Hence the acronym LAMP....

Yared says developers far and wide are creating a new generation of Internet-based applications with LAMP and related technologies rather than with Java. Can it possibly be that Java -- once the hippest of hip software -- has become a legacy technology, as old and out of style as IBM's (IBM) mainframe computers and SAP's corporate applications? Mounting evidence points to yes.

Reports by Evans Data Corp., which does annual surveys of the activities of software developers, show Java use is slipping as LAMP and Microsoft's (MSFT) .NET technology gain traction. In North America, the percentage of developers who use Java as one of their principal programming languages declined to 47.9 in Evans' fall survey, vs. 51.4% in the fall of 2002. The same surveys show that while Java use is climbing in Asia, it's on the decline in Europe.

Meanwhile, .NET usage increased to 54.1% from 40.3% in the same period in North America, and exceeded Java use in Europe and Asia. In a different survey series, the use of PHP in North America grew to 36.1% this fall, from 26% in the fall of 2003. It grew almost as quickly in Europe and Asia. "There's more competition out there," says Evans President John Andrews. "These other technologies are catching hold. They're biting away at [Java's] share."...

Steve gives a little room to the other side:

Sun is adamant that Java isn't losing momentum. "I vehemently disagree," says John Loiacono, executive vice-president of Sun's software division. "Is Java at the end of life? We think Java is just kicking in." He points to the continued strength of Java as a mainstay of large, complex corporate applications, as well as the popularity of Java in cell phones, where 600 individual models run Java, and seven of the top 10 mobile games are based on the technology.

But his analysis points to the death of Java.

Again, in my experience, this is a gross mischaracterization of what is happening in the market. I like ActiveGrid and think it has a very bright future, but it's not a future that comes at the expense of Java. It's a future that complements Java, if anything. At least, that's what the BigCos are saying with their IT budgets.

Work from home daddying

The December 12 edition of BusinessWeek has an interesting story on telecommuting. Interesting, because it mostly captures the benefits of being able to work some or all of the time from home, and interesting because it largely glosses over the problems attendant to telecommuting.

I've been working from home since 1999, and I'm no longer sure it's a good thing for my family. There is no distinct line between home and office. My kids are used to seeing me, sure. I've always prided myself on being around for the early years of every one of my children, both for those born during my school years and those during the my employment years. When I'm not traveling I tend to work from the kitchen table.

But I'm not sure it's healthy for them to have Dad ignore them, gaping endlessly at my laptop. I don't want my kids growing up to be like me in this way. I don't want them to go on vacation with their laptops and Treos, the way I do. I want them to know how to leave work behind them. Given that one of their two primary role models (me) can't seem to leave email behind, how can I expect them to be better?

Bryce Roberts and I were discussing this the other night. Our respective spouses went out to see a movie together, and we crowded the kids together at my house. While the kids played downstairs, Bryce and I worked on our Macs across the table from each other. On a Saturday night. In the same room. Barely talking to each other.

We are/were, in short, complete losers.

I'd be very interested in learning from anyone how to kick this habit. In part, it's a problem with work - I enjoy work too much. But it's mostly a problem with email - I like to be uber-connected with a lean in-box. As a result, I'm a slave to my email. Because I work from home, my kids have to put up with it.

Any advice?

Talking the Talk

Though I never got into the Boy Scouts program as a youth, I'm now responsible for the 14-15 year-old Scouts in my neighborhood. Surely, someone, somewhere has a sense of humor.

Anyway, I went to one of my Scout's basketball game today, and felt like I was watching two dueling open source companies. (Yes, it's pathetic. I see open source everywhere. This is not something to which anyone should aspire.) My Scout's team was up 46 - 4 when I left. But before I left, I laughed hearing the coach of the losing team. Why? Because he was saying all the right things. "It's a 1-and-1, boys. Box them out!" "Drive to the hoop!" "Work the perimeter!"

But they didn't. They couldn't, probably. They were heavily out-manned.

It's the same with many of the open source companies today. They talk the talk, but can't walk the walk. They still think of open source as a clever marketing slogan or science project, and haven't internalized the most critical lessons of open source success. Namely

1. Product precedes community,
   
2. Community doesn't come cheap (as in beer), and
   
3. Great architecture is essential to both a great product and a great community.

My favorite open source companies are product companies first. They don't try to build community around a dung hill. It's possible, but not the kind of community they want. In short, they understand that open source may well be a better, more efficient way to develop and distribute software, but it's not a substitute for great software.

In short, there really are no shortcuts to building a quality company. Seems obvious, but you'd be surprised at how many pretenders there are currently crowding the open source landscape. (That said, if you're like me, you'd also be surprised at how many really strong open source product/community companies there are. More every day.)

Why open source works (Smith, Dewey, Peirce, Rousseau, and Hobbes)

On a flight from BOS to DCA last week, I happened across a book that had been left in the plane's seat pocket. The book was Proceedings and Addresses of The American Philosophical Association (November 2005, Volume 79, Issue 2), and contained a gem of a presentation given by Philip Kitcher of Columbia University entitled "The Hall of Mirrors."

The lecture/speech addresses the problem of morality. That is, how do egocentric individuals learn to act morally within and against society. Or, to ask the question posed by David Hume, how do we determine when we are motivated by concerns for self versus motivated by concern for others?

To answer this question, Kitcher turns to Adam Smith (of invisible hand fame) and Thomas Dewey (of library card catalogue fame). I find Smith's answer the more appealing, turning as he does to his free market core philosophy to suggest that it is the composite of our individual deficiencies, and deficient perspectives, that makes for True perception.

Software is much the same. Left to our own devices, it will inherit our strengths and weaknesses. And there it will sit. Software becomes more robust, more useful and used, when it becomes an amalgamation of different methods and minds. This is why open source works so well. Because it ferrets out our individual weaknesses by holding us up our code to a "hall of mirrors" where each individual mirror (coder) is subsumed into a larger whole.

Hence, to cite C.S. Peirce, "our job is not to worry about the foundations, grounds, or coherence of what we now have but to treat our current corpus of belief as a starting point from which we can hope to improve" (72). In the software world, quality of code is a process, not an architecture or one-off endeavor. Yes, closing it off does not impede in-house developers from revising and improving the code. But it is precisely the problem of who employs them that stagnates the code: it can never be better than that group of developers happens to be, compounded by the problem of quarterly earnings reports, etc.

Thomas Dewey, speaking of morality, could very well have been speaking of software, then, when he writes of moral decision-making:

The struggle is not between a good which is clear to [a person] and something else which attracts him but which he knows to be wrong. It is between values each of which is an undoubted good in its place but which now get in each other's way. (73)

In other words, life is messy. We're in the middle of Rousseau's "social contract," having thereby emerged from Hobbes state of nature (Nasty, brutish, and short).

Software can be messy like that, too. That's its "state of nature," in the closed, proprietary world where individuals or small clumps of individuals hack out software in solitude, beating up on others who get in their way. Not healthy for the developers or the code.

In open source software, your way of writing code frequently will intersect mine in more positive ways, in well-run communities (like the Linux kernel community). The struggle between "right" and "wrong" happens on mailing lists, rather than between warring companies (though this happens, too), and the code, community, and customers benefit as a result.

Anyway, not news to anyone reading this. But an interesting analogue from the world of philosophy.

Truth better than fiction (Open source in the SME market)

I talked with a great new open source startup today, LucidEra, founded by a few former execs from Salesforce.com, BEA, Siebel, and Oracle, and funded by Bob Lisbonne at Matrix (about time he did an open source deal - he's got more open source blood in him than most VCs, having been the one running Netscape's browser division when they pushed to open source it) and Mark Kremer from Benchmark. Bryce and I were so impressed I immediately gave them one of our most prized slots at OSBC: the CIO pitch panel.

In talking with Ken, he mentioned a fable that I think he believed, and would have no reason to disbelieve: that open source applications are still SME-bound. As I explained to him, however, my experience with Alfresco and SugarCRM (on whose advisory board I sit) has taught me that while the SME market does tend to pick up open source, if his product is any good it will quickly migrate up to the enterprise. Salesforce.com experienced this, too.

As it turns out, the enterprise is really sick of Big Enterprise License Sales, and quickly take to those who can provide a quality product with a smaller upfront investment.

So, by all means, tell the VCs that you're focused on the SME market. Just don't be surprised when the enterprise market proves to the short-term target buyer. It's a very good "problem" to have. :-)

Learning from SugarCRM: The New Face of Community

I just got off the phone with Clint Oram, one of the founders of SugarCRM and currently their VP of Products and Services. That's a somewhat non-descript way of saying what Clint really does: build and leverage SugarCRM's development community. In other words, Chief Community Officer (which, as I've argued elsewhere, is something that every technology company needs.)

Clint said something that set me back on my heels. Of the top 50 contributors to SugarCRM's development community, 95% of them are SugarCRM partners. Think about that.

Some will say, "Of course. That's because SugarCRM doesn't get open source, and can only attract commercial developers."

The less ignorant (yes, the statement above reflects deep-seated ignorance of where open source has been and where it's going - it has always been a commercial phenomenon in the sense of developers solving business problems through technology) will note that it means that the open source development community around SugarCRM is very well aligned with the company. The community wants to make money around SugarCRM and so is going to contribute plug-ins, add-ons, extensions, etc. that make it a robust, superior product.

This is precisely what Microsoft has built over the years: a "keystone" strategy that fosters a benevolent network around itself (Note: I'm not sure I buy Microsoft as a benevolent dictator. Dictator, maybe... :-). The difference in Sugar's case, however, is that the software is open and, hence, the value network is more open, as well.

Where does Clint look for community inspiration? The top 5 open source (web) CMS communities (e.g., PHP-Nukes, Mambo, etc.). These, like SugarCRM, have always been about helping solve business problems (in their case, web content management), and have a strong core of commercial developers surrounding them.

This will be fascinating to watch. Commercial open source projects creating a highly virtuous community constellation of development and revenues.

Where are they now? Open source executive moves, III

More movement recently in the open source business ranks...[UPDATED: Billy Marshall is still the CEO of rPath. It was bad information from *r*c* R*b*r*s (He's so unreliable :-) ] that made me think Billy was stepping aside. Sorry about that, Billy.]

• Stormy Peters, former chair of HP's Open Source Review Board, has moved on (but not far geographically) to OpenLogic, as announced today. She's their new Director of Product Management. Stormy's a fantastic addition to the team there. But will she make the commute down from Fort Collins...? TBD.
  
  
• Mark de Visser, former VP of Marketing for Red Hat, has joined Zend as its Chief Marketing Officer. Another great catch.
  
  
• rPathrPath is getting a new VP of Marketing, Keith Boswell of TogetherSoft past. As one of our OSBC sessions addresses, finding good open source marketers is incredibly difficult. Looks like Zend and rPath are a few that have navigated that difficult chasm with excellent hires.

Windows: The new lightweight?

I'm sitting in on Alfresco's partner training in Boston, and just heard something that got me to thinking....

Paul Holmes-Higgin, our chief scientist, was asked whether the bulk of our downloads are for Windows or Linux (we support either, as well as Mac OS X, NetWare, Unix, etc.). The answer intrigued me. The majority of our downloads are for Windows, but the majority of our installs/customer implementations are for Linux.

It might be (not enough data to support this hunch - I'm just speculating) that developers or system administrators download Windows because it's easiest - they already have it on their desktop, they download the product, they use it, and then they go to Linux when they want to make sure it works for real. Is Windows the new lightweight, toy system?

It's official: open source is righteous

Matt Harrison of Spike sent me this job posting from the LDS (Mormon) Church's website. I'm an active member of the LDS church, and am glad to see that we are officially blessing the open source development process. :-) (Actually, I know that Doug Cutting of Lucene fame has been flown out to SLC for genealogy search work, and that the Church is working on a range of other open source stuff. Time to corral all that open sourceness, I guess....)

I can see it now...

1 In the beginning God created the software.

2 And the software was without form, and void, and was proprietary, causing anguish through lock-in; and closed-source darkness was upon the face of the deep. And the Spirit of God moved upon the face of the waters.

3 And God said, Let there be open source light: and there was light.

4 And God saw the open source light, that it was good: and God divided the light from the closed-source darkness....

:-)

Here's the position from the LDS Church's job site. A good form to follow for anyone looking for an open source development lead?

Lead Software Developer (Open Source) – 0500203
 
Description
The Church is embarking on an effort to leverage the community of LDS developers willing to contribute their time and programming skills to further the mission of the Church. The Church intends to employ someone whose primary responsibility will be to lead this effort and organize the community. This person will develop and document the processes that will govern the interaction with the community and the integration of their contributions into systems maintained and managed by the Church. In addition, this person will be responsible for gathering and understanding user requirements, guiding the creation of an environment in which the community can contribute, communicating needs, organizing teams of programmers, providing design oversight and guidance for each project and coordinating the deployment and life-cycle management of these systems. This person will effectively manage multiple geographically dispersed teams of developers and will be responsible to ensure progress is being made and high-quality systems are being developed and managed.

Qualifications

• Must be a member of the Church of Jesus Christ of Latter-day Saints [Sorry about that]
  
• The candidate will have 5 or more years of professional software development experience. He or she will have an in-depth knowledge of software development methodologies and processes as well as a solid understanding of good programming and design practices. Further, the ideal candidate will be able to apply these to open source projects.
  The candidate will be able to demonstrate significant participation in multiple open source projects. Ideally, the candidate will not only have contributed code, but also taken a leadership role and managed the contributions of others.
  
• The candidate should have strong analytical problem solving skills, with the ability to logically decompose problems and tasks into their elemental parts. He or she should have strong debugging and troubleshooting skills and be capable of identifying project risks and dependencies.
  
• The candidate should have experience with Unix/Linux or Windows environments, distributed systems, database programming, network programming, and/or developing large software systems. He or she should have operational experience and be able to demonstrate an awareness of the operational impact of design decisions.
  
• The candidate should have 5 or more years experience in a commercial software product development environment with demonstrated proficiency in Java (J2EE), Oracle, and web programming. He or she should have a sound understanding of database systems and data modeling. He or she should have experience with version control and change management.
  The candidate should have experience in cross-platform development. He or she should have a strong knowledge of issues related to the internationalization of software systems.
  
• The candidate must be a strong communicator with the ability to interact in a positive way with the community of volunteer developers. He or she should have well-developed oral and written skills. He or she must be able to keep management apprised on progress and potential problems that jeopardize deadlines and milestones.
  
• The candidate must be self-managing and highly motivated. He or she must have the ability to work independently as well as in collaborative team environments. He or she must have the ability to remain positive in the face of challenges and problems and be able to think clearly and logically under pressure. He or she must be able to facilitate issue resolution across geographically distributed teams of developers.
  
• The candidate must have enthusiasm for solving problems and should be passionate about leveraging the power of the LDS developer community to better fulfill the mission of the Church.

I know I'm biased, but that description sounds like a pretty good description for any open source lead developer, church-employed or not.

Google: Developers only need apply

Kudos to Business 2.0 for resolving (for me, anyway) the answer to one of the great questions of the universe:

Why is Google weak in nearly everything new it tries to do?

Yes, its search is manna from heaven. And yes, that's a big deal: I love Google's search, use it hourly, and shudder at the thought of pre-Google life on the Internet. (Of course, I can also remember buying the Internet Yellow Pages, when there just wasn't much out there to search, but....)

But Google stumbles on everything else it tries. Some swear by Gmail but, come on, it's a toy compared to Outlook, and it's not even very good vis-a-vis Zimbra/Oddpost (Yahoo!)/etc. But think beyond Gmail and what do you have? Not much. (Btw, I put Maps and Print under the search rubric - they're cut from the same cloth and serve the same basic ends.)

Anyway, Business 2.0 captures (for me) the reason behind Google's stumblings (as well as open source's, for that matter):

Does Google have some kind of grand strategic plan for the new products it creates?

Virtually everything new seems to come from the 20 percent of their time engineers here are expected to spend on side projects. They certainly don’t come out of the management team....[We ask ourselves:] Are the things that we’re doing consistent with the mission of the company?...[W]e’re in the business of making all the world’s information accessible and useful....

The test that I apply -- and we do this every day, 70/20/10 -- is to ask how a feature will extend the core, the adjacent, or the innovative stuff to fulfill our mission. That’s the sort of drug that we all take, and it works really quite well.

In other words, Google succeeds where developer-driven engineering succeeds, and fails everywhere else. A bit like the open source development community...?

This isn't to diss developers. But just as it would be dumb to build a company with only marketing people so, too, would it be dumb to build a company with only (empowered) engineering types. You really do need both, whatever Google might think. I think the proof is in the product: every Google project besides search (and those things tightly coupled with it, like Maps, though arguably this isn't a commercial success) has failed to the extent and precisely because they put engineering over UI/consumer experience/etc. Turns out, consumers don't like that. So they don't use (Picasa, Gmail (No, they really don't - it has a tiny user base by comparison to its competitors), Froogle, Talk, etc.

Given all the cheerleading Google gets for its trivial forays into everything beyond search/advertising, I'm glad to see that the company really does care about money. Spends 70% of its time making it. Imagine what would happen if it devoted all 100% to that end....Slightly evil, maybe, but comforting if you're a shareholder.

Source no longer matters...

...from a competitive standpoint. I'm talking commercial vendors, not the value of open source. Open source is now big enough that it's not enough to be the "open source widget X" or "open source widget y." You won't get press anymore, and you won't get customer traction this way.

No, it's not a differentiator anymore to be open. The value is reverting to what it's always been: you must be better.

Alfresco was interesting for me because of its superior product: 5-10X faster than Documentum, SharePoint, etc.; easier interface (CIFS so end users can interact with the CMS through a file system, the way people use documents today); and lower cost (1/5 to 1/3 the cost).

Oh, and we're open source, too.

The point is, open source is important to Alfresco because it helps build community around our product and because it is a more efficient distribution mechanism. But without a great product, open source would be wasted on us.

So...if you're thinking of starting an open source company, make sure you haven't forgotten the most essential component: something worth open sourcing.

Sponsorships and community (Conferences)

Matt Harrison (aka "The Real Matt," as my eldest daughter dubbed him - we were neighbors at Stanford) of SpikeSource is in India attending the FOSS.in conference. I'm not there, but Matt is reporting on the sessions on his blog.

On the third day, Andrew Cowle gave a keynote that asked some interesting questions:

Today started off with a bang with a keynote by Andrew Cowie. He did an excellent job speaking and wasn't afraid to ruffle feathers. One of the best parts was his proposal to change Linux to "Open Office/Mozilla/Apache/Perl/Gnu/Linux" (talking RMS's GNU/Linux fixation one step further). He also questioned the lack of sponsorship by Open Source companies such as RedHat and Novell. "Why do they not sponsor community conferences?" I wondered the same thing at OSCON.

I don't work at Novell anymore, and have never worked at Red Hat, but I think I can answer this latter question.

The answer is time/resources/attention. No one has the time to investigate the myriad of small, developer community-led open source conferences to see which are worthy of sponsorship (some are good, some are lame). No one has the resources to sponsor them all. And no one has the attention necessary to find them all.

This time/resources/attention problem applies to the broader open source world (and life in general), as well. Why do we focus in on very few Linux distributions? Because it's too cumbersome to support the myriad options out there. Why do so many IT administrators go with Microsoft/Windows? Because they appreciate the simplification Microsoft does for them. We trust brands to shortcut our decision-making, and (usually) rightly so.

Conferences are no different. Which is why I can happily say that there are very few open source events that sponsors and attendees should not miss. The rest? Well, they serve the "long tail" of the open source community very well, and will continue to do so, with or without Novell's and Red Hat's sponsorship. (Frankly, knowing as I do the sometimes poisonous effect sponsorship can have on an event, I think these community-led events are best left to the community without BigCo sponsorship.)

Christmas gift to Mac users: Handbrake

Handbrake. I think it's one of the best programs I've ever used. What does it do? Rip movies to your hard drive (highly compressed but with great picture quality). From the developer's website:

HandBrake is a GPL'd multiplatform, multithreaded DVD to MPEG-4 ripper/converter. HandBrake was originally available on the BeOS, but now has been ported over to MacOS X and to GNU/Linux. There is no Windows version of HandBrake.

• Supported sources:
  • Any DVD-like source: VIDEO_TS folder, DVD image or real DVD (even encrypted)
    
  • PAL or NTSC
    
  • AC-3, LPCM or MPEG audio tracks
• Outputs:
  • File format: MP4, AVI or OGM
    
  • Video: MPEG-4 or H.264 (1 or 2 passes or constant quantizer encoding)
    
  • Audio: AAC, MP3, Vorbis or AC-3 pass-through (supports encoding of several audio tracks)

Why do I use it? Because I like to have movies on my hard drive when I travel, as it removes the risk that I'll either lose or damage my DVDs on the road. It also means I can take more movies with me, and that I'll use less battery power while watching them. My kids love the fact that I have their entire DVD library on my hard drive.

If you have never used Handbrake, you should. It's mind-numbingly easy to use (which is why even I know how to use it), and works great. My most recent project? Ripping the BBC's excellent adaptation (Nearly 26 hours of it!) of Anthony Trollope's Palliser series. Merry Christmas.

Btw, Handbrake is free (as in beer and as in freedom). However, you can contribute to the developer - Eric Petit - through iTunes/DVDs. Click here for more information. I just sent him an iTunes gift certificate. Excellent work, Eric, and available on the only two OS platforms that matter: Mac OS X and Linux. :-)

Just what is open, anyway?

I had an interesting dinner conversation last night, in which the subject of "openness" was discussed. We were sorting out which open source business models truly count as open source, and which are just along for the ride.

Apparently, the question isn't confined to source code. Wikipedia is struggling with how open its "source" should be, as well, in the wake of Adam Curry's hack job on rival podcasting entries (How could anyone trust the host of Headbanger's Ball, anyway, when he clearly didn't listen to the music? :-) and some gross inaccuracies.

The problem isn't Wikipedia, however. The problem is people. And whenever people are involved - in encyclopedias, source code, or cheese making - things can go very right, or very wrong.

Which brings us back to control. Many in the open source development community have a knee-jerk reaction to control. For them, control automatically equates to "evil." (In the Google sense, whatever that means.) In the open source business community, control can instead mean "profitability." So long as you can achieve profitability while still promoting openness, isn't this a good thing? Shouldn't we, vis-a-vis Wikipedia, be glad that Jimmy Wales is exercising some control in the interest of making the product better for all (in terms of quality), even if it sacrifices some openness?

I think so.

If free software then support revenues (?)

Jonathan Schwartz writes one of the only open source blogs worth reading. The other day he had a particularly insightful post, but I'm still trying to figure out just how far the insight goes. Clearly, free software is a complement to Sun's hardware, and neither Sun's hardware nor software is likely to be used without a support contract.

But would the same hold true for a CRM system? How about ECM? ERP? Probably. But what about email? (Mission imperative - what else would we waste our time on? - but mission critical?) Etc. The point is, how far can we extend this simple proposition that one can open source everything, and give it away for free, because serious customers will always pay for support?

(One of the industry's top open source startups, as I've noted in this blog before, actually has a serious problem with this today, because the product is good enough that customers are not renewing support contracts, if they buy them in the first place.)

Anyway, to Jonathan's blog post. I think he's going to address this topic at the Open Source Business Conference in February. He was worth the price of admission last year. You should plan to listen to him in 2006.

...[M]ost folks, especially some journalists, have [fundamental misunderstandings] about free software in the enterprise - that it somehow implies a sacrifice in revenue....So I thought I'd tell my most representative free software story, and highlight why free software GROWS revenue, not diminishes it.

A while back, nearly three years ago, I was visiting a very fast growing customer. It was not a pleasant meeting - we'd had quality problems in the account, service problems, and a variety of price/performance problems - all of which had made the life of the CIO and CTO squarely miserable.

After talking the customer off the ceiling for an hour or so, the lead executive in the meeting looked at me and said, "we're about to go into our Christmas season. We'll sign up more customers on that one day than in the rest of the year combined." Fair enough. Sounds like a good business opportunity. "So give me your home phone number."

"What?" I said? "Why do you want it?" He had a perfectly logical answer. "Because I'm betting on you. And if I have any issue whatsoever on Christmas Day, I'm calling you at home, and I want you on the hook with me." Ok, fine. I gave him my number. (After I reassured myself I had every one of my staff member's home phone numbers on speed dial.)

Christmas came. He didn't call. Whew.

Just a few weeks ago, I met with this same customer in my office. He's had two years of huge growth (and an account team that's done a stellar job creating a partnership between our two companies).

He'd come in to drop off a big order, and to get caught up. In the process of delivering the order, he'd inquired about a Sun software product he'd begun using. We talked for a while, and said he'd like to include it on the order. To manage his expectations, I let him know it was about to go to free/open source. He looked to his sales rep and said "HA! I'm not paying you $500K for something that's free!" The Sun sales rep looked at me like I'd gone insane.

So I looked at the customer and said, "Oh, there's only one little thing you should know.""If you're using the free product without a support contract, don't bother calling me on Christmas Day. You'll have to look to the community."

The customer paused, smiled, and said "Ok, ok - put it back on the order."

The point being, Sun doesn't have a single customer, worldwide, that will run an unsupported product in their datacenter. Do such customers exist? Surely. They're called developers. Or startups. Or companies or economies that want to build their own internal support teams. That's the target for the Solaris Enterprise System. That's who uses free software without support contracts. And you're not going to win them over if you don't provide them with free and open source products. And if you don't provide them with the technology to use, they'll find someone else's free products.

Jason Matusow moves on...and Bill Hilf steps up

I've been waiting for this announcement, and it has finally come: Jason Matusow is moving on from his current role as Microsoft's Shared Source guru. Jason has been beating the open source pavement for at least three years now (I first bumped into him at the eGOVOS 2003 conference in Washington D.C., and we've been friends since then), and well deserves a break.

Unfortunately, he's not getting one. Neither are we.

Thus spake Jason (my favorite sophist):

I’m pleased with the work done over the past 5 years by a community of hundreds within Microsoft. I’ve been just one piece of that larger puzzle, and now it is time for me to take on a new set of challenges. As of this week I’m taking on a new role as a Director in the Corporate Standards Strategy Team. I’ll be looking at the issues surrounding standards from strategy, policy, and communications points of view.

Just when you thought it was safe to get in the standards water, along comes Jason. If he proves as potent an interlocutor in the standards world as he did in open source, I have a feeling we'll all be worshiping at the .doc/.xls/.ppt "standards" shrine soon enough. :-)

As for his replacement, I don't think Microsoft could have done better. Bill Hilf. Bill, as you likely know, runs the LInux/open source lab at Microsoft, and was the architect behind the Microsoft/JBoss partnership. Importantly he, like Jason before him, is great to be around - someone that will be fun to do battle against.

To both Jason and Bill, I wish you the best of luck. Great people deserve great things.

My holographic will

I'm driving down to Death Valley today to run my first marathon. I've run Envirosports races before, back when we lived in the Bay Area, and they do a great job. I just hope there will be someone available to bury me when I drop dead in the land of desolation.

In case I don't make it back, I leave:

• All the unfinished OSBC program work to Dave Rosenberg, so he and Melinda can hang out together;
  
• My 2-year old daughter, Dread-A, to whomever is foolish enough to take her;
  
• My new job at Alfresco to Novell - here's the SharePoint killer that Novell and everyone else needs; and
  
• My Mac to Stanford University - an object of beauty to study, worship, and emulate. :-)

Zack, will you please arrange for the search party and burial if I don't show up at OSBC? You've done these marathon things before. You'll know what to do....

Downloads and conversion rates

Who has the most downloads? Mozilla FireFox, with MySQL not far behind. (Thanks for pointing out the picture, Zack.) Here's the chart from the WSJ:



Now, keep in mind that this chart doesn't include all major open source projects (the article in which it appeared focuses on Sun) - no Linux there, nor Apache. But I think it's interesting to see who has volume.

More interesting, though, is what companies are doing with that volume. MySQL is on a $40 million run-rate for this year, last time I checked. The big question for them (and for every commercial open source provider) is how to turn downloads into dollars. That is, how to improve conversion rates.

MySQL had some 10 million downloads (give or take 2 million) in 2003. It also had ~$10 million in revenue. That's (very) roughly $1/download. (See Christof Wittig's excellent Stanford GSB paper here (PDF).)

I don't have current figures (though someone with the inclination could easily do this), but $1/download isn't a model that MySQL wants. (Nor is it one that the company has kept, shifting as it did last year to a subscription network rather than a license approach.) The company needs to find ways to improve both its conversion rate and the deal size per conversion.

But this isn't something exclusive to MySQL (just the healthy volume ;-), as I mentioned above. For example, I've been out with Alfresco raising our Series B. A few VCs have asked us to correlate downloads into dollars, and we're so early into the process that it's extremely hard to do. In our few months of existence we've had 70,000+ downloads, a decent percentage of which are turning into paying customers. But the secret will be to improve that number.

If we can, it's mostly profit. An example: I closed my first deal today for Alfresco. Fortune 500 company. They downloaded the product late into an RFP process, when they had already settled on a vendor. They liked what they saw, emailed us, and one month, no visits, and 2-3 phone calls later, they're an Alfresco customer. That is the power of open source. If we (and other open source companies) can replicate that experience for a greater percentage of our downloads, we'll be immensely profitable.

What do you think it will take to boost conversion rates for open source companies, generally? I'd be very interested in feedback.