Download the episode directly, or subscribe to the podcast feed for auto-magic downloading in iTunes or other podcatcher. Also, I realize – but have yet to do anything about – that for this podcast, out of all of them, I should probably have an OGG version and feed.
Stephen O’Grady and I finally embrace out podcasting destiny: to start a (more or less) weekly podcast on the topic of open source. While I note it’s Valentine’s day, Steve notes the much important occurrence that this date marks: the start of baseball.
On more open source topics, we start out discussing Sun’s acquisition this week of VirtualBox. We both agree that we haven’t heard an incredible amount about VirtualBox until now, but that the overall RedMonk community has told us to check it out, evidencing their interest, in the recent past. On the overall note of virtualization, we speculate that there’s an interesting “family tree” of virtualization to be ferreted out: did all this PC virtualization start from the same original clutch of people?
As a to do item for ourselves, we note that we’re not too clear on the proprietary vs. open source license that VirtualBox is available under.
This launches us into a discussion of Sun’s dual-license (I incorrectly start out by calling it “dual source”) approach to open source as account to us at the Sun Analyst Summit last week. Stephen points out that this “Fedora model” of delivering open source software is something that we’ll probably see more and more of. He uses Sun’s Project Indiana as an example of this model in action and an example of the difficulties it can cause with existing communities, such as the OpenSolaris community. In doing so, Steve gives a pretty nice overview of the stink-up around Sun creating an OpenSolaris distro.
Relevant to this, Steve points out that the influence of the MySQL model will probably bleed into Sun. Even more so, I note that it’s my hope that the MySQL folks put together a sort of “The MySQL Way” manual to pass out to all of Sun.
Steve notes that Sun clearly has an affinity for the GPL license, MySQL being so, as well as Java, and the GPLv3’s Sun Open xVM.
I then see if Steve has a snappy phrase for what Tarus Balog have called “shareware open source.” In this model, a company offers a “purely” open source version of their software, but then offers a closed, proprietary “extensions” and features on-top of the open source version as a for-pay only distribution. Steve says that he would just call this the Fedora model.
Moving onto – we hope – non-Sun talk, I ask Steve to go off on what he sees happening the open source acquisition scene. Though we try to avoid Sun-talk, this topic is of course driven by Sun spending $1 billion on MySQL. When it comes to that valuation in question, Steve notes and I agree that if we knew the valuations that other potential MySQL acquires placed on the company, we might better appreciate why Sun ended up paying $1B. That is, we speculate, there must have been several other potential buyers, and Sun offered the highest price, driven by the offers that other put.
Steve then recounts a recent posting on the topic where he outlines the idea that large tech companies like to acquire innovation rather than do it themselves (my phrasing, here, for the idea). That is, better to acquire a company that’s moved he ball forward, innovation wise, than rely on internal innovation at large companies.
We also get into a discussion of open source companies providing network enabled services on-top of their open source stacks as a route to monetization. On that note, Steve prompts me to discuss the Spiceworks model of making money while giving away software. The key to the general idea is to get a huge amount of volume when it comes to your install base and then focus on monetizing a relatively low amount of that large volume. The large the overall number, the more money there is in even “small” conversion rates. Spiceworks follows this model by giving away a free software stack to a very targeted demographic – IT admins at small companies – and then selling that channel of attention to advertisers.
Google, of course, is another version of this high volume, low conversion model in that Google gives away it’s services (largely) for free and collects money through ads on a small percentage of all of the transactions between Google and it’s user-base.
Sun’s model is, as I call it, “the pink dot” model wherein Sun’s looking to get as wide a distribution of Sun IP across the world as possible, and then focus on turning some profitable percentage of those dots “green”: getting cash for them.
In noting another open source “volume embedded platform,” Sleepycat (acquired by Oracle several years ago), we mention that Mike Olson (former CEO of Sleepycat, and then Oracle VP) has recently left Oracle, hoping to take some time off to ski. After the skiing, we both hope to see what exciting adventure Mike’s up to next.
On the note of this volume/pink dot strategy, I note that I’ve often worried about the sustainability of an open source model for smaller companies. Reflecting on discussing I’ve had with people recently, including Tarus of The OpenNMS Group, I note that as long as you’re fine with working a fair amount, it seems like a profit can be squeezed out from that model.
As we note at the end, we’ll hopefully be back next week, shooting for weekly episodes. More importantly, we’ll try to hustle up some guests for you, dear readers.
Disclaimer: Sun is a client, as is Spiceworks. See the RedMonk client list for other RedMonk clients mentioned.
The link to the podcast is missing on this page?
podcast feed has it however, so got it from there…
Thanks for the pointer, I added in the direct link 😉