5 Things I Learned from the MySQL Conference

In an event short on bottled water but long on reliable wireless, MySQL proved much to this first time User Conference attendee. Not that I came in as a disbeliever, of course: even before they (smartly 😉 became RedMonk clients, MySQL impressed me as a project that possessed that rare awareness not only of where it was but where it could get to. The past three days have nothing but reinforce that assessment; it’s tough to disbelieve, after all, when all you hear from its developers, partners, hell, even competitors, is that the product is, to borrow Dan Patrick’s words, en fuego.

A quick aside on the tone of the show. As is typical of conferences with an open source flavor, the days here have had a decidedly informal, conversational flavor (you get three guesses as to whether I prefer that over the formal alternatives, and the first two don’t count). The fact that I can get away with jeans doesn’t hurt, admittedly, but the tone was set on day 1 when I heard a “Steve” en route to the press room, and discovered Jeff lounging on one of the oversized chairs common to the lounge at the Hyatt. Given that the man better known as jdub to #redmonk regulars lives more or less on the opposite side of the planet, it was excellent to catch up
on everything from GMAE (his current baby and a fascinating project) to Waugh Partners (his and Pia‘s day job).

That is what conferences should be for, in my opinion. Cisco’s HD telepresence solution may one day make scheduled face to face meetings unnecessary, but it’s not going to kill the need for conference travel – much as I’d like it to (travel gets old quickly, just trust me on this). You just can’t match the serendipity offered by conferences; the interactions that are directly dependent on proximity.

There are the conversations with the folks you “know” but have never met – i.e. Ben Rockwood, Lisa Sheeran, Mark Atwood, Paola Lubet, and Steve Curry, in this case. There are the folks you “know” from from past interactions, like Dave Gynn, Jason Hoffman (thanks to Jason for both a.) organizing last night’s event and b.) picking up the tab), Marten Mickos, Matt Mullenweg, Zack Urlocker, et al. And there are the dozens of casual interactions with folks you don’t know from a hole in the wall and would be unlikely to meet outside the context of this shared interest, this shared event.

Ok, so maybe the aside wasn’t that brief, but you can’t honestly tell me you’re not acclimated to the extreme lack of brevity in this space. So with no further delay, the five things you need to know about the MySQL User’s Conference.

The Competitors Recognize the Opportunity and Threat

Don’t get me wrong, I’m in no way contending that the IBM/MySQL deal to bring MySQL to System I represents some strategic decommitment to DB2. It’s no more that than the embrace of PHP was a strategic departure from Java. But all that said, if you think this was an easy and simple decision for IBM, you’re fooling yourself. DB2, like Oracle and to a lesser extent SQL Server is still struggling with the unique equation that simplicity and open source and low cost that MySQL represents, so supporting that – despite the fact that DB2 for System I is built out of the Rochester facility unlike its namesake – is a non-trivial decision. Credit IBM here for being willing, just as it was in acquiring Gluecode, to look beyond the obvious threat to see the opportunity. As I told Eric, I think this is a potentially excellent deal for both sides: MySQL gets access to the perennially underappreciated System I channel, and DB2 gets access to the applications that run on PHP (via the earlier deal with Zend) and MySQL.

The Developers Love the Product

Anyone who’s read through an O’Reilly, Pragmatic Programmer, or related text lately knows this drill: MySQL is far and away the choice of choice amongst developers. Praised alternately for its simplicity, application compatibility or merely ease of acquisition, MySQL is absurdly popular. Even amongst the admittedly self-selecting population of developers at a conference dedicated to the product in question, the affection heard from the product – and more importantly its creators – was something to behold. Open source purists may occasionally take exception to MySQL’s chosen business model – dual licensing – but as nearly as I can determine, developers don’t care. Even the potentially controversial fork (my words, not theirs) between Community and Enterprise seems to have had little impact; certainly it’s doing nothing to throttle interest and better adoption.

The Ecosystem is Bigger Than You Think

It’s becoming increasingly well understood that it’s very difficult, if not impossible, to accurately measure traction and adoption of open source products. MySQL’s no exception to this rule; as I’m fond of telling reporters seeking accurate metrics, we run at least 8 instances of MySQL internally for a variety purposes – and MySQL has no way to know about any of them. Possessed of the mindset that only what can be measured can be managed – as HP’s Mark Hurd apparently is – many will cast about desperately for some metric to measure momentum, settling eventually on ISV traction. This conference, interestingly, seemingly belies the actual traction in that arena. If one were to merely visit the conference, it would be easy to conclude from the show floor that MySQL is a vendor of middling traction – the exhibitors were far from legion. But take a walk around – nearly every one of the attendees (1400+ registered) is doing something with MySQL, and many are building on top of it. The database isn’t the developer’s choice for no reason.

The Plan is Not to be Acquired

Like Web 2.0 firms, the exit strategy for a great many open source vendor most likely lies in acquisition – trust us, we have those conversations all the time. And if you can’t trust us, just ask Winston Damarillo. MySQL, however, appears to be the rare exception to that rule. Although I didn’t get the opportunity to chat with Marten on the topic, more than a few reporters confessed that he’d mentioned an IPO as the logical next step for the firm. Computerworld’s Eric Lai and I, in fact, chatted on just this subject; my response to whether or not this would invariably change the way they do business was “not necessarily.” Wall Street, after all, primarily demands growth, and having achieved (or nearly so) the goal of virtually every open source project – ubiquity – MySQL’s well poised to deliver decent growth. Just look at the success that its enterprise licensing deal achieved. ESPN and the New York Times were early wins – how hard do you think it will be to upsell them from the Gold level of support to something more financially rewarding.

The Product is the Firefox of its Market

About a year and a half ago, I challenged vendors to answer how they were “the Firefox” of their respective markets. If there’s one thing that MySQL definitively answered at this show, it’s that question. If you don’t believe that, read Farhan Mashraqi’s summary of Mark Atwood’s presentation detailing the creation of his S3 Storage Engine. Or go straight to the horse’s mouth, as it were, and get the presentation and presentation notes themselves over here (ODF fans, rejoince, Mark’s got your back). If you’re link averse, here’s the short version: Mark created a storage engine for Amazon’s S3 service. MySQL’s pluggable storage engine mechanism was eloquently described by mark as a “Protocol Translator” – a piece of software that will translate from one protocol/schema/model to MySQL’s way of looking at the world. The importance here, in case it’s not already obvious, is this: rather than try and solve every known database problem, MySQL can rely on the community to build solutions for it. Need better and faster text search, as at least one very notable Web 2.0 developer has pined for? Check out Sphinx. It’s almost amusing at this point to look back on the Fear, Uncertainty and Doubt that surrounded the acquisition of InnoDB by Oracle (was hugely flattered, incidentally, to have some of my analyses called “bright” by one of the InnoDB developers).