Last week I got to represent Zimbra at a dinner of open source luminaries that coincided with OSBC West. At one point, I was asked what was new in Zimbra-land, and I got to mention that we’d rolled out our GA in early February, and (I confess I made a bit of a dramatic pause here) we were also rolling out our first 100,000 seat enterprise deployment. (In service provider land, Zimbra is now ramping up seven figure deployments.) This garnered Zimbra warm congratulations, but also some incredulity. “After all, you’d only publicly launched Zimbra 4 months ago, and you’ve already got a 100,000 seat enterprise customer ramping up?”
What I did say was that Zimbra team had been hard at work since 03, and that the technology had been in private customer trials since early 05. Weak. What I should have said was …
… that Zimbra first and foremost owed a tremendous debt to the open source technologies that enabled the Zimbra Community to get so far so fast.
I believe Zimbra makes heavier use of open source software than any messaging solution to date: Linux file system (message store), Apache (server container), MySQL (meta-data), Lucene (search), Postfix (MTA), OpenLDAP (config. data), SpamAssassin (anti-spam), ClamAV (anti-virus), and so on. (Click here for a comprehensive listing.) Without all of that sweat and toil from our brethren, you would not have heard of Zimbra yet, let alone be reading about its large-scale deployment.
Eric Hahn, who really has forgotten more about messaging than I’m ever going to learn, pointed out at our last board meeting that Zimbra actually had substantially fewer defects than he would have expected at this stage of its lifecycle. (No doubt, our Zimbra community members would still like to see fewer bugs; you can track our progress here.) While we have invested in sophisticated automated testing infrastructure, and Zimbra is now running some heavy messaging loads around the world (both essential for hardening), I think the (relatively) low defect count is again reflective of Zimbra being built with such solid open source building blocks.
As I compare Zimbra’s progress to that of WebLogic—the other server I’m intimately familiar with, I am also struck by the role that the Web technical community now plays in accelerating technology validation: the tech. bloggers, Slashdot, Digg, O’Reilly, and so on. Compelling innovation appears to be able to percolate to the top much more quickly thanks to such insightful peer review being universally available. I think the good news for the geeks is that technology counts for more in this world—great marketing and sales strategies cannot compensate even in the short term for a lack of innovation and quality.
Just my $.02.
on February 20th, 2006 at 12:25 pm
Congrats to Zimbra! I recently wrote an article on my blog about your app and the AJAXTK — my point in mentioning this here is because I appreciate your company’s culture embracing and supporting the world of developers. It’s evident throughout your company’s site. Awesome and inspiring stuff you guys have accomplished. Please keep pushing forward.
I wanted to comment that your last sentence about innovation and quality rings very true. While billions in marketing dollars can compensate for inferior technology in the short term, the best solution wins in the long run more times than not.
Thanks,
Mark Holton
on April 5th, 2006 at 8:46 pm
> I think the good news for the geeks is that technology counts for more in this world—great marketing and sales strategies cannot compensate even in the short term for a lack of innovation and quality.
Really?
Your conclusion, particularly regarding Zimbra’s situation, seems like a stretch. Zimbra made a significant market impact simply by “embracing open source” (vigorous hand waves). I don’t see how you can compartmentalize, and discuss how much goodness came from quality in the source code and how much from the OSS feel-good vibe. In other words, how much from OSS, the technology, and how much from OSS, the sales and marketing strategy.
But getting back to the point you made…
Look at IBM and BEA - that pair is the quintessential counterexample to the argument that technical excellence trumps business execution. No-one holds IBM on a particularly high pedestal for excellence in software (though they are none too shabby). Yet Websphere is unquestionably a much more successful product in the marketplace than WebLogic today. Has excellence in technology prevailed 0over business execution?
“Technology matters more than business execution” - it’s a nice dream for dweebs but has never been true.