What I found most interesting about Tuesday’s Sun event, which gathered financial and industry analysts at their offices in midtown for an update, had little to do with the content and nothing whatosever to do with any of the presenters. No disrespect to them, of course – everyone involved did a perfectly credible job. But what Sun had to say about Sun was less compelling to me than the fact that numerous external parties cared.
I’ve been going to these things since, oh, sometime in ’01 I’d guess. In all of that time, I can’t recall a single instance of a third party contacting me prior or during the events for details on what was said. Afterwards, sure. But during? Not one. On Wednesday, three separate people had checked in on the subject by 10 AM. Not reporters looking for scoops, but outsiders curious as to where Sun was and where they were going.
As the very definition of a small sample size, it’s necessary to resist the temptation to read too much into the piqued interest, but still, it’s curious. And to satisy those and the other folks looking for answers, I thought it might be a good time to turn to an old friend – the Q&A.
Q: Before we continue, is there anything to disclose?
A: Yes indeed. Sun is a RedMonk customer, and comped my hotel for this particular event. Travel was on RedMonk, but the AR team of Megan Patterson, Allison Murphy and co were kind enough to arrange transport out to JFK given the two day taxi strike in NYC – hugely appreciated, ladies.
Q: Setting aside your own view of what was interesting, what would you say the intended theme of yesterday’s event was, from Sun’s perspective?
A: In some respects, it was eerily similar to a post I wrote a few weeks ago. The focus of discussion seemed to be, “well, we’ve proven that we can make some money, now we’re focused on how to make even more money.” Much of the ensuing discussion, in other words, centered on growth.
Q: And what was the answer on how to make more money?
A: Increasing volume, primarily. On the hardware front, it would appear they’re looking primarily towards partnerships, and if the goal is volume folks like IBM and Intel are unquestionably good partners to have. There’s room for innovation, of course, with Thumper’s role in the remake of Sun’s long moribund storage business receiving some attention, but it appears that partners will carry most of the load for short and mid term growth.
On the software front, the bulk of the discussion centered around the two largest assets in Java and Solaris.
Java volume is solid both on the server side and on the handset front, and the recent decisions to a.) open source the technology (though the Apache crowd might argue the point) and b.) emphasize Java the platform over Java the language – opening the doors to PHP, Ruby, et al – are likely to have additional benefit in spurring interest.
As for Solaris, the Sun folks tout both current differentiators like DTrace and ZFS as well as the coming improvements of Project Indiana, which, if it can deliver on its promise, is likely to lower the barriers to entry to the technology by a significant margin.
While these explanations largely make sense to me, however, they often seem to be less convincing to the financial analysts.
Q: Why do think that should be?
A: Part of it is just a natural and justified skepticism, I think, for a corporation that bled for a substantial period of years and now must prove that its newly minted recovery is sustainable in the longer term.
But the larger part of the problem, in my view, is that there are some fairly fundamental and intrinsic disconnects between the business of software pre-open source and the business of software post-open source. Disconnects which are anything but Sun specific.
While financial analysts have largely graduated from questions that express a skepticism that open source could ever make money – recent history suggests strongly that they can – they seem frustrated at times by the transition. One can sympathize; if your valuation processes depend heavily on the ability to measure with any degree of accuracy shipments of software, as if software was a tangible, rivalrous asset like a server, life in the post-open source world is more difficult. In the old world, shipments – or downloads, as the case may be – correlated in nice, easy to predict ways with revenues deriving from licensing, sales and support, and so on. In the new world, you probably can’t measure downloads with any degree of accuracy in the first place, and even if you could the relationship between that and purchase is abstract in the extreme. Maybe a customer will pay for the bits they’ve downloaded, but maybe they won’t: this fundamental reality tends to drive numbers oriented people absolutely crazy, because, after all, “if you can’t measure it, you can’t manage it.”
So for all Sun’s efforts to ease their pain by providing insight into the telemetry coming back from installed instances, the revelation that 50% of their software transactions are now taking the form of renewable subscription agreements, and the attempt to paint the big picture, people who only focus on the numbers are and will continue to be unhappy. Until Sun proves it can make “even more” money.
Q: And they do that, they say, via growth?
A: Indeed. And, I think, via the establishment of new, non-traditional revenue streams. Jonathan said in his presentation on the day that Sun builds “systems,” and that “the dividing lines between traditional industries” is going away. I happen to agree, and there’s more proof of this trend every day. Consider virtualization, for example, which is blurring all sorts of lines between hardware types, hardware instances, hardware physical locations, not to mention operating system, application, and workload.
From a Sun perspective, however, I’m most interested to see how quickly and effectively they can build systems that guarantee them network services revenue – money for something other than simply answering the phone when something goes wrong.
What do I mean by that? Consider the fledgling Landscape and Network Monitoring and Advisory services offered by Canonical and MySQL, respectively, in their efforts to make “even more” money. Then stop and consider the assets that Sun has at its disposal: community (check), hardware-as-a-service (via its Joyent partnership), operating system (Solaris), connection/telemetry infrastructure (Sun Connect), services capabilities (Sun Services). Can those be combined; collapsed into a “system” that goes far beyond just bits on a disk, one that begins to inject people and community into the equation? Obviously, I think so. The pieces they don’t have, after all, they could get from Amazon. And I see hints that Sun thinks the same, though they’re perhaps not quite as aggressive on the subject as I’d hope for.
Q: Speaking of virtualization, any discussion of the subject at the event?
A: Tons. Solaris Containers, Brand Z, LDOMs, the xVM project, Project eTude, all were mentioned and play a role. There’s also a forthcoming product, referred to in the session as Project Virginia, that will target the space. It’s a complicated enough picture, however, that I should probably table the subject for another post and simply say that Sun appears to get fully that virtualization is both here to stay and poised to overhaul the datacenter.
Q: How will their strategy play out in an increasingly competitive landscape, though?
A: My answer to that would be what I’m telling people who ask us about virtualization these days: pretty much everyone agrees that virtualization will change everything, it’s just that no one’s quite sure exactly how. The better question, until the clouds dissipate a bit, is to ask whether or not Sun will be relevant to the conversation, and my answer to that would be an unqualified yes.
Q: You also mentioned ZFS, the new Solaris filesystem. Was there any mention or discussion of the patent litigation initiated by NetApp to attack it?
A: Nope. The news hit that day, and Jonathan wasn’t even aware of it until afterwards.
Q: What is your reaction to the news?
A: I’ve long been of the opinion that the patent system in the United States as it pertains to software is both broken and beyond repair, although on the latter point I’m not entirely opposed to efforts to try. This ZFS suit has done nothing but reinforce those views. Based on the he said/he said byplay, it’s impossible for me to say precisely where the truth lies, but I am likely to condemn IP litigation regardless of the circumstances. As do many others, I find the choice of venue suspect, and will unfortunately be required to track the suit as it progresses, but apart from that I’m just like you: we’ll have to see how it plays out.
Q: What about the reverse stock split – was that mentioned?
A: It was, with drivers like “perception” cited in justification. Personally, having grown up in family heavy in financial services workers, I pay very little attention to share price machinations and understand even less about their longer term implications.
Q: How about the change in the ticker symbol, from SUNW to JAVA?
A: Yes, that was discussed briefly as well. The justifications here were fairly simple – a.) Java’s a more widely known brand than Sun and b.) Sun’s not a workstation station company anymore – and there was little additional commentary on the subject. So while I personally didn’t understand the decision, it impacts me not at all and seems to have bothered the financial folks even less.
Q: Anything else to add?
A: No, I think those were the highlights. If you have questions I didn’t get to, feel free to send them in and I’ll try and answer them or maybe do an Ask RedMonk.