SUNW last week announced that it would open source ALL OF ITS SOFTWARE and pursue a subscription business model à la JBoss. WOW.
A good article that captures most of it is Information's week Charles Babcock's Sun Bets On Open-Source Middleware For Turnaround
I was in Munich at the time of the announcement drinking beers at night with Sacha Labourey and Michel Goosens as we were doing a European press tour (London, Munich, Paris) and we dialed into the conference call. I was trying to put a question to Jonathan Scwhartz and John Loicano to congratulate them on the bold move and ask them how they hell they were intending to pull it all off. Needless to say I was filtered out of the call.
As Charles noticed Jonathan Schwartz, SUN's COO, must have mentioned JBoss 10 times as the proof point of why the business model was working. Essentially, this is a 180 degree turn from statements two years ago when Jonathan was also announcing that SUN would be the new dominant PROPRIETARY software player. Now, Jonathan has seen the light and decides to open source everything, mentioning JBoss as a proof point of the business model of FOSS and subscription. SUN's turnaround is going to be bet on the JBoss business model?
Just a little point of correction for the SUNW marketing team and the press. As the Wall Street Journal picked up, Sun's appserver SES with it's claimed 3.5 M downloads is NOT THE VOLUME LEADER. JBoss has 6 M on AS alone (as verified by Sourceforge) and north of 8M downloads on the JEMS suite, and that is not even including Tomcat. So, please correct your slides.
But the numbers are details. On a grander scale, betting the future of SUNW on FOSS and subscription is a bold and visionary move by Scwhartz and his team. The writing was on the wall and I predicted something like this would happen back when IBM announced their acquisition of Gluecode, specifically that it would "force Sun’s hand". However I didn't expect them to answer this quickly and this strongly. They have met IBM’s challenge and raised the bar.
The last time Sun open sourced something I wrote "Sun open sources XYZ, who cares?" but this time around there is a significant dimension that I want to focus on. Just as our acquisition of the Arjuna/HP transactional technology represents a significant acceleration of the timeline at which we open source technology, SUNW's OSS'ing of SeeBeyond is an explosion in the integration sky. We saw it, and cheered. I am sure IBM took notice, but I am not sure they are cheering.
All of the sudden, integration and distributed transactions, two technologies that were supposed to hit FOSS maturity in the year 2030 timeframe are available now, mature now, stable now in FOSS. SeeBeyond was a solid if not overly exciting tier 2 integration play, but nonetheless I claim this represents a steep escalation of violence against IBM’s reluctant FOSS strategy.
Both moves go directly to the heart of IBM's software business in Websphere. Half of their license business is driven by integration and high-end transactions. That is not even counting the Global Services revenue that comes from integration because Websphere is so complex. The acceleration of the timeline must be spinning heads in Armonk. A trap door just opened under their feet. Fear and loathing of FOSS must be at an all time high over there. "What did we do with Linux? We wanted linux only... Not ALL of software, ohmygod!".
I don't know why SUNW is choosing to market this as a move against MSFT. Being an ex SUNW employee I know that the anti-microsoft crack is pervasive in their ranks but come on guys, MSFT is laughing their heads off at this move. We are eating each other up right now, throwing bombs over trenches. I believe it is a marketing mistake to target MSFT with this because they don't care. In fact, I am sure they applaud the short term implications of this move...
While we are on the topic of botched marketing communication, I really think SUNW should drop the "Solaris link" with SES. It is a mistake. People will use it in other platforms, remember. Our own user base at JBoss is 50% WINDOWS. You guys invented Java for cross platform purposes. I hear in the market that it is pretty good :) why you decide to go against that technical fact is beyond me. Selling hardware nowadays is done on hardware merits alone, due to the dominance of virtual machines. You either compete on metal or you don't.
Let's talk about the long run. Jonathan, if you read this, congrats again on this move, but seriously you have nowhere to run now. You need to succeed. Fail and Wall Street will open source you (as the comments on yahoo financial boards called for in the wake of these announcements); succeed and you will be hailed as The Man.
Sun has a poor track record of execution in middleware. Quality of implementation has always been sub par and customer satisfaction is a sore problem. SeeBeyond has its own legacy bloat to deal with. Solaris is a shining star in your line up, but face it JES has been an embarrassing failure so far. Even after being open sourced, it still fights for 8th place marketshare in the low-end. What next? You already OSS'd it, are you going to do like IBM and give away plasma TV's for people to download your middleware? Free and OSS doesn't cut it nowaday, it is not the silver bullet some people think it is. JBoss succeeded because the SOFTWARE AND THE SUPPORT are world class.
Charlie Babcok’s analysis is right on the money, the devil is in the details. Can Sun open source everything as they claimend? On what timeline? Through our partner network that has OEM relations with SeeBeyond, we know FOR A FACT that Sun CANNOT open source certain parts of that stack. Already Sun is back-tracking in the press as to the timeline, oy! Come on, Sun, execute well on this one. This is your last chance, but it’s your best shot yet. I would hate to see them end up with the short straw on this one.
As far as competition is concerned this is further fragementation of the number 2 open source middleware contenders. SES will be fighting Jonas, Geronogo for distant second place in open source and maybe 5 or 6 th across the board, in the market. ONJava survey puts JBoss at a 40% market share as of a couple of weeks ago and IBM and BEA are at 20% each. How in the world is anyone going to encroach on our respective positions? For what reason? For sub-par software and support? JBoss ranks first in customer service satisfaction, which is normal since this is the only place we make money. Hmmmm, anyone considering migration would be taking a step backwards on the software and support front. SUNW would be wiser to work with us, as opposed to fighting us.
But don't let me downplay the Sun announcement. It says "the future is now" and "the future of the business model is Professional Open Source". This represents a tremendous acceleration of the timeline of OSS: buying SeeBeyond for $400M and then OSS'ing it was a bold move. Just as IBM forced Sun's hand 6 month ago, Sun just forced the hand of ALL THE BIG PLAYERS NOW. The dominos are starting to fall and JBoss sits in the sweet spot , in a war of the worlds that involve the big boys, JBoss represents a premium brand of volume OSS distribution. Everyone has to react to an open source stack, again, I got made fun of the last time for saying this but where will it end? The future is now and I congratulate SUN for calling the end game on the business models and putting its software where its vision is, will it be enough for SUN to pull it of? Probably not, but we will see,
I am starting to seriously enjoy this,