Posts
Oracle bought Sun. Here are my predictions/suggestions.
HARDWARE
What I Think Should Happen: I think Oracle should spin off and sell the server division, maybe to HP. Oracle's never been a hardware company (similar to Sun really never have been a software company, more on that later) and they probably will just muck it up. On the storage side however, Oracle has a great play. Between Sun's StorageTek acquisitions and Thumper (it's storage array) Oracle has an awesome opportunity to combine it's industry standard DB marketing power with a hardware storage play. They can/will do much better than Sun did at pushing this product line. They should keep it.
What You Should Do: If you like Sun hardware, continue to buy it. It's valuable, has it's place in the Enterprise and someone will eventually support/move it forward, even if it's not ORCL.
SOFTWARE
HARDWARE
What I Think Should Happen: I think Oracle should spin off and sell the server division, maybe to HP. Oracle's never been a hardware company (similar to Sun really never have been a software company, more on that later) and they probably will just muck it up. On the storage side however, Oracle has a great play. Between Sun's StorageTek acquisitions and Thumper (it's storage array) Oracle has an awesome opportunity to combine it's industry standard DB marketing power with a hardware storage play. They can/will do much better than Sun did at pushing this product line. They should keep it.
What You Should Do: If you like Sun hardware, continue to buy it. It's valuable, has it's place in the Enterprise and someone will eventually support/move it forward, even if it's not ORCL.
SOFTWARE
What I Think Should Happen: Larry Ellison said there were two software products that were instrumental in Oracle's decision to acquire Sun: Solaris and Java. I believe it. Here's how I think the overall software play should break out:
Solaris - It's a great OS, continue investing heavily. Oracle not being as OSS friendly may kill Open Solaris (or let it flounder). This would be a bad move politically though, so I would hope the project will continue.
Java -This will move forward (duh). The community essentially owns Java at this point anyway. If Oracle continues to do a good job facilitating it's evolution, which arguably Sun did well, they will gain points and appear a bit "softer". Having said that, .NET's evolution is IMO eclipsing Java's due to the focus of a single vendor, MSFT, at the helm (unlike Java having to appease multiple stakeholders). Oracle's business saviness may be what Java needs to focus the JCP to accelerate it's maturation. Right now it's definitely floundering in some areas (see EJB).
Database - Oracle wants MySQL dead, but there's no way they'll kill it and they'll probably figure out a better way to commercialize it. I think the project will fork, since a lot of the MySQL community probably doesn't trust ORCL. This will ultimately be a good thing, because I think we'll get another great OS DB alternative in addition to MySQL and PostGres.
Enterprise Software - There's a lot of it so here's a rundown.
OK, that should stir up enough controversy. Now tell me why I'm crazy in the comments below (or silently agree with me :-). And if you want to chat more on this, you can always contact me.
Solaris - It's a great OS, continue investing heavily. Oracle not being as OSS friendly may kill Open Solaris (or let it flounder). This would be a bad move politically though, so I would hope the project will continue.
Java -This will move forward (duh). The community essentially owns Java at this point anyway. If Oracle continues to do a good job facilitating it's evolution, which arguably Sun did well, they will gain points and appear a bit "softer". Having said that, .NET's evolution is IMO eclipsing Java's due to the focus of a single vendor, MSFT, at the helm (unlike Java having to appease multiple stakeholders). Oracle's business saviness may be what Java needs to focus the JCP to accelerate it's maturation. Right now it's definitely floundering in some areas (see EJB).
Database - Oracle wants MySQL dead, but there's no way they'll kill it and they'll probably figure out a better way to commercialize it. I think the project will fork, since a lot of the MySQL community probably doesn't trust ORCL. This will ultimately be a good thing, because I think we'll get another great OS DB alternative in addition to MySQL and PostGres.
Enterprise Software - There's a lot of it so here's a rundown.
- Sun's Portal R.I.P. it sucked anyway and doesn't hold a candle to Weblogic's.
- Sun JCAPS R.I.P. it's convuluted and Oracle's ESB/BPM plays are stronger.
- Sun Glassfish R.I.P. if Oracle was smart they'd release a community edition of Weblogic in it's place
- Sun Netbeans. Please kill it already. And kill JDeveloper while you're at it. Standardize on Eclipse as THE Java EE IDE platform and invest like crazy into it so the Java development environment can start holding a candle to .NETs (Visual Studio). Ask any developer who's played on both sides of the fence which is slicker (and for those of you crying for competition, Visual Studio IS the competition).
- Security Software: Sun Identity Manager will most likely merge with Oracle Identity Manager. Both have there strengths. Same goes with their Role Manager products. Sun Access Manager - R.I.P. Oracle Access Manager has a couple years on it.
- Sun LDAP - lives on. It's more hardened than Oracles (which really is a directory fascade on top of their RDBMS anyway).
- Sun iPlanet - who is still using this? OK, I know who is, why?
- What else am I forgetting?
OK, that should stir up enough controversy. Now tell me why I'm crazy in the comments below (or silently agree with me :-). And if you want to chat more on this, you can always contact me.
2 comments:
A provocative and insightful post, and an equally provocative merger.
I do take issue with the statement "Sun is not a software company". While this is 100% true from a revenue standpoint, the future of Sun's hardware and storage business was what had them trading at less than $5 for most of the last 6 months.
StorageTek, Solaris, and UltraSPARC are 2009's Gateway, Friendster, and SGI. All run below a distant 3rd in their respective markets. Sun has its niche in financials and telecom, and I don't see either of those sectors as big tech buyers in 2009/2010.
Sun actually has some compelling virtualization plays from desktop (VirtualBox) to cloud (xVM) that you didn't mention. Between that and Java alone, I would wager software makes up at least half of Larry's $5B check for Sun.
Two of the most promising contributors to future value, Java and MySQL are in my opinion officially DOA as of 4/20/09. Of course Java, like AIG, is too big to die, so the OpenSource community will undoubtedly move toward open containers and JVMs. Oracle will continue to ship a leading Java container and JVM with Weblogic, but let's forget about leadership of the Java community. Hooray Scala!?
Having purchased several million USD of both Oracle and BEA software, I can agree the salespeople for those organizations were well-aligned, and they probably will cherry-pick some value from the ranks of the old SUNW sales squad. Good luck to any of those guys trying to connect with the majority of CTO's and developers who have embraced Java.
Unfortunately, I attribute about $1B in value to the old SUNW, and about $4B to the software portfolio. I'm not sure how you could have come to the conclusion that Sun isn't a software company.
My take (please forgive the length of this monologue, since this is your blog. Heh, you did ask for thoughts, and FWIW I agree with most of what you wrote :)
Solaris is (still) dead. But for a few niches, this has been true since 2001.
UltraSPARC? Are we kidding? I've got an old DEC Alpha machine I can part with if anyone's interested. Itanium anyone?
MySQL. Sales = inconsequential, deployment = #1. Oracle isn't the company to figure out how to monetize that product. It's dead. Huzzah Postgres and MyOpenSQL!
As of yesterday, Java is dead as a viable commercial platform. Commercial JVMs will continue to diversify, as will Java containers. Fragmentation will decimate the market (ala Linux). Hooray Microsoft and .NET! I think you're 110% right on that. If anyone profits from Java, I predict it will not be Oracle/BEA. It will be another large, more innovative company that gets behind the successor to Sun's Java community.
Fun diversion
----------
1. Name a materially relevant open-source Oracle project.
2. Visit oss.oracle.com
3. Name a materially relevant open-source Oracle project.
4. GOTO step 2
Sun had many of the best ideas ever. Sun had many of the brightest people ever. How they eventually lost out to HP and IBM, it's a tragedy. I personally think it was Scott McNealy's reluctance to give up his chip. IBM knew when to drop the PowerPC (my apologies to the pour souls still purchasing AS/400 hardware).
Even intelligent people still don't seem to realize that innovation always happens at the fringes of enterprise. It is the CTO to look toward for future trends, the CIO is an highly-astute businessperson who often knows the optimum time to investigate a new market trend, but rarely the source of innovation.
Oracle, in much the same way, will not continue to foster innovation (have they ever?) Oracle's genesis is the outcome of a properly timed launch of a product derived from other people's research. Right place + right time + good sales and marketing. Sun's innovative culture (and the source of it's value) will fit like a square peg in the round O of Mr. Ellison's outfit.
Larry Ellison is a cancer on innovation. Bold statement, perhaps, but since he left his post on the board of AAPL (2002) it's up 1100%. He's Charles Wang with a slightly better track record (for now).
Winners: MSFT!, IBM, HP, Apple, Postgres, Scala, JBOSS, Apache, any other RDBMS vendor, OpenSource.
Losers: ORCL, JAVA, Commercial Java, MySQL, OpenOffice, VirtualBox, Innovation.
Meh, Sun unfortunately became as irrelevant as SGI. Both companies had people far ahead of the curve in terms of innovation, but far less skilled in building a long-term brand. If Steve Jobs could ever figure out enterprise computing like he's figured out personal/mobile computing, he'd eat all of these guys for lunch.
Jeezus, I hope this isn't longer than your original post, but I guess you could say you touched a nerve. Very insightful and interesting post -- Thanks J.
David, great comment.
What I meant by Sun not being a software company was not about the software it owned, but how it managed the software it owned. They made some GREAT acquisitions over the years (Waveset's Lighthouse, Seebeyond's JCAPS, Vaau's Role Manager). This gave them some great IP and even better people. But every acquisition came with huge issues around support, services and maintenance of the product. I've been on the delivery end of this one too many times. Arguably ORCL has had similar issues but they eventually figure it out. Sun never seemed too. So although they had great software, they were never a great software company. I think ORCL can help change that.
Thanks for your insights, and for reminding me about their virtualization plays. Great points!
Post a Comment