Shaping Technology Strategy

CTO Journal

Subscribe to CTO Journal: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get CTO Journal: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn

CTO Authors: David H Deans, AppDynamics Blog, Louis Nauges, William Schmarzo, Shelly Palmer


i-Technology Viewpoint: OpenSolaris, Get It While It's Hot!

Is this move going to bolster Sun, and turn around this excellent piece of engineering?

Go ahead, download the source code for Sun's amazing operating system, Solaris. It's now been released. I just downloaded it in under 5 minutes using BitTorrent. Why? I have no idea. Because I could I guess. Or because I didn't believe it would ever happen. I used to work at Sun at one point, so I have some appreciation for what it took politically, not technically, to get to this point. Congratulations, Sun.

The code to Solaris is now available to everyone. Why would Sun give up such an asset and is it too late for Solaris? Is this move going to bolster Sun, and turn around this excellent piece of engineering? It's hard for me to imagine Linux or BSD kernel hackers leaving behind their beloved code to poke around in Solaris' guts. In fact, wouldn't that make it legally hard for them to return to the BSD or Linux code bases? Aren't there IP issues involved? Or is OpenSolaris open enough that it's okay to study the competitor's implementation before designing your own? I don't know enough about their license, and there are no precedents from which to guess at an answer. My guess is that Sun has learned that it needs to let Solaris grow and influence the community, that type of stewardship will help their OS and Java businesses (such as they are) and potentially their hardware business at some point.

So I ask, "who cares?" Not because there are viable alternatives such as Linux and BSD, but because the industry is at an inflection point and this is only one of the signs of the time, not some amazing breakthrough. The release of Solaris as open source is really more about recognizing the shift toward a new reality more than anything else.

If you ask me, its now clear that UNIX has won the Von Neumann architecture award. Although it had its birth on a time-sharing machine, it's really this architecture that allowed UNIX to grow and thrive and eventually win the hearts and minds of the average computer user. Neither of these pillars of modern day computing is going away, neither is the language that gave birth to UNIX, ANSI C. However, with the release of Solaris as open source Sun is the last to admit that UNIX is a commodity, and possibly a dead end for future architectures.

Consider the latest news from our chip manufacturers. More and more we read about hyper-threading, multi-core, large numbers of specialized parallel vector units (such as in the Cell processor or the G5's Altivec), and just recently the advent of CMT (Chip Multi-Threading) techniques by Sun. I believe that Sun has realized that it has lost the advantage of RISC, UNIX, and that Java might be next. These are the foundational components of Sun's existence, if they don't find new bedrock they will die. They know this better than anyone.

Tomorrow's systems will require new languages, to complement these new systems. Today's programmers do not have the tools or the skills required to build systems able to scale with such new hardware architectures. Concurrency will be the dominant buzzword for the next 10 years. Sun knows this too. It's why Guy Steele is working on Fortress, the follow up to Fortran. Other languages like Erlang will influence new as yet unimaginable programing methods for managing concurrency much as Smalltalk led to the object-oriented revolution culminating in C++ and Java.

What is clear from this shift toward concurrency is that just as the C language led to UNIX, some other language best suited for these new architectures will lead to a new operating system design.

UNIX, ANSI C, Von Neumann, and even the notion of threads will all soon be history. Software will simply be concurrent, and somehow we'll be able to deal with that.

More Stories By Gregory Burd

Gregory Burd is the Product Manager for Sleepycat Software, now a part of Oracle. Prior to Sleepycat, he was on the business team at KnowNow, a Kleiner Perkins startup in the San Francisco Bay Area. He has many years of software development and product leadership within companies such as JavaSoft, a division of Sun Microsystems, Marble Associates, a consulting company, and NeXT Computer, now part of Apple Computer.

Comments (1)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.