By ColmSmyth on Mar 11, 2005
In the last 6 months, I've seen a serious up-tick in the number of folks who spontaneously tell me about their renewed love affair with Sun servers, and this is in Dublin Ireland where there is intense awareness of IT costs and where Dell's local presence is particularly strong due to a local manufacturing facility. But at a global level, it's even nicer to see large companies like Microsoft not only acknowledging but talking jubilantly about their reliance on Sun's servers - one of Microsoft's bloggers says more about this.
It makes my day to hear a customer talk passionately about Sun's products; you know that the hard work has been worth it when you see that you've made that kind of connection. I'm confident that StarOffice 8 is going to evoke the same kind of excitement; this time, we really listened hard to what customers were telling us. The existing evangelists out there will be able to point their colleagues to the product and demonstrate that it's now easier to use than Microsoft Office, and it has near perfect interoperability with 10 years of different MS Office product versions, not just the recent releases. The StarOffice 8 Beta release will give a flavour of what's to come. StarOffice and OpenOffice.org already lead the market on Linux and Solaris, but I think StarOffice 8 will give even entrenched Windows deployments a high value upgrade for MS Office 2000 and earlier, while opening up the option to reduce costs through incremental deployments of low TCO Linux traditional desktops or Solaris SunRay consolidated desktops.
On a related note, it surprises me that some analysts haven't really grasped the essence of Sun's strategy, even though it's surprisingly simple.
- Java (J2SE, J2EE, J2ME) has enabled a level playing field for products to be deployed across different hardware and software platforms.
- Sun competes within this benign ecology by delivering high value implementations of hardware and software, and by focussing innovation at the core of solving real customer problems, rather than just expanding the edge with unused immature features
- The existence of alternate implementations of hardware, applications, tools and middlware throughout the entire stack enables customers to choose and prevents a mono-culture of pyramid products that are inextricably bound to one vendor's stack
Following this thought, I would be very interested in seeing Sun and Microsoft co-operate not just at the level of services within a web services standards-based SOA, but at the container level (J2EE and .NET). That would further expand choice and enable customers to achieve more fine-grained sharing of critical resources (servers, load-balancing, clusters, grids) and unify the security models (role-based access control and authentication) of their IT architecture. Technologies like UNO and open-source implementations of the .NET CLR such as Mono can today enable early implementations of several architectures (container peers, hub-and-spoke, dispatcher, etc.), but such a bi-lingual container needs Microsoft and Sun to define common standards in order for this model to become a reality.
While you consider these entirely practical solutions, here's a slightly wilder thought - many people think of Wine (the Windows emulation available on Linux and Solaris) as a solution for running desktop applications, but it could also be an option for running Windows server-based applications and containers say on Solaris 10!
Ah yes - the future's so bright, you've to to wear shades ;)