Sun's Apparent Ooooooooooops! (How many o's?)
Sun's decisions with respect to Java could be an oops, but it could also be that they are diversifying their business strategy to include the possibility (and a reasonable one at that) that hardware is already fairly commoditized (as you readily admit in your article), or at least if it isn't wholly so, that it will be increasingly commoditized to the point where Sun's margins on hardware do start to come tumbling down. In that circumstance without Java they don't have anywhere near as strong a fallback position to maintain profits.
However, while it would be true to say that the existence of Java increases the likelihood of hardware becoming increasingly commoditized, Sun MAY have asked themselves the question, and this is the important bit: "What happens to us if this happens ANYWAY?" It is better to be first at establishing a new market for a product which perceived as a threat than to ignore that possible threat or to do a Microsoft and attempt to thwart it (see Linux FUD...), since neither of these really works too well in the LONG term.
This is what people will often forget about Sun, it IS a LONG TERM BUSINESS STRATEGIST. Bill Joy and Scott McNealy for their failings are no dills. that is why they appear to be the odd one out, it is because they are often ahead of the curve. They are akin to the countercyclical trader: buy when everyone else is selling, and sell when everyone else is buying.
Scott McNealy, in relation to commoditisation of hardware, is quoted as saying that Sun would be happy to be in IBM's situation of having just a slice of such a huge market. They even used the commodity approach to sourcing their hardware components consequently and ever since 1982. That is why Sun not recently gives away or licences most everthing of their intellectual property.
Sun understands and has understood for a long time, the one dictum which is relatively new in economics (are you listening?) that there is an inverse relationship between supply and demand to that which is usual in microeconomics for protocols. That is that the more people who use a protocol the more value it has. This is because a protocol doesn't get used up. It is naturally inclined to be commoditized.
Microsoft understands this and wants to move from owning the desktop to owning the protocols for the higher layers of the internet with .NET. Were it to be or have been possible Microsoft might have considered licencing the Windows source code, rather than selling programming tools for it. Same effect, different strategy.
Applied to Java protocol economic theory means that if Sun can determine the marketspace for this kind of product then they will make their small percentage of the huge pie they just created.
Their participation with CORBA is perhaps a more reasonable basis for argument about the goodness of their strategies, but even so one needs to enable some form of easy substitutability (as you have written) and CORBA provides that for Java amongst OO languages.
Of course this may be a brilliant justification for some pretty amazing accidents, but I'm not so sure about that...
Tuesday, June 18, 2002
My guess is that java was a response directly to visual basic. That is: "We would buy sun boxes, but VB/PB is soooo easy to develop in and all we can use your crummy boxes for is C/C++/Perl"
Some things I don't understand though:
here they are creating the JDK's/writing the specs etc...
yet they cant make a decent app-server (the biggest java money maker so far) granted they make good money of licensing JDK to bea etc ... but still
Also, I disagree with joels analysis that multi-platform commodotizes sun, what it does do is give their sales people an in everywhere java is used, i.e...
If your servers are slow comon down to crazy bobs starcat/e10000 emporium!!!!!
I would say java commodotizes software, and lets Sun compete on nhardware!
Tuesday, June 18, 2002
Fog Creek Home