>> The hawk maintained that the imminent abundance of bandwidth ...
>> The accountant countered that the turnpike and corridor effects ..
Hi Frank ... I've been hiding out in Win2k internals and writing business plans.
I am having major deja vu here. This sounds like the same conversation that held the mike at the 1974 Fall Joint Computer Conference ... only the nouns plugged into the dynamic at that time were about the desirability of more powerful programming languages.
"More Power! More power in the language will solve everything" ...
same-same with bandwidth now ... then, we were fixated on the ability to express abstractions coherently ... now, we are fixated on the ability to move data around coherently.
The meeting divided into two camps, around the mike, the More is Better camp and the Less is Beautiful camp. The more guys liked PL1-like ideas ... the less guys liked C.
Neither of these two arguments makes any sense without a "context of expectation" ... the act of "wanting" ... or having a goal ... itself defines a horizon of the possible beyond which one cannot look without invalidating that short-term goal.
I commented then that no matter how much "concept power" designers built into new languages, programmers would conceive of programs that lay beyond the power of the language to express coherently.
Today, the same is true. No matter how much bandwidth we have to play with, we will design systems that demand more.
It's all about cycles of expectation.
And whenever this structure appears, it means there is some underlying truth that's being ignored.
I believe the underlying truth here is that the current economic model of the web is not aligned with the eventual technical model.
This thought has immensely powerful implications ... and I believe it to be a compelling investment concept for long-termers like me.
1. The technical model of the web must necessarily evolve toward a generalized system which automatically positions pages optimally, managing servers and links just as a PC operating system manages PC resources.
2. The concept of "ownership" will change radically. Currently we think "this is my server", "this is my rack", "this is my wire", "I sell you the use of these". This socio-economic gestalt is doomed by the realities of inevitable progress toward a truly intelligent web.
3. The concept of "innovation" will change radically. Years ago, we internals programmers made money by innovating new enhancements to "operating systems" ... we wrote cool TSRs, applets that did this or that ... add-ons for an incompletely envisioned environment. Now this market is nearly dead ... Microsoft has "envisioned" itself into a position which subtends all such efforts ... if anything you invent is useful, it will become part of Windows.
Similarly, when the web is "completely envisioned", then all these petty differences in web infrastructure technology will become moot ... the surviving methodologies will be incorporated into the larger paradigm and the innovation process itself is transformed in the process ... becoming more standardized and less diverse and more locally efficent.
4. Some companies are more likely to prosper in this transition than others. Some companies have Microsoft-like traits such as a willingness to beg/steal/borrow external ideas, flexibility is self-description, the ability to morph their structure, etc, etc. These are more likely survivors than companies whose "identity" depends on a specific technology.
5. The facts of the transition are more knowable that the relative details of competing technologies. Remember the competing desktops systems of the early 80's ... the competing languages ... the competing bus systems? Who could really predict one would win over another? Many surprises occured. One could, however, predict how the underlying trends would move toward a conclusion ... amd who was attitudinally positioned to survive.
Anyway ... that's what comes to mind. |
