From: bcox@virtualschool.edu
 To:  David Herschman &<74014.3021@compuserve.com&>
 CC: &
 BCC: !CEM
 

David Herschman's comments help to bring out several crucial points about superdistribution that I didn't have space to address in my article.

Superdistribution is not intended as a short term solution to the practical CD ROM problems he faces right now. There are simpler solutions available for this, such as calling an 800 number for a decryption key.

Superdistribution's goal is to address two deep-seated limitations of such acquisition-based approaches. The first is that once a paying customer decrypts a product, existing solutions provide no obstacle to making additional unpaid copies. The second is that existing approaches don't support revenue flows to sub-component providers; only to the owner of top-level components such as computer applications.

Both of these problems are so fundamental that they're rarely articulated. But since they determine who can get paid for what granularity of product, they shape the entire software industry as we think of that industry today. Superdistribution's purpose is to provide a revenue mechanism that might allow the software industry to redefine itself as manufacturing redefined itself during the industrial revolution. That is, instead of each programmer fabricating everything he/she needs from first principles, to assemble large components from stockrooms of prefabricated subcomponents.

Mr. Herschman says "I'm troubled by the piracy issue in a big way.". He should be. Existing solutions leave him completely exposed to piracy. Once the software is paid for, decrypted, and copied to a hard drive, there are no obstacles to further copying. With superdistribution, copying is no longer a threat since each copy means another paying customer. His concern with superdistribution arises if the metering infrastructure is implemented in a tamperable form such as software. This is why the metered approaches I know of (Ryoichi Mori's Superdistribution system and Wave Systems' chip) emphasize tamper-resistent silicon technology.

Mr. Herschman's concern about who'd decide the revenue sharing split is best handled by asking the same question of the baker, the miller, and the wheat farmer. Except in centrally planned economies such as pre-collapse Russia, nobody does, yet everybody does. Prices are set in a distributed fashion, by the price signaling that healthy market processes encourage. Superdistribution aims to bring the same multigranular market processes to bear on software.

I accept Mr. Herschman's point that Macromind Director provides a no charge run-time license. His objection confirms the point I was trying to make, that there *is* no robust market for information subcomponents, such as the software for reading multimedia documents (and the even smaller components inside it). Macromind's product is an information writer program. They give the reader away for free.

Mr. Herschman's closing point was "publishers can not afford to be watchdogs in this case". Again, he may be correct in the short term, but he articulates a mind set that can be exceedingly dangerous in the long run. Unlike the tangible goods of antiquity, information age goods can be replicated and transported at the speed of light. But at the same time, the global economy is clearly moving in a direction that forces all nations to engage the ownership issues that easily copied goods imply.

A safer attitude is to continually ask ourselves the question that the British neglected when their backwater colony began to displace them at the dawn of the industrial revolution. "What if interchangeable parts turn out to be the silver bullet for the problems that plague us? And what if the competition gets it first?"

A Evolving a Distributed Learning Community
	Spring 1997 (97a) Wired; Response to David Herschman Letter to Editor; Wired Magazine by Brad Cox