Re: Sweat Copyrights

At 5:33 PM 2/28/96 -0800, Karjala, Dennis wrote:
>
> I have just read the full opinion in Compaq Computer Corp.
> v. Procom Technology, Inc., 908 F. Supp. 1409 (S.D. Tex. 1995).
> I wonder whether participants in this list, especially those who
> believe that Feist rules out "thin" (or "sweat") copyrights,
> think that this decision is compatible with Feist.
....
> Assuming that the court was correct in deciding that no
> third party would independently determine the PRECISE thresholds
> chosen by Compaq, what do we do with five numbers that are close?
> Anybody making these choices is trying to optimize the same
> things, and consequently we would expect most choices, even those
> independently made, to be in the same ballpark. Suppose Compaq's
> first threshold number is 4.12. (I've chosen this arbitrarily,
> as the opinion gives no hint as to actual values.) Does 4.10 for
> the first threshold infringe? 4.14? 4.0? 4.2? I don't think
> we can tell by looking at the defendant's numbers alone. Rather,
> we have to ask how the defendant arrived at its numbers. If the
> defendant proves independent assessment (through some sort of
> paper trail, perhaps), presumably there is no infringement. On
> the other hand, if the defendant has simply started from Compaq's
> number and made a slight variation, I presume infringement must
> be found. (If not, Compaq's copyright may be rather
> meaningless.)
>
> If this analysis is correct, what we have here is a "thin"
> copyright in specific numbers, which has raised so much objection
> on this list. The numbers themselves are not protected; others
> are simply prohibited from copying those numbers from Compaq's
> disk files. Infringement hinges on the METHOD by which defendant
> determines its threshold values. If the result in this case is
> acceptable on the basis of the intellectual input involved in
> determining the threshold numbers (again putting to the side
> whether functionality concerns shouldn't have denied the
> copyright in any event), why not--as a policy matter--an equally
> thin copyright for "sweat" numbers or facts (or public domain
> materials in electronic form) that are costly to produce but easy
> to copy?

This decision is so misbegotten as to defy even your (valiant) effort to give it a venere of sensibility. One could guess, but only guess, that if the threshold values used by Procom had varied even slightly, the court would have adopted a virtual identity type of standard in light of their obvious functionality, and held that they escaped infringement. In other words, I think (hope) Compaq's copyright is meaningless.

You're right I guess that infringement hinges on the "method" by which defendant determines its threshold values. But only I think insofar as independent creation is always a defense, and access is always an element. As soon as you vary the number even slightly, when you are just comparing two sets of five numbers, how can there be finding that the numbers were "copied."

This, unlike I think the Lotus menu command structure where the argument was recently advanced, has all the hallmarks of a patent claim: Method for predicting hard drive failure, comprising assignment of threshold values reflecting optimum performance; means for reading current values; means for comparing threshold and current values; and means for reporting divergence exceeding predetermined tolerance. (I know, you said to ingore the 102(b) issue, but how can you -- its standing there naked in the middle of Times Square.)

This decision is even more bizarre in its treatment of the Trademark and contributory copyright infringement claims. After ProCom got sued for copying the threshold values directly to its hard drives, it changed course. They started writing Compaq's vendor i.d. to the firmware on their hard drives. Then when the customer configured the hard drive using Compaq configuration software that came with their Compaq computer on which the ProCom drive was installed, the Compaq software would write the coveted threshold values to the ProCom hard drive. The Court held that appropriating the Compaq vendor id was not false designation, but was contributory copyright infringement because it contributed to the unauthorized reproduction of the threshold values by the end user.

The false designation analysis goes something like this. Compaq converted its vendor id to a functional use (prompting its config software to write the threshold values to its own hard drives). Procom had no other legitimate means of replicating this functionality because writing the threshold values to its own hard drives directly is copyright infringement. So ProCom is entitled to appropriate Compaq's vendor id under the Lanham Act even though doing so constitutes contributory copyright infringement.

I can't wait to see what the 5th Circuit does with this. Pray they don't settle.

J. Noble
<jnoble[_at_]dgs.dgsys.com> Received on Sat Mar 02 1996 - 13:34:19 GMT