The ABI files--which allegation are they?

July 27, 2006


Ryan Tibbets' sent a "Dear Unix Licensee" letter on Dec 19, 2003. It put Linux users on notice that use of 71 ABI files was considered a violation of the DMCA. The files in question are described as header files subject to the BSD-ATT settlement, they can be grouped in 9 generic names in forms for several different architectures. This is the infamous *ernno.h* group.

SCO 725 Appendix A -- Objection to Wells motion filing associates the Dear Unix letter in some manner with 5 allegations: 183, 184, 201, 202, 203. (all ongoing, and not dismissed)

Descriptions of the allegations offered elsewhere , and the tab count (single integer) associated with these allegation numbers raise the question as to what form the ABI files were included, or are they included in the final disclosure at all.

The Tibbetts letter lists 71 files in the body and says it attaches a "complete listing of Unix derived files" otherwise subject to restriction. The language "Unix derived files" corresponds to Appendix A of the BSD-ATT Settlement Agreement that lists 91 files. These 91 are generic names, so would more properly correspond to many times the 9 groups in the body of the Tibbetts letter. The attachment file listing is not included in the text of the letter on Groklaw or elsewhere. see: or
ATT settlement:

The objection Appendix A. says "The code identified in this letter was part of the October 2005 Submisssion (Item Nos. 201-203)" and further "and a material part of the code identified in this letter was part of the December 2005 Submission (Items Nos. 183, 184)"

The use of the odd circumlocution "material part of the code identified in this letter" means the 183-4 allegations must stand in some abstraction to the ABI letter and its precise delineation of files.

Appendix A claims 201-203 was part of the October submission (Item Nos. 201-203), yet these are reverse engineered to involve only a single "tab" of supporting evidence, for a misappropriation list that involved 71 named files, and portions of perhaps 91 generic classes of files.

Clearly the broad claims of the Dear Unix letter, and the paucity of apparent supporting evidence in the Final Allegations are incompatible. Has SCO dropped the erron.h allegations entirely? Is it trying to recharacterize the letter by using the bizarre circumlocution: "material part of the code identified in this letter"

< EOM >

10:42:30 PM

Source: Investor Village SCO Board [ ]

Copyright 2006