SCO's fatal error?

By claiming copyright in the "Copyrighted Binary Interfaces" of C language header files, SCO is rendering its claim into the land of virtual fiction.

Header files do nothing except label things for compatibility through a system but as such they do not implment binary interfacing - they are ASCII files, not binary, and they only define common names between one useage and another.

In as much as the nuts and bolts of Unix are defined as a sequence of labels with values attached and that is a very generic definition indeed, is to say only that it is a computer language system. If some of the nouns or labels coincide that is due to a compatible operating system (note, it is not a binary compatible operating system, but object compatible) software may be developed with a "common binary interface". To claim that such as interface is unique is oxymoronic - it is a little like two countries that border over a river giving a bridge two different names.

Only, translattors are a bit of a nuisance in software - so a bunch of labels with names provides a common interface for reasons of planned compatibility.

It is a little like Noam Chomsky claiming copyright for "parts of speech".

It is not valid.

The Register

Ownership of intellectual property

See Lawrence Lessig on the Latest "Clueless" SCO Letter (LinuxWorld)

Some of us have an issue with SCO and its stand against Linux, Open Source and the Free Software Foundation. They believe that the "progress of science" is best advanced by protecting authorship rights and the rights of investors to earn a profit from their work.

Some of us would like to be paid more by large corporations who make outrageous profits by fostering products that are inferior.

If a less than efficient methodology is locked in and this wastes the time of most, and that copyright is defended in a way designed to disallow competition, it could be seen to be asking for a licence to slow progress. That is contrary to the ideals of progress.

Before the advent of Linux, Unix was a "mid range system". SCO ( in its former incarnation, perhaps, as SCO of the 1980s) and Microsoft both made x86 processor based version of Unix which were not ground breaking or state of the art in their day. They were just reasonably good and mostly solid implementations of the AT&T Unix system, or a derivative kernal that ran on "IBM PC" machines.

It was a text based implementation, and it was okay. But when the Windows operating system changed the paradigm, text based systems fell out of the window. It distracted the main body of potential users into an "easier" interface. Nobody really needed to read the manual so much now, or learn all those tricky command line tools. You just dragged and dropped.

SCO did nothng with Unix to bring it into the new paradigm. The FSF and Open Source community did that. Now the Linux desktop could run without anything remotely resembling Unix. The operating system part of it is just a set of standard functions that are easily rewritten. It is not that hard to write an OS. Marketing it in the present environment may be difficult.