🚀 go-pugleaf

In article <4721@peora.sdc.ccur.com> tran@peora.sdc.ccur.com (Nhan Tran) writes:
>Are Data format of commercial software like WordStart, Lotus 1-2-3, dBase, 
>etc.. copyrightable?   If they are, then can other software use those
>data format, i.e. a dBase clone use dBase data format as its own format.

Copyright protection is only applicable to expressions, e.g. the written or
recorded version of something (e.g. a printed book, a recorded musical
performance or movie).  It has to be something that can be copied, so that
you can disallow people from copying it without permission.

A data format is just an idea, not a recorded version of something.  A
particular data file in a format may be copyrightable, but generally not
the format itself.

However, it's possible that the data format may be patentable.  This would
probably only be so if non-trivial technology is necessary to encode or
decode the format.  The format itself actually wouldn't be patented, but
the process of encoding or decoding it might be.  Thus, the format of
compressed files isn't patented, but the LZW compression program is.  If
you can come up with a different compression algorithm that produces
equivalent files you're probably safe (I wouldn't suggest anyone go out now
and try to solve this problem -- the relationship between the LZW algorithm
and its output is very close).

>What about ASCII table?  The latter half of 8bit ASCII table is not defined.
>So for PC, IBM create some line-graphic symbols and put them there.  Can
>IBM copyright that? 

They might be able to copyright the individual character designs, but
probably not the assignment of character codes.

[Disclaimer: I'm not an expert in this field, I only know what I've read
casually.]
--
Barry Margolin, Thinking Machines Corp.

barmar@think.com
{uunet,harvard}!think!barmar