Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Revised statement of policy on CIF

Brian McMahon wrote:
> ...a revised policy document for CIF.

I am afraid that I do have problems with the policy listed below. While
I support the intent, It has been my experience that most commercial
software vendors are very uncomfortable with releasing source code. I
fear that with this requirement, many companies would avoid creating CIF
implementations rather than expose themselves to be require to release
source code.

Besides, source code seems unneeded. If CIF is indeed a robust standard,
the files must comply with the standard not the software. Compliance
should be assured by a validation kit, not by analysis of coding. We
already have tools for analysis of CIFs so analysis of written CIFs is
easy. What we need are a set of test files that to validate software
that reads CIFs.

I am uncomfortable with several areas where I feel the document is vague
or overly strenuous. For example, my reading of this policy is that all
CIF software must handle all CIFs and thus any application that can only
read DDL 2 CIF's is not CIF-compliant and cannot be distributed. Also,
how must the notice be attached? Is including the notice inside 50,000
lines of source code sufficient?

> At Herbert's suggestion, I have drafted a paragraph (the final one below)
> touching on the rights associated with CIFs themselves. Comments on this
> paragraph will be especially welcome.

The last paragraph is very important and very good. I would like to see
the point made that a bit more clearly that one does not give away any
rights to the information by encoding that information into a CIF. How
about: "While submission of a CIF to a journal or other organization may
require a transfer the copyright, no ownership rights are transferred
when information is written in the CIF format. Also, I find the
following sentence is a bit ambiguous. 

> larger data collections so long as the rights of the copyright owner are
> preserved and respected.

I would like to see it clear that "copyright owner" refers to the owner
of the information in the CIF rather than the owner of the CIF format.

Brian T.