Accepted Version

Copy of accepted version with added suggested minor edits from Ref 1. This is the copy returned to the publisher after receiving notice of acceptance.

Revised Submission

This is the paper which was revised and submitted back to the A&C.

Re-submission draft

This is the edited draft, accounting for referee comments, which is proposed for resubmission.

Re-submitted Paper

New submission, per request of A&C editors, to remove the ordinal from title, disengage the paper from
association with Slava's and Tim's data format submissions to the journal.

A&C Submitted Draft

This is the draft submitted to the A&C journal.

Release Candidate 5

Here's release candidate 5. Other than minor grammatical and typographical fixes, it features a rewriting of the "versioning" section, 2.1.1.

Release Candidate 4

Here's (hopefully) the last release candidate, number 4. This one features very small grammar and spelling fixes. The only 'major' change occurs in section 2.1.2 which adds text to describe informal variants better and point out that the content in figure 1 is an example because of the use of the EPOCH keyword.

Release Candidate 3

Another release candidate. Various improvements and changes in this draft. Main content change is the removal of section 2.4.5 (Limited support for image compression) because the technical arguments needed to suppport this section appear to be long, and are better addressed in Slava's paper III.

Release Candidate 2

Release candidate 2 has numerous fixes to the prior release. Improved wording,
addressed all outstanding notes and fixed errors/missing references. This draft
is the first one which might be considered submission-worthy. Please review.

Release Candidate 1

First release candidate for draft which will be submitted to the A&C. This draft has only a small amount of redded text. One bit relates to potential fleshing out of problems with FITS tables, the other text may wind up being part of a potential argument/lession Tim is drafting on why conventions are not sufficient to the needs we have in software development with FITS.

This draft should otherwise be taken very seriously and read critically by all CO-authors.