[rfc-i] issue: canonical formats
touch at isi.edu
Fri Jun 1 07:38:33 PDT 2012
On 5/31/2012 7:36 PM, John Levine wrote:
> As I read the wiki page, I see notes about a canonical source version
> and a canonical display version. I would like to suggest that there
> be only one canonical version, and whatever it is, it should be a
> source versiont has structure and metadata at roughly the level that
> xml2rfc does.
I agree that there should be one canonical version, but I consider that
necessary only for the archival version.
As to metadata, I don't think that needs to be inside that version per
se; it needs to be available but not necessarily integrated. That said,
integration of metadata about the doc itself (author, title, date)
should be easy to include. Metadata about other docs (refs) serves no
useful purpose since there's no standard for document metadata that we
can assume will help link docs together.
As to structure, there should be two considerations:
- required structure should be minimal. AFAICT, that includes location
marks and references to locations, and might include tags around code
- optional structure should optional - i.e., removable. that includes
containment markings around non code/abnf (e.g., sections), lists, and
other aspects of document structure
Optional structure serves only one purpose AFAICT - support for future
editing, but the cost is more complexity in the archival format. Since
most revisions are rewrites (after the RFC is published), we should not
be trading stability and persistence for author convenience.
More information about the rfc-interest