Re: [office] ODF, OCF & UCF -- Some container file options

[Bruce D'Arcus <bdarcus@gmail.com>]

Thanks Rob,

On Dec 18, 2006, at 11:33 AM, robert_weir@us.ibm.com wrote:

> ...
> A summary of what OCF added beyond ODF can be read in this note:   
> http://lists.w3.org/Archives/Public/public-appformats/2006Dec/ 
> 0131.html
>
> And you can read the full OCF 1.0 specification here:  http:// 
> www.idpf.org/ocf/ocf1.0/index.htm

I think the requirements that obviously drove their design are  
sensible, and overlap with some of what we've discussed in the  
metadata SC. In other words, it certainly ought to be technically  
possible for us to extend the manifest schema in ODF to cover these  
needs.

However, their actual solutions seem rather strange, in particular  
using hard-coded paths to denote particular kinds of files  
(signatures, metadata, etc.). For example, the spec says:

"A file with the reserved name “metadata.xml” within the “META-INF”  
directory at the root level of the container file system may appear  
in a valid OCF container. This file, if present, MUST be used for  
container-level metadata"

If we need to essentially type a given file, then we should add that  
to the manifest. But using file names to do that implcitly is not  
very scalable.

What we ought to look at, then, is a more extensible manifest schema.

Earlier I had mentioned that the manifest actually validates as RDF,  
and so gives us built-in extensibility (a minimal example is to just  
say any foreign-namespaced attribute of a file-entry is a property of  
that entry).

However, in looking more carefully recently, I've discovered there is  
one minor detail that would make some manifest files non-valid.

Bruce