Re: [oasis-member-discuss] Re: [members] Membership and Public Reviewof OASIS Artifact Standard Identification Scheme for Metadata

["Eve L. Maler" <Eve.Maler@Sun.COM>]

As the person who first perpetrated the canonical hyphen-separated 
form on OASIS documents in the pre-Kavi era, I'd like to speak in 
favor of it.  (I was largely influenced by IETF naming rules but 
didn't follow their lead completely; I was going for 
general-to-specific fields rather than their all-over-the-place and 
somewhat more verbose style.)

I don't want to rely on a fancy system that has to look inside a 
document to tell me what it is; filenames are the easiest way to 
label the content "out of band", and making them regularly ordered 
means that it's useful to view and manipulate a pile of them in a 
directory listing or folder, as editors often have to do.

(W3C forces creation of a unique URL/filename on the occasion of any 
change, significant or no, in a document, and although that's a 
pretty Draconian rule I've come to appreciate it.  That way, no one 
is ever tricked into pointing to something whose content changes out 
from under them.)

I don't think it's strictly necessary to make the filenames 
machine-parsable, but reserving hyphens for formal filename parts 
avoids ambiguity even for eyeballs.  If that's truly too onerous, 
however, I'd be willing to take the risk of adding ambiguity by 
allowing field-internal hyphens.

	Eve

Christopher B Ferris wrote:
> 
> Bill,
> 
> As I have tried to make abundantly clear, I think it folly to try to 
> imbue the filename with
> parsable metadata. Metadata should be contained within the artifact (and 
> machine processable
> if possible, in addition to being human readable) or be associated with 
> the artifact
> by some other means (such as I have suggested in other missives).
> 
> As others have pointed out, forcing filenames to be changed just because 
> the
> version/revision changes, or the status, etc. will create more problems 
> than
> it is intended to solve.
> 
> That aside...
> 
> While I could live with the constraint of not allowing hyphens in the 
> filename save those
> that are used as metadata separators, I am fundamentally oposed to 
> precluding their
> use in the non-terminal path component of a URI, especially given that 
> the OASIS
> persistent URI space has already violated this constraint.
> 
> Cheers,
> 
> Christopher Ferris
> STSM, Software Group Standards Strategy
> email: chrisfer@us.ibm.com
> blog: http://www.ibm.com/developerworks/blogs/dw_blog.jspa?blog=440
> phone: +1 508 377 9295
-- 
Eve Maler                                         +1 425 947 4522
Technology Director                           eve.maler @ sun.com
CTO Business Alliances group                Sun Microsystems, Inc.