Subject: Re: [regrep] ebXML Registry "Granularity"
Joe +1 on the need to address fine grained sub-document level content. It is already on teh list for V4 but I can see the need to at least do a TN right away on it. I would be glad to work with you on this. Chiusano Joseph wrote: >Team, > >This week I participated in a series of federal XML meetings in which >the OASIS/ebXML Registry standard was mentioned as a candidate for >various potential roles within multiple federal government initiatives. > >That's great news. > >In most if not all of these references, there is a *very strongly* >expressed requirement for granular registration - that is, registration >of elements, attributes, and datatypes - as well as XML namespace >support. There has been a very big push within the federal government in >the past several years for component-based architectures, both from a >software module standpoint as well as a data standpoint. > >I know that if an implementer chose to, they could include such >functionality in their ebXML Registry product, given the abstract nature >of our information model. But the reality is that people read our >specifications and they don't explicitly glean from the specs that the >ebXML Registry standard can accomodate granular registration and >maintenance - so they close the specs and potentially move on to other >possible solutions (solutions which do exist in the federal space and do >allow such granular registration and maintenance, as well as namespace >support). > >That's not-so-great news. > >What I would like to propose is that, in no greater than 3 months, we >make a big technical and marketing push to educate the world as to the >potential of granular registration and maintenance for an ebXML >Registry. I think a great first step would be the creation of a series >of Technical Notes that address (1) how to register and maintain >elements, attributes and datatypes in an ebXML Registry and (2) how to >register and maintain namespaces, and associate elements, attributes, >and datatypes with their namespaces. > >Then, we can combine this capability with CAM, and allow registry users >to assemble schemas using registered artifacts such as namespaces, >elements, attributes, and datatypes. > >Is anyone with me on this challenge? > >Kind Regards, >Joe > -- Farrukh