- From: Norman Walsh <ndw@nwalsh.com>
- Date: Tue, 17 Jul 2007 09:33:48 -0400
- To: public-xml-processing-model-wg@w3.org
- Message-ID: <87y7hfgnnn.fsf_-_@nwalsh.com>
/ Innovimax SARL <innovimax@gmail.com> was heard to say: | Dear, | | Fine so for this special purpose, I propose to add a p:map-ports | | map-ports = element p:map-ports { p:port+ } | port = element p:port { attribute base-uri { xsd:anyURI }, (p:pipe | | p:document | p:inline) } | | I know it is a strong requirement, but I want to start discussion to | see where are the acceptable limits for every one | | The idea behind it to give to the component a mapping it could use for | its URIResolver to link to accessible content that could be defined in | the pipeline. The reason why, I let p:document and p:inline is that it | could be useful to have such a way to make XProc aware of what could | be in use. | | Any thoughts ? I think this is too pervasive a change for V1. It would require that pipeline processors have the ability to interact in a very concrete and intimate way with the underlying URI resolver. We know that there will be implementations that have no resolver at all. Even those implementations which do have a resolver may not have access to an API for that resolver which gives them sufficient control to implement this proposal. I think for V1 all we can say is that the behavior of, and even the existence of, a URI or entity resolver is implementation-dependent. Be seeing you, norm -- Norman Walsh <ndw@nwalsh.com> | In a universe of electrons and selfish http://nwalsh.com/ | genes, blind physical forces and | genetic replication, some people are | going to get hurt, other people are | going to get lucky, and you won't find | any rhyme or reason in it, nor any | justice.--Richard Dawkins
Received on Tuesday, 17 July 2007 13:34:06 UTC