- From: Norman Walsh <ndw@nwalsh.com>
- Date: Mon, 29 Mar 2010 06:51:19 -0700
- To: XProc Dev <xproc-dev@w3.org>
- Message-ID: <m2mxxrdync.fsf@nwalsh.com>
"Toman_Vojtech@emc.com" <Toman_Vojtech@emc.com> writes: >> You have to import a library. I was imagining that we'd wind up with >> something like http://exproc.org/steps/fileutils.xpl but of course, >> implementations would be encouraged to recognize that URI and load the >> values from some cached location so that every attempt to use the >> exproc steps doesn't actually require hitting the web. > > Since the EXProc "package" specifies both extension steps and extension > XPath functions, we will need to say that importing the EXProc libraries > will also magically enable a couple of XPath extension functions. So I > think that EXProc compliant processors will have to recognize the EXProc > library URIs in any case - and actually do something more than just a > simple import (as defined in the XProc spec.) Well, I think a processor could choose to support the fileutils library of extension steps without supporting any additional packaging system or functions. Be seeing you, norm -- Norman Walsh <ndw@nwalsh.com> | Look for the ridiculous in everything http://nwalsh.com/ | and you will find it.--Jules Renard
Received on Monday, 29 March 2010 13:51:57 UTC