XSLT version

At the face-to-face, we've been discussing the issue of XSLT 1.0 vs
XSLT 2.0.

Michael Kay pushed back, other commenters have pushed back, and the
XSL WG has pushed back on our decision to require XSLT 1.0 and make
XSLT 2.0 optional.

So, suppose we don't require it?

Suppose instead we say that there's a single, required p:xslt step and
an implementation may implement XSLT 1.0 or XSLT 2.0 (or both) at its
option.

Here's the proposed signature:

<p:declare-step type="p:xslt">
     <p:input port="source" sequence="true" primary="true"/>
     <p:input port="stylesheet"/>
     <p:input port="parameters" kind="parameter" sequence="true"/>
     <p:output port="result" primary="true"/>
     <p:output port="secondary" sequence="true"/>
     <p:option name="initial-mode"/>
     <p:option name="template-name"/>
     <p:option name="output-base-uri"/>
     <p:option name="version"/>
</p:declare-step>

If the pipeline author requests a version of XSLT that the
implementation does not support, the implementation SHOULD report a
dynamic error. If the pipeline author doesn't request a version, the
implementation may run any version it wishes, and may use any criteria
it wants, including sniffing the XSLT stylesheet to make that
determination.

If XSLT 1.0 is run, it is a dynamic error unless exactly one document
appears on the source port. The secondary result port will always
return an empty sequence.

It is implementation-defined if the initial-mode, template-name, and
output-base-uri options are supported for the version of XSLT that is
used.

The allow-version-mismatch option is removed; you get what the
implementation gives you.

The allow-collections option is removed; if you run XSLT 2, the
sequence of documents that appears on the source port is always the
default collection.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | Everything should be made as simple as
http://nwalsh.com/            | possible, but no simpler.