W3C home > Mailing lists > Public > public-xml-processing-model-wg@w3.org > September 2006

Re: the nested view and the flat / subgraph view (re-send, with standoff images)

From: Norman Walsh <Norman.Walsh@Sun.COM>
Date: Thu, 21 Sep 2006 10:17:44 -0400
To: public-xml-processing-model-wg@w3.org
Message-ID: <87venhgv7b.fsf@nwalsh.com>
/ Alex Milowski <alex@milowski.org> was heard to say:
| In a similar fashion, I've taked the internal "flow graph" inside my
| compiler and generated a flat/subgraph view of use case 5.30 (which
| is faily complex).

To me, this seems very analagous to compiler optimizations. Compilers
can, and often do, unroll loops and inline procedures (which is what
the flat graph looks like to me) but no modern language is described
in these terms.

I still think we should describe our language in terms of components
containing flow graphs that are distinct from the flow graph that the
component itself participates in. The fact that we allow the syntactic
convenience of referring to (some) steps outside the component is
something we'll have to deal with.

Returning to the compiler/procedure analogy, those look like
references to global variables (or other variables in the same scope
as the component, anyway). I think we should approach them that way.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh
XML Standards Architect
Sun Microsystems, Inc.

Received on Thursday, 21 September 2006 14:17:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:48 GMT