Re: (atomic/compound) steps, containers and related issues

/ ht@inf.ed.ac.uk (Henry S. Thompson) was heard to say:
| I've been doing a close reading of the new alternative draft, and
| (independently of its intended focus on options/vars/params) this has
| reinforced a lingering concern that we just don't have our ontology
| right and it's biting us badly.

I don't feel that particular bite, but I'm going to give this a try
anyway...

| I propose the following analysis as the basis for some suggestions:
|
|            container            multi-container          atomic
|
| step       p:for-each,          p:choose, p:try          p:add-attribute,
|            p:viewport,                                   . . ., p:xslt,
|            p:group[1]                                    pfx:user-declared-ppln
|
| non-step   p:when, p:otherwise,
|            p:catch, p:group[2]
|
|            where p:group[2] is first child
|            of p:try, p:group[1] is all
|            other p:group
|
| * Containers that have single subpipelines as their content;
| * Multi-containers have multiple subpipelines as their content, each
|   wrapped in an appropriate non-step container;
| * Atomics have only binding contents
|
| The environment of a *step* is the inherited environment from its
| *container* (definition in the spec. for this can be retained).
|
| The environment of a *non-step* *container* is the environment
| From its *multi-container*.  [I _think_ this is right, i.e. there is
| _no_ modification necessary]

I think I finessed this by defining "container" and "contained steps"
such that the intervening non-step wrappers are transparent. That is,
the container for a px:foo inside a p:for-each is the p:for-each; the
container for a px:bar inside a p:when is the p:choose above the
p:when.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | A man can believe a considerable deal
http://nwalsh.com/            | of rubbish, and yet go about his daily
                              | work in a rational and cheerful
                              | manner.--Norman Douglas