Re: For-each

/ Innovimax SARL <innovimax@gmail.com> was heard to say:
| <p:viewport
|  name? = NCName
|  match = xpath expression
|  p:ignore-prefixes? = prefix list>
|   (p:viewport-source?,
|    p:output,
|    p:journal,
|    (p:option |
|    p:parameter)*,
|    subpipeline)
| </p:viewport>
|
| it is read
| @name is optional (OK)
| @match is required (OK)
| @p:ignore-prefixes is optional (OK)
| p:viewport is optional as a first child (OK)

You meant to say p:viewport-source is optional, right?

| p:output is required (OK since we have only one output)

Right. We must have one output.

| p:journal is required (WHY ??[1])

Bug. Fixed.

| (p:option or p:parameter) zero or more (ok for p:option but what is
| the aim of p:parameter here ? [2])

With all the parameter proposals floating around, I don't remember :-)
I'll revisit all the p:parameters after we decide what we're going to
do with parameters.

| and a subpipeline
|
| shouldn't it better be
|
| <p:viewport
|  name? = NCName
|  match = xpath expression
|  p:ignore-prefixes? = prefix list>
|   (p:viewport-source?,
|    p:output,
|    p:journal?,
|    p:option*,
|    subpipeline)
| </p:viewport>
|
| Since we mandate a perticular order for p:viewport, I propose to do
| the same for the other component

Well. Ok. We used to have an order for all of them, then we were
persuaded to allow the elements in any order, but I forgot to change
viewport. When I proposed to allow viewport elements to appear in any
order, I got pushback.

I don't care what we do wrt to ordered or unordered children, but I
think we should do it for *all* the steps uniformly.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | The wonder is, not that the field of
http://nwalsh.com/            | stars is so vast, but that man has
                              | measured it.--Anatole France