Norman Walsh wrote:
> / Dave Pawson <> was heard to say:
> | Norman Walsh wrote:
> |> / Dave Pawson <> was heard to say:
> |> | I wonder what the original reason was, unless it was M$ type of pressure
> |> | to have everything everywhere.
> |>
> |> See http://docbook.org/tdg5/en/html/para.html :-)
> |>
> |> But more helpfully, consider the following example:
> |>
> |> There are times when it may be necessary to frob the foobar. These
> |> can be summarized as follows:
> |>
> |> <some table goes here>
> |>
> |> where anything that falls outside the boundaries of column 1 must
> |> be considered an error.
> |>
> |> Logically, that's a single paragraph with a table in the middle. To
> |> mark that up as two paragraphs with a table in between fails to
> |> capture what the author intended.
> |>
> |> Years of struggling with HTML has mostly trained me not to write that
> |> way, or not to worry about the mangled markup that results from making
> |> that three sibling elements, but it's still a rational markup model.
> |
> | I'm probably being 'too strict' /picky, but two paragraphs and a table
> | seems right to me (and is what I'd do).
>
> Sure, but if you do that, you also have to rewrite the two paragraphs
> so that they stand alone, yes? It's surely very bad style to introduce
> a paragraph break in the middle of a sentence (poetry and literature
> excepted, perhaps).
Literally speaking, I don't think a phrase is intended to hold a table
(or any "block" element) in the middle of it. 1) Reading a phrase with a
big chunk of something else in the middle is uneasy. 2) That's even
worse with respect to accessibility.
Also, one specificity of a paragraph is that it starts on a new line. So
we have a vocabulary issue here, as above paragraph contains many new lines.
Now if you need to enclose a single idea into a structuring element able
to contain paragraphs and other structures, there are already a bunch of
them in DocBook. Though maybe none neutral enough for all purposes...
Camille.
> At the end of the day, I don't think there's a technical argument to
> be made here. Some authors absolutely think paragraphs can contain
> block structures, some don't.
>
> | Just an observation.
> | If the current simpara content had been in para and a new
> | 'messy' / complex model presented (complexpara for want of a better
> | name) then I'd have ignored complexpara and used the para.
>
> Alas, we crossed that bridge sixteen years ago.
>
> | [Or I could start using docbook-- with the content model of simpara
> | substituted for para]
>
> Sure. That's a perfectly reasonable subset of DocBook.
>
> Be seeing you,
> norm
>
