No - in your example I would not want a container for the <para> elements. It *is* a matter of instinct (or Sprachgefühl, as I would say), but our instinctive reactions can be reflected and condensed into helpful guidelines. Guidelines are important when we are concerned about quality and consistency, especially when dealing with large and many schemas, and when cooperating with others. Instinct is paramount - (a) when developing/maintaining the rules and (b) when deciding whether to apply the rules - but rules are important. I think you would agree, and if not, I would like to know.
In the current case, I agree with your objection to the suggested rule, so there must be a reason. I prefer refining the rule, rather than discarding it. I feel that the children of <section> are more perceived like a stream of items, rather than a list of items from which to choose. This streaming character is typical for document-oriented XML, while I had only thought of data-oriented XML. The rule should be refined accordingly.
No, I cannot quantify the "weight", it was an unfortunate attempt to use an image, intended to hint at a sense of proportion which is awakened when glimpsing a structure, documents included. (How quantify and express "proportion"? It is about the relationship between the extent ("weight") of parts forming a whole.)
Finally, Stephen points to a practical aspect: container elements serve as "collapsing points", which are extremely useful when inspecting data visualy.
Hans-Jürgen
Michael Kay <mike@saxonica.com> schrieb am 1:39 Dienstag, 5.Mai 2015:
>Dare I propose an esthetic rule that child elements should (if possible) have comparable weight, and the weight is obviously rather uncoupled from the number of value items.
Can you quantify what you mean by “weight”?
Do I read you as saying that you don’t like
<section>
<head>..</head>
<para>..</para>
<para>..</para>
<para>..</para>
</section>
and would prefer
<section>
<head>..</head>
<body>
<para>..</para>
<para>..</para>
<para>..</para>
</body>
</section>
Personally, I’ve found it impossible to come up with a general rule on this. I rely on instinct. In the above example the wrapper seems to serve no conceivable purpose.
