RE: [Yaml-core] Naive reader confused in section 2 of Jan 29 2004 draft spec

[Peter W. <pwo...@qu...>]

PJW> | ... would beg the question how to distinguish between
PJW> | the header comment associated with a document in a
PJW> | stream and a footer comment at the end the previous
PJW> | document ...
CCE>
CCE> Comments in YAML exist only in the "Presentation" level,
CCE> and are not
CCE> considered part of the serialized content ("Document")...
OBK>
OBK> Right. Since comments aren't part of the serialized content,
OBK> YAML doesn't specify any rules associating the comments with
OBK> specific parts of the content. The human reader will, of
OBK> course, associate a comment with some content - that's what
OBK> comments are for, after all, specifying human-only=20
OBK> explanation of the content. Presumably, the human reader knows
OBK> whether a given comment refers to the previous document or
OBK> to the next one (or both, or neither).
OBK> ...
OBK> - comments are not formally associated with specific content,
OBK> ever. However, if you wrote a program that split a YAML stream
OBK> to separate documents, you could use this marker (BOM) to control
OBK> where the comments end up.

Thanks for the clarifications, Clark & Oren.

I think my confusion would be eliminated if the text attatched to
example 2.7 were changed from "Two documents in a stream each with
a leading comment" to something like "Three documents (the first
containing no real data) with 'throwaway comments' interspersed".
The "leading comment" phrase sounds too much like "header comment"
in the HTML/XML world, and really encouraged me think that each
comment *was* supposed to be attatched to a particular document
(if not a particular element).

Cheers,

    Peter