[relax-ng] spec typos

[Michael Fitzgerald <mike@wyeast.net>]

Hello,

I mention here a few typos I found in my final reading of the RELAX NG spec,
none of which, if not corrected, would keep me from approving the spec as a
1.0 committee document. Take these from a copy editor's point of view.

- In section 1, drop the period from the second bullet or add a semicolon
after the first bullet. (Compare the bullet list in section 1 of the
compatibility spec.)

- In section 5.1, in the note, "is instance of the data model" should read
"is an instance of the data model"

- In section 6.1, in the parenthetical remark, "(NCName, name without
prefixes)" might better read "(NCName or a name without a prefix)"

- In section 6.2.8, in the sentence that begins "The datatypeEqual
function...," "that is the following inference" might better read "that is,
the following inference"

- In section 7.2, it is suitable to follow both instances of "i.e." with a
comma ("i.e.,")

A curiosity: the meaning of the actual symbol NCName is never given in W3C
NS spec but it is given in the RELAX NG spec and in the XML Schema spec. (I
don't mean the production that NCName matches which of course is in the NS
spec; I mean an explanation such as "a non-colonized name" or "a name
without a colon" or "a name without prefixes.")

Another curiosity: the Latin /id est/ is abbreviated /i.e./ and means /that
is/. It is good practice to stick with either /i.e./ or /that is/ and rather
avoid a mixture of both in a text.

I really like this spec. Particularly nice is the explanation of how to read
inference rules. I also really like the coordination of a single example
throughout the spec. Congratulations!

Mike