wording defining "RDF Literals" needs clarification

[TallTed]

Originally posted by @TallTed in #64 (comment)

@afs said

This is stylistic. I will leave it to the editors of this document.

rdf-semantics/spec/index.html

Lines 778 to 783 in 53b171f

	In summary: with two exceptions, RDF literals combine a string and an IRI <a data-lt="identify">identifying</a> a datatype.
	The exceptions are <a data-cite="RDF12-CONCEPTS#dfn-language-tagged-string">language-tagged strings</a>, assigned the type <code>rdf:langString</code>,
	which have two syntactic components, a string and a language tag; and
	<a data-cite="RDF12-CONCEPTS#dfn-directional-language-tagged string">directional language-tagged strings</a>,
	assigned the type <code>rdf:dirLangString</code>,
	which have three syntactic components, a string, a language tag, and a base direction.

Lines 778-783 can easily be read as listing a total of four or six things -- 2 syntactic components plus 1 string and 1 language tag; or 3 syntactic components plus 1 string, 1 language tag, and 1 base direction -- rather than the intended two or three syntactic components, being enumerated as "1 string and 1 language tag" or "1 string, 1 language tag, and 1 base direction".

I think too much is currently being asked of a single, very long sentence. Here I break it into two sentences. It could be broken into three or even four.

    In summary: with two exceptions, RDF literals combine a string and an IRI <a data-lt="identify">identifying</a> a datatype.
    The exceptions are <a data-cite="RDF12-CONCEPTS#dfn-language-tagged-string">language-tagged strings</a>,
    which are assigned the type <code>rdf:langString</code>, and
    <a data-cite="RDF12-CONCEPTS#dfn-directional-language-tagged string">directional language-tagged strings</a>,
    which are assigned the type <code>rdf:dirLangString</code>.
    <a data-cite="RDF12-CONCEPTS#dfn-language-tagged-string">language-tagged strings</a>
    have two syntactic components, these being a string and a language tag;
    <a data-cite="RDF12-CONCEPTS#dfn-directional-language-tagged string">directional language-tagged strings</a>
    have three syntactic components, these being a string, a language tag, and a base direction.

[pfps]

This is a summary of a definition from elsewhere. It is of the form "something, something; something, something." The semicolon is a stronger separator than the commas so there is, in my mind, no ambiguity.

There are lots of ways of writing this. I find the current way concise, unambiguous, and quite readable.

[TallTed]

It is of the form "something, something; something, and something."

Rather, it is of the form "3 somethings, something, something, and something."

It would be better to make it of the form "3 somethings: something, something, and something."

[pfps]

I was referring to a larger context, where there is a semi-colon.

I don't see that using a colon would be better in the two cases.

[TallTed]

I don't see that using a colon would be better in the two cases.

With it being an enormous sentence, colons don't work, because there's no "best way" to end the first list and start the next phrase. Hence, my rephrasing as two sentences.

[pfps]

I still don't see any problem.

[TallTed]

[@pfps] I still don't see any problem.

Do you see a problem in my currently suggested rephrasing?

If not, since I do see a problem with the original, my suggested change should be applied.

If so, please elaborate, so I may try another rephrasing that may resolve both the problem I see in the original and the problem you see in my currently suggested rephrasing.

[franconi]

Like @pfps, I don't see any problem with the original.