W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2009

"start", "first" and "last" in link relations draft, RDFa and (X)HTML

From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Mon, 16 Nov 2009 02:25:43 +0100
Message-ID: <4B00AA17.3040306@xn--mlform-iua.no>
To: ietf-http-wg@w3.org
CC: RDFa mailing list <public-rdf-in-xhtml-tf@w3.org>, HTMLWG <public-html@w3.org>
This is input on the http-link-header draft. CC to public-html@ and 
public-rdf-in-xhtml@.

    Please see the following 2 links to some comments by Alexandre 
Alapetite on the HTML 5 draft's understanding of link relations. The 
latter text (the table) seeks to document the most important compound 
document link relations as defined/understood by implementations and 
specs [note that the table is still a work in progress - it is still 
being expanded and updated/corrected/discussed - e.g. at the moment it 
doesn't mention whether the Nottingham draft or RDFa-in-XHTML.]:

[Message] http://lists.w3.org/Archives/Public/public-html/2009Nov/0373.html
[  Table] 
http://alexandre.alapetite.fr/divers/vrac/20091115_HTML_link_rel.html

    I note that the latest http-link-header draft defines "first" and 
"last", and that it defines them as opposites. (Quite naturally to see 
them as opposites, I must say.)

http://www.mnot.net/drafts/draft-nottingham-http-link-header-07.txt

    The http-link-header draft says that it is the first to define 
"first" and "last". However, "last" is already defined in the XHTML 
Vocabulary document, which sees "last" as the opposite of "start":

http://www.w3.org/1999/xhtml/vocab/#last
http://www.w3.org/1999/xhtml/vocab/#start

    The XHTML Vocabulary builds on RDFa in XHTML:

http://www.w3.org/TR/rdfa-syntax/#relValues

However, when we look at RDFa in XHTML, then it also defines "first" - 
without there being any corresponding definition of "first" in the 
Vocabulary document ... That is: "first" is lacking from the Vocabulary 
document ...

RDFa also operates with "start" and "top". I don't understand this 
discrepancy between the Vocabulary and RDFa-in-XHTML ... There may also 
be other values that are in RDFa but which are not in the Vocabulary ... 
The RDFa spec claims to be older than the Vocabulary ...


    HTML 4 did not define "last" but it defined "start". It's definition 
of "start" includes the word "first". However, when W3C in a "W3C QA 
Tips" explains what "start" means, then it shows an example that has to 
do with file/folder/directory hierarchy:

    <link rel="Start" href="/solar-system/" />
    <link rel="Prev"  href="/solar-system/venus/" />
    <link rel="Next"  href="/solar-system/mars/" />

http://www.w3.org/QA/Tips/use-links


     Thus it seems like "start" as defined in HTML 4 (at least as 
understood by the QA Tips page) is more similar to "top" and "home". Or 
in plain English: "homepage".  As one can see in Alexandre's table, 
almost all the implementations (all web browsers and browser extensions) 
that support link relations have "top", "home" and "start" as synonyms. 
None of them see "start" and "first" as synonyms. (And they all support 
"first"!)

So in a summary:
1. The nottingham draft is not correct in saying that it is the first to 
define "first" and "last".
2. Secondly,  in my view, the best thing would be to follow the 
implementations and RDFa-in-XHTML with regard to he meaning of "start" 
versus "first". 3. Thus we should separate "start" and "first" - as the 
nottingham draft does.
4. I hope the XHTML Vocabulary document gets updated so that it actually 
documents what it claims to document, namely the RDFa-in-XHTML syntax 
specifications.
-- 
Leif Halvard Silli
Received on Monday, 16 November 2009 01:26:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:13 GMT