Re: top level domain name in "author" tag of RSS feed.

Frank Ellermann wrote:
> Stephane Saux wrote:
> 
>> However, the list of top-level domain names include some that have more 
>> than four characters, including "travel" and "museum" (and some realy 
>> wacky ones: 'XN--11B5BS3A9AJ6G')
> 
> Indeed, the eleven I18N example.test domains are supposed to work since
> yesterday.  And the syntax of mail addresses allows weirder constructs,
> especially domain literals like '[127.8.9.10]'.
> 
>> So the validator could be both stricter (only accept 2-3-and 4-letter 
>> top domain name in the list,) and more tolerant (a corollary, accept 
>> longer top-level domain name.)
> 
> What you call "stricter" would be wrong for addresses in the 'museum'
> or 'travel' TLDs.  Or for addresses in a 'localhost' domain, that's
> at least syntactically allowed.
> 
> Check out <http://feedvalidator.org/>, it reports the same non-error,
> and it finds another non-error "missing atom rel self".  I think the
> W3C validator is an older variant of <http://feedvalidator.org/>. (?)

The W3C validator periodically re-synchs with feedvalidator.org.

Feedvalidator.org has had a number of recent updates, including:

  * just this morning, I took the suggestion of Stephane and checked
    for the TLDs in the IANA registry -- both short, longer, and IDNAs.
  * should the name not match the registry, the feed validator will
    check to see if the name resolves via DNS.  If so, no warning is
    issued.
  * recommendations are more clearly marked as merely being
    recommendations and not being errors.

> BTW, has RSS 2.0 some kind of "official" schema or at least a DTD ?

RSS 0.91 as a DTD, RSS 1.0 and Atom 1.0 have schemas (RDF and RelaxNG 
respectively).

>  Frank

- Sam Ruby

Received on Friday, 2 November 2007 13:45:52 UTC