Just following up on these tests. Both are in need of flanking links to
catch aggregators that just return the first or the last link regardless of
the @rel. For the first test I'd recommend something like this:
<link rel="http://example.org/random";
href="http://www.snellspace.com/public/random"; />
<link rel="http://www.iana.org/assignments/relation/alternate";
href="http://www.snellspace.com/public/alternate"; />
<link rel="http://example.org/random";
href="http://www.snellspace.com/public/random"; />
For the second I'd recommend this:
<link rel="ALTERNATE"
href="http://www.snellspace.com/public/alternate2"; />
<link rel="alternate"
href="http://www.snellspace.com/public/alternate"; />
<link rel="ALTERNATE"
href="http://www.snellspace.com/public/alternate2"; />
As an example, JetBrains Omea currently passes both your tests as they are.
If they were changed as recommended, it would fail both of them.
Also, where does it say that the @rel attribute is case sensitive? I
couldn't find anything in RFC4287, but in HTML at least link types are
definately case INsensitive [1].
Regards
James
[1] http://www.w3.org/TR/html401/types.html#h-6.12
James M Snell wrote:
Yes, I'll make those changes this evening.
James Holderness wrote:
May I suggest you add a third link with rel="ALTERNATE" in your second
test. As it is you're catching aggregators that choose the first of
multiple alternate links, but not catching those that choose the last
link.
