W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2008

RE: Feedback for draft-nottingham-http-link-header-03

From: Ian Hickson <ian@hixie.ch>
Date: Sat, 6 Dec 2008 10:18:58 +0000 (UTC)
To: HTTP Working Group <ietf-http-wg@w3.org>
Message-ID: <Pine.LNX.4.62.0812061005460.17401@hixie.dreamhostps.com>

On Fri, 5 Dec 2008, Drummond Reed wrote:
> Much better to either: a) drop support for "rev" but leave "rel" 
> unambiguously asserting outbound links [...]

In HTML5, <link> can only assert outbound links. The rel="" attribute 
merely gives the kind of link (e.g. an external resource with particular 
processing requirements, or a hyperlink with certain semantics).

(We can't really define inbound links with HTML for two reasons: one is 
that we don't really have a way for the user agents to know that the links 
exist if they aren't outbound, since user agents tend to work on a 
per-document basis; the other is that even if we had a mechanism, we 
couldn't trust other documents to assert links on a document because it 
would be too heavily abused for spamming, phishing, and the like.)

On Sat, 6 Dec 2008, Phil Archer wrote:
> It's the same basic argument used by the WHAT WG to decide whether to 
> keep or remove features in HTML 5. On the surface it's an attractive 
> one: if no one uses a feature properly, or worse, where a feature is 
> used it's generally used improperly, then it should probably be dropped. 
> The flaw with the argument is that it can only ever lead to a 
> diminishing feature set and imposes restrictions on current and future 
> use cases.

It demonstrably doesn't lead to a diminishing feature set, just look at 
the size of HTML5. :-)

It does result in features being trimmed, though. That's a good thing; it 
reduces featuritis and keeps the language focused on important things.

> To emphasise this a little, try these two:
> Link: <foo.bar> rel="describedby";
> Link: <foo.bar> rev="describes";
> Here the rel/rev values are the inverse of each other so that in effect 
> the same relationship is being asserted in two different ways and there 
> is no obvious reason to put more weight in one rather than the other.

Right, it isn't clear to me what rev="" really brings to the table. As you 
show here, pretty much any "inbound" relationship can be described as an 
"outbound" relationship (in the extreme case by just minting a new value 
with "rev-" on the front and defining it as such).

> Most Links will use rel, not rev, but if we can clearly articulate the 
> semantics as being equal and opposite, whilst making it clear that this 
> says nothing about trust, then perhaps we can retain rev and leave the 
> door open to future possibilities.

There are many, many features we could add with that justification. Is it 
enough? For HTML5, it hasn't been.

Anyway, I'm not trying to make any recommendations for HTTP or for the 
Link: header itself, just trying to give some background on what went into 
the current decision for HTML5. Hopefully that will help one way or the 

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Saturday, 6 December 2008 10:33:52 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:47 UTC