W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1997

Re: A broken browser

From: Daniel DuBois <dan@spyglass.com>
Date: Thu, 02 Jan 1997 21:04:52 GMT
To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <32ce2172.174854960@spyglass.com>
On Thu, 02 Jan 1997 13:31:08 -0500 (EST), you wrote:
>	if (q < 0.001)
>	    q = 0.001;
>We don't know how the compiler is losing that.

Are you using a Pentium?

(Sorry couldn't resist.)

>	My understanding, so correct me if I'm wrong, is that the
>q's are preference ratings, so q=0.000 would be lowest preference,
>not an "I don't want that".  If no wild MIME type is sent, and

In the history of q's, there was a HTTP/1.0 draft that specified the
algorithm for q's that is now left unspecified (awaiting definition, like
in the TCN draft).  IIRC, it said q's should be multiplied together, and
that a q of zero meant 'None acceptable' in the literal sense AKA "I don't
want that".  I've implemented in the past to that kind of mindset, so maybe
there are more like me out there.  I don't believe a negotiating server
should send a document that has a q value of 0, and have said so in the
past.  You may have indexing spiders that insist on only retreiving
text/html and text/plain for instance.

-----
Daniel DuBois, Traveling Coderman        www.spyglass.com/~ddubois
Received on Thursday, 2 January 1997 13:09:24 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:25 EDT