W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1995

Re: Semantics of "Accept:"

From: M. Hedlund <march@europa.com>
Date: Wed, 22 Mar 1995 11:40:14 -0800 (PST)
To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <Pine.SUN.3.91.950322110036.8542A-100000@thetics.europa.com>

On Wed, 22 Mar 1995, Jim Seidman wrote:
> My biggest problem is that current practice for most browsers seems 
> to be an accept line like:
> 
> Accept: text/html, text/plain, image/gif, image/jpeg, */*; q=0.3

Some notes on current practice (from the latest versions I've seen):

* Sends Accept: with at least one q value --
	Spyglass 2.00a1, OmniWeb/0.7.5.5, ATSON-W3LM/1.0 beta

* Sends a list of accepted types, including */* but without q values --
	Lynx 2.3.7, WinMosaic 2.0a8, NCSA Mosaic for X 2.5b4, AIR_Mosaic
	3.07.04.02, GWHIS/110, Netscape 1.0

* Sends just a list of accepted types, without */* or q values --
	MacMosaic 2.0.0a17, Netcruiser 1.10 beta, MidasWWW 2.2, EI*Net/0.1

In addition, Emacs-W3 sends a list with some subtypes wildcarded, but 
does not list */* (yay Bill); and Chimera 1.0 just sends */*.  Lynx and 
Emacs-W3 (at least -- probably others as well) modify the Accept: list 
based on your preferences.

[Of those that send q values, ATSON-W3LM and Spyglass 1.x appear to use 
them correctly.  OmniWeb sends '*; q=0.500', omitting the wildcard for 
subtype.  Spyglass 2.00a1 (Win32) sends '*/*, q=0.300', using a comma 
instead of a semi-colon, but presumably I have old information on that 
count.]

M. Hedlund <march@europa.com>
Received on Wednesday, 22 March 1995 11:52:42 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:14 EDT