- From: Dave J Woolley <DJW@bts.co.uk>
- Date: Fri, 14 Apr 2000 11:26:25 +0100
- To: "'www-html@w3.org'" <www-html@w3.org>
Someone has just sent a bug report to the lynx-dev mailing list claiming that Lynx is at fault for choosing a default radio button if none are marked checked in the document. Reading the HTML 4.01 specification leaves me confused as to what is really intended, although all of Amaya, Netscape and IE seem to start with no buttons down in that case. The wording of the definition of checked could be read as logical "if and only if" or as logical implication; Lynx seems to take the latter view. The wording of mutual exclusion rule could be interpreted as "at most one". More of a problem is that radio buttons are described as behaving like check boxes, except for the mutual exclusion rule. This would imply that you could restore an all buttons up state by "clicking" the down button, but none of the browsers does this. The overall result is that the Amaya/Netscape/IE interpretation allows a legal state for the buttons that can only be restored by resetting the whole form (or client side scripting). On real radios, give or take mechanical imperfections, all buttons up requires one to press a non-latching button which could equally well really latch, but all up can be cosmetically better.
Received on Friday, 14 April 2000 06:33:11 UTC