From: email@example.com (Scott Matthewman) To: <firstname.lastname@example.org>, "Jordan Reiter" <email@example.com> Subject: Handling of script attributes (was Re: HTML 4.0 draft available) Date: Wed, 9 Jul 1997 09:00:03 +0100 Message-ID: <19970709080343421.AAA261@scott.danielson.co.uk> > Why then (as Russell O'Connor has pointed out also) does this new > specification includes the onclick, etc. attributes? The implications of > this are, in my opinion, somewhat more dubious. Here we are getting into > the question of function and behavior, which I think is (to a certain > extent) wrong. We have simultaneously a documentation which warns against > using style sheets to force paragraphs to render without whitespace > in-between (in order not to confuse users), but allows the use of > scripting, which can completely change the way that a browser works > (clicking on a link might open a new window, or turn the page black). But the content of HTML is by its very nature interactive, so there needs to be some means of applying controls to certain events. Specifying which events a UA should monitor in the DTD will help to standardise function across platforms. I do think, though, that the insertion of script events into tags should be discouraged more firmly, to be replaced by binding from a script in the <HEAD> element. Individual event attributes in tags should be deprecated, to be replaced by a SCRIPT attribute, in much the same way as the STYLE attribute can incorporate style changes. Would this go any way to meeting your reservations? What it would then be doing would essentially split a fully-laden HTML document into three distinct sections: 1) Main document, including text, graphics and hyperlinks 2) Style sheet, governing presentational output 3) Script, governing interactivity. ---- Scott A. Matthewman, Danielson Limited <firstname.lastname@example.org> Tel: +44 (0)1296 24478. Fax: +44 (0)1296 392141 ---- "He's not as stupid as he looks." "My dear, no-one could be as stupid as HE looks."