W3C home > Mailing lists > Public > www-style@w3.org > July 2004

Re: Standards mode and Quirks mode (was Re: [CSS21] Test Suite)

From: Malcolm Rowe <malcolm-www-style@farside.org.uk>
Date: Tue, 27 Jul 2004 13:02:30 +0100
To: Justin Wood <jw6057@bacon.qcc.mass.edu>
Cc: W3C Style List <www-style@w3.org>
Message-ID: <courier.41064456.00007C37@mail.farside.org.uk>

Justin Wood writes:
> Technically, unless I misread in places, we (any UA) *can* render up to a 
> point where a "mal-formedness" occurres and then either "drop" all 
> rendering up to taht point and put up a malformed error, or just leave
> all rendering up to that point and note the error. 
> 
> But to do so reliably would be harder to code than our current "get all 
> document first" method.... the "incremental rendering" (in XHTML) would 
> have to assume that any open tag is closed correctly, until it is not...a 
> bit harder than it sounds at first..,

How is this any harder than incremental rendering in HTML? 

I think that everyone agrees that a UA should not render a non-well-formed 
document; the only point in disagreement seems to be whether the UA is 
required to *ensure* the well-formedness of the document before it starts 
rendering; I don't believe that this is the case (indeed, the XHTML spec 
does not seem to describe *any* behaviour for non-well-formed documents, 
only that documents must be checked for well-formedness [at some undefined 
point in time]. Well, I checked, it's not well-formed, now what?). 

Regards,
Malcolm
Received on Tuesday, 27 July 2004 08:03:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:31 GMT