Re: HTML interpreter vs. HTML user agent

On May 29, 2009, at 14:07, Henri Sivonen wrote:

> * The </head> tag should be ignored (and inferred by <body>).

This point was wrong, since it changes the output for conforming  
cases--that is, it isn't a pure error handling change. A pure error  
handling change would allow non-body elements to appear between 'head'  
and 'body' when the tags appear between </head> and <body>

> * </body> and </html> tags should be ignored (and inferred by EOF).

This point is unnecessary if the implementation is ignoring comments  
and wrong if the implementation is reporting errors. If the  
implementation exposes comments to the application, it's impossible to  
make the algorithm streamable without affecting conforming cases.

-- 
Henri Sivonen
hsivonen@iki.fi
http://hsivonen.iki.fi/

Received on Friday, 29 May 2009 12:19:18 UTC