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

On Tuesday, July 27, 2004, 7:35:17 PM, David wrote:

DH> In fact, this is what the not-yet-released Safari (that has switched
DH> over to libxml and SAX) does.  It will render up to the point of the
DH> first error, and then at the top of the page will show you a list of
DH> the errors while still rendering everything up to the first error 
DH> below.

I think that is a very useful method - it allows debugging of where the
wf error is (browsers are used as authoring tools by content authors)
and it gives a visible indication that the document is malformed, so its
unlikely to be just shipped as is.

DH>   I plan to use libxml's push mode to enable incremental 
DH> rendering of XHTML in Safari fairly soon.

Excellent.

DH> On Jul 27, 2004, at 12:54 AM, Chris Lilley wrote:

>>
>> On Tuesday, July 27, 2004, 9:51:36 AM, Andrew wrote:
>>
>>
>> AF> ----- Original Message -----
>> AF> From: "Henri Sivonen" <hsivonen@iki.fi>
>> AF> To: "Chris Lilley" <chris@w3.org>
>>
>>>>> Where does it say in the XHTML spec that incremental display must be
>>>>> disabled for that media type?
>>>>
>>>> Nowhere. It is an just unimplemented feature in those browsers.
>>>>
>>
>> AF> Strictly speaking, XHTML parser (UA) must read </html> to decide if
>> AF> document well-formed (valid) or not.
>>
>> And you further claim, presumably, that it must wait for this closing
>> tag, thus precluding incremental rendering. This is incorrect.
>>
>> The parser is required to go into error when it finds the document is
>> not well formed.
>>
>> It is not required to buffer up the entire document before doing
>> anything else. Indeed, a popular parsing method (sax) does the exact
>> opposite, firing a stream of events as the document arrives, and is
>> ideally suited to incremental rendering.
>>
>> AF> And only ather that render document as it *must* be well-formed 
>> [1].
>>
>> Please point to a specification that requires this chronological
>> sequence.
>>
>> AF> This the end of era of incremental rendering...
>>
>> No, it isn't.
>>
>> AF> [1] http://www.w3.org/TR/xhtml1/#h-4.1
>>
>> AF> Andrew Fedoniouk.
>> AF> http://terrainformatica.com
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> -- 
>>  Chris Lilley                    mailto:chris@w3.org
>>  Chair, W3C SVG Working Group
>>  Member, W3C Technical Architecture Group
>>




-- 
 Chris Lilley                    mailto:chris@w3.org
 Chair, W3C SVG Working Group
 Member, W3C Technical Architecture Group

Received on Wednesday, 28 July 2004 10:34:11 UTC