On Jan 27, 2006, at 09:16, Ian Hickson wrote:
In that case I don't understand your proposal.
We need something that, given this:
<em><p>X
...results in this DOM:
BODY
EM
P
...and given this:
<em><p>X</em></p>
...results in this DOM:
BODY
EM
P
EM
Given that "X" can be arbitrarily long, how do you generate the
above DOMs
using your system without sacrificing incremental rendering?
I understood your system to be:
[TOKENISER] --> [PREPROCESSOR] --> [PARSER]
...where the preprocessor rearranges tokens, and the parser creates
the
DOM. How does your system handle the cases above without blocking
in the
preprocessor stage to wait for the end of X?
By producing the second DOM in both cases.
Based on Hyatt's comments, I take it that this would make the DOM too
heavy and would not work.
--
Henri Sivonen
[EMAIL PROTECTED]
http://hsivonen.iki.fi/
