Re: Mozilla's rendering of slightly buggy HTML

[Henri Sivonen]

In article <[EMAIL PROTECTED]>, 
[EMAIL PROTECTED] (Jason Bassford) wrote:

> 1. What part of the coding has Mozilla stop at 196 (Confirmed?) lines?

There is a nesting limit in the parser.

>  What's the reasoning for this particular number?

Someone figured that allowing deeper nesting would make a stack overflow 
somewhere. (It has happened. That's why the limit was put in.)

I should point out that reaching a nesting level of even 20 is rare in 
valid HTML documents.

> 2. Shouldn't it be consistent?  I.e. shouldn't it display either 0
> lines or ALL lines?

No.

> If not, what's the (technical) reasoning behind
> the behaviour as it stands?

Most real-world cases of bogus HTML will display with the current stack 
limit. There is so much bogus HTML out there that Mozilla can't afford 
to display nothing if it discovers a mild case of faulty HTML markup. 
(XML is another story.) OTOH, increasing the stack limit to allow even 
more broken pages is most likely not worth the effort. Authors should 
just fix their HTML.

-- 
Henri Sivonen
[EMAIL PROTECTED]
http://www.clinet.fi/~henris/