Re: Error handling and legacy content

On May 2, 2007, at 2:00 AM, Gareth Hay wrote:

>
> On 2 May 2007, at 09:18, Anne van Kesteren wrote:
>
>>
>> I have a hard time understanding what people actually have against  
>> handling legacy content. This doesn't affect authors. Authors are  
>> currently not allowed to write "<em> <strong> </em>" and HTML5  
>> will not allow them to that either. (Also, to be clear, HTML5  
>> defines what authors, user agents, scripting enabled user agents,  
>> scripting disabled user agents, interactive user agents, non- 
>> interactive user agents, etc. have to do. It's certainly not just  
>> "desktop browsers".)
>>
> There is a world of a difference from "it's best practice not to  
> write" and "it doesn't work if you write it like that".
> Authors will /never/ stop writing tag soup unless there is some  
> barrier there. I suggest that HTML5 should present this barrier.
> By all means support legacy content written in bad ways, but don't  
> support html5 written in bad ways.
> That is my point.

Let's think about this. Why is non-conforming content harmful? The  
reason usually given is that it's unpredictable how it will be  
handled, and may be different between different UAs. However, if we  
define exact error-handling rules, that solves the problem. And the  
cost to the industry as a whole is much less than trying to wipe out  
non-conforming content.

Do you have any reasons in mind why non-conforming content is harmful  
other than the interoperability problems?

> Well, I think I have made my point.
> It appears that there is a great resistance amongst people with  
> differing interests in html to listen to this side of the argument,  
> and they simple dismiss it.
>
> I think, in particular, that the browser vendors have made their  
> stance clear. They are not even inclined to think about this  
> approach, and in that regard, I think I am banging my head against  
> a very large brick wall, with maybe only one or 2 others.

We think there's strong practical benefits to defined error handling  
that matches legacy error handling. If you want to advocate a  
different approach, I think you need to relate it to practical  
benefits. So far, you've suggested that draconian error handling in  
HTML5 will eliminate non-conforming content. Personally, I'm not so  
sure - as long as pre-HTML5 browsers are popular, and apply error  
handling, a new HTML5 UA would not be able to implement the draconian  
rule. But even given that, I'm not sure why eliminating invalid  
content is a benefit. Could you please explain what the practical  
problem is with nonconforming content?

Regards,
Maciej

Received on Wednesday, 2 May 2007 19:28:43 UTC