Re: http content type authoritative for object data?

Tab Atkins Jr. wrote:
> On Thu, Dec 10, 2009 at 4:51 AM, Julian Reschke <julian.reschke@gmx.de> wrote:
>> So if what the spec requires isn't your top consideration you probably
>> shouldn't care whether it requires or just allows ignoring the HTTP
>> content-type.
> 
> Please don't be ridiculous.  The meaning of that statement is that the
> spec is less important than reality when making decisions.  That
> doesn't imply that he or any other browser developer 'doesn't care'
> what the spec says.  It's quite the opposite in fact; it's in

I said: "do not care whether it requires or allows". That's different 
from not caring at all.

> everyone's best interests that the spec declare something that matches
> reality, so that corner cases aren't different between impls and they
> don't have to waste QA time and money on checking if the way other
> browsers do things is better.

On the other hand it's also important that the spec doesn't disallow 
something that may be the cleaner approach.

> The willingness to ignore the spec when it's incorrect is not a
> statement that the spec is worthless, as long as the spec is willing
> to update itself.

The spec doesn't update itself, we do. And we do that by discussing the 
current contents, looking at alternatives, and then making decisions.

The thing I proposed was that the spec *allows* UAs to follow the HTTP 
Content-Type (similarly as the content sniffing internet draft already 
does). Do you have an opinion related to that proposal?

Best regards, Julian

Received on Thursday, 10 December 2009 14:42:54 UTC