Re: Conformance Criteria (was: remedy for click event)

On Sep 21, 2009, at 1:16 AM, Doug Schepers wrote:

> Hi, Folks-
>
> In response to recent threads about deprecation and conformance, I  
> adjusted the definition of 'deprecated' not to exclude UAs from  
> supporting deprecated features, added a definition for 'obsolete',  
> cleaned up and expanded the details of conformance criteria (with  
> specific conformance classes) [1], and added warnings and links to  
> the definition of 'deprecated' to each deprecated event (not just  
> their interfaces, as before).

It seems like deprecated and obsolete have the exact same effect on  
conformance:

- MAY be implemented by implementations
- SHOULD NOT be used by authors

As far as I can tell, the difference is only in non-normative "flavor  
text". Given this, I think we should have only one concept, so people  
don't have to wonder what the difference is.

>
> I also added an explicit warning about keyCode/charCode being  
> obsolete in the section on Keyboard Events.

keyCode and charCode are used a lot in existing content. I think it  
would be unwise to say that they are optional for implementations,  
since any implementation that wishes to handle general public Web  
content will have to implement them. It further does a disservice to  
implementations to leave keyCode and charCode undefined. Also,  
keyIdentifier is not yet reliably implemented in current browsers, so  
a Web page script that wants to figure out what key was pressed has to  
use keyCode or charCode. Under these circumstances, it seems like  
wishful thinking to declare these attributes obsolete.

Regards,
Maciej

Received on Monday, 21 September 2009 15:03:57 UTC