[whatwg] Re: several messages

[Matthew Raymond]

Greg Kilwein wrote:
> The key is _easily_ test the fallback content.  For those authors who 
> just don't care, there's no default fallback content, which I think is 
> worse than none at all.  Besides, if the upper-end authors really want 
> to replace <input type="date"> with three select boxes in legacy UAs, it 
> can be done with some script without imposing some sort of 
> comment/uncomment testing method on lower-end authors who just don't 
> know or care.

    The easy way to test pages for legacy fallback is to use a legacy 
UA, but you were specifically asking how someone would test markup in a 
_WF2_ user agent. In the end, the only way to know for sure that 
something degrades gracefully on a legacy browser is to check it with a 
legacy browser. For instance, you would not be able to test if your 
client or server-side validation of the date works on a legacy browser 
by testing it in a WF2 browser unless the WF2 browser had a feature for 
turning WF2 off, in which case <idate> would benefit in a similar manner.