Re: CR exit criteria and features at risk for HTML5

On 8/16/12 10:50 AM, Sam Ruby wrote:
> (once again, assuming that the assumption spelled out in the previous
> email[1] is operative for the purposes of this discussion)
>
> Such parts can be enumerated, and then either stubbed out or removed.
> This process can be repeated as often as necessary.
>
> I'll assert that large parts of provably interoperable behavior will
> remain.  Do you disagree?

I suspect that pretty much everything other than maybe the parser 
depends on the event queue setup and on navigation in the end....  We 
can perhaps work around this by just saying that parts of the event 
processing model and of navigation are undefined, I guess.

> Jame's default assumption[2] seems to be that people would prefer to
> publish something known not to work than to exclude it.

That really depends on the uses the REC will be put to.  If it's going 
to be used as a club, with demands that UAs implement the REC exactly as 
written and legislation to that effect (not hypothetical; has happened 
with specs before), then it's better to exclude things than to publish 
wrong things.

-Boris

Received on Thursday, 16 August 2012 15:47:12 UTC