On Wed, Mar 2, 2011 at 1:25 PM, Anton Prowse <prowse@moonhenge.net> wrote: > On 02/03/2011 20:23, Alex Mogilevsky wrote: >> I guess it is fine with me if for the purposes of the spec we mention >> "box tree" if it helps to get clear definitions but it doesn't really >> need to exist in implementation... If analogy with anonymous blocks >> in text is any good, those definitely don’t need to be implemented >> while being fully compliant with spec behavior. > > I've always been rather fascinated by this idea. Can an implementation > claim to be compliant if it produces the same rendering in all cases as if > it did implement these "invisible" abstractions, even if it actually doesn't > implement them? I think this should be made clear in the spec. Yup, and in fact I depend on this in places; it allows me to describe algorithms that are inefficient but clear and simple to understand, and depend on implementations to actually implement a more complex and performant algorithm. ~TJReceived on Thursday, 3 March 2011 17:52:45 GMT
This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:38 GMT