On Tue, 23 Oct 2012, Ian Hickson wrote:
Having multiple specs means an implementor has to refer to multiple specs
to implement one algorithm, which is not a way to get interoperability.
Bugs creep in much faster when implementors have to switch between specs
just in the implementation of one algorithm.
That is true to the extent that all developers have the same objective but
we've already established that there at least two case:
a) Products which can assume correct URI (STD 66) syntax and reject errors
b) Products which need to handle human mangled input
By writing merging (a) and (b)->(a) you make life more difficult and hence
error prone for (a). I would suggest that the developers of (b) will be
better served by a clear specification of (a) w/o the (b)->(a) concerns
because that will improve their ability to validate URIs generated by
their logic as well as have meaningful discussions with folks who
assume (a).