Null characters are almost always avoided in interchanged plain texts.
This is not a practicle problem. The use of nulls as significant
characters is extremely exceptional
Yes, but still I think that the BOM ambiguity needs to be documented. If
it already is, the documentation isn't visible or hasn't been visible
enough to the public.
After all, Unicode is a reference standard, among other things. It's
like the detail about parsing of >> in C++: this sort of thing might not
matter in everyday coding, but as soon as someone manages to abuse a
security hole by going through silent (un(der)documented) assumptions,
there is a problem. Even if not, it's nice to have things be
descriptively complete in the IT world.
Stephan
