template<typename> void foo();
template<typename = T> void foo();
template<typename> void foo() { }

Since the default template argument isn't always present (and may not even be declared in some translation units where the function template is declared) it would not make sense for it to be part of the signature. Otherwise the signature for the same function template would be different in different translation units.

@jwakely , you seem to agree we need to be more precise than just saying template-head in the definition of signature?.

Since the default template argument isn't always present (and may not even be declared in some translation units where the function template is declared) it would not make sense for it to be part of the signature.

It's not clear to me how intentional that is. For function templates, we have [dcl.fct.default]p4:

"For a given inline function defined in different translation units, the accumulated sets of default arguments at the end of the translation units shall be the same; see 6.2."

... which seems to suggest that someone at some point thought that the ODR implied this requirement. (And if so, presumably it would also imply a similar requirement on default template arguments.)

Editorial meeting consensus: Add "excluding default template arguments" in parens everywhere we refer to template-head in a definition of signature would be fine.