Mh… I don't like it, they are after all inline-functions, and as a bonus help people, that don't know much about the headers to build them.

Can you show me a bin-diff, that this actually brings any advantage?

I argue that these inline functions bring any advantage for no-one.

This PR tries not optimize code size but keep the API slim and I don't see any advantage in a do_foo_xyz(..-) function as a wrapper for do_foo(xyz, ...).

I see it, because it is convenient (at least for me).

If we do it your way, gnrc_icmpv6_echo_build() would need a check if the type is correct btw.

I don't see the necessity for that. Whoever calls this function should not what she/he's doing.

I think this is a point of contention between us. If you want this merged, I think you have to reassign :D

Did anyone of you had a change of heart since the last discussion? Personally, I also would let go of the inlined wrapper functions and opt for a slimmer API. So ACK from my side

I still think it is inconvenient, but since it seems that I'm overruled let's do it.

travis approves and everybody seems to be happy. GO

I'm not happy, I'm just complying to an argument I'm tired of having :P