--- Alexandre Julliard julliard@winehq.org wrote:
Surely there's a better way to fix this than by forcing generation of the header even when it's not requested.
Certainly, and I did this partly out of laziness. But I also notice that midl generates the header every time as well. At least, "midl /client stub foo.idl" generates foo.h as well as foo_[ip].c and dlldata.c. And that brings up a further question: should widl flags be made (somewhat) compatible with midl ones to ease porting?
--Juan
__________________________________ Do you Yahoo!? Make Yahoo! your home page http://www.yahoo.com/r/hs
Juan Lang juan_lang@yahoo.com writes:
--- Alexandre Julliard julliard@winehq.org wrote:
Surely there's a better way to fix this than by forcing generation of the header even when it's not requested.
Certainly, and I did this partly out of laziness. But I also notice that midl generates the header every time as well. At least, "midl /client stub foo.idl" generates foo.h as well as foo_[ip].c and dlldata.c. And that brings up a further question: should widl flags be made (somewhat) compatible with midl ones to ease porting?
Well, there can be a set of compatible flags, but it's very important that there is a way to generate each individual file separately, otherwise it's pretty much impossible to write a correct makefile.