On Mon, 2014-09-29 at 20:21 +0000, David Woodhouse wrote: > >> For the string ownership thing I think we're almost there, and the Java > >> bits seem to be working. We do need to fix gnutls_get_cert_details() > >> which currently returns an allocated string for the caller to free. > > > > You could provide a function for the caller to use to free it. > > That requires a user of the library to use it. If we let them pass in an > optional malloc function of their own then they can happily just not > bother. I don't know if that's really worth the effort. The API would become quite ugly, and the implementation inside quite uglier as it would have to allocate and then reallocate using the user's malloc. That's only for some windows systems, and I believe it is reasonable to expect _them_ to properly know and use the provided deallocation function. regards, Nikos