I suppose I'm comparing this to the openssl method where there's a per-thread queue of error codes which can then be translated to strings. You might get 0 or 1 back from RSA_verify(), for example, but if you want detail you have to go into the per-thread error stack, extract codes, and translate them to strings.
It may be the case that the GNUTLS equivalent functions are more descriptive. If that's the case, then I probably don't need this capability after all. -MSK _______________________________________________ Help-gnutls mailing list [email protected] https://lists.gnu.org/mailman/listinfo/help-gnutls
