ret->data.file.do_index = GNUNET_NO;
ret->data.file.have_hash = GNUNET_NO;
ret->data.file.index_start_confirmed = GNUNET_NO;
+ /* FIXME: what's our approach for dealing with the
+ 'reader' and 'reader_cls' fields? I guess the only
+ good way would be to dump "small" files into
+ 'rh' and to not support serialization of "large"
+ files (!?) */
break;
case 1: /* file-index, no hash */
if (GNUNET_OK !=
ret->data.file.do_index = GNUNET_YES;
ret->data.file.have_hash = GNUNET_NO;
ret->data.file.index_start_confirmed = GNUNET_NO;
+ /* FIXME: what's our approach for dealing with the
+ 'reader' and 'reader_cls' fields?
+ (should be easy for indexing since we must have a file) */
break;
case 2: /* file-index-with-hash */
if ( (GNUNET_OK !=
ret->data.file.do_index = GNUNET_YES;
ret->data.file.have_hash = GNUNET_YES;
ret->data.file.index_start_confirmed = GNUNET_NO;
+ /* FIXME: what's our approach for dealing with the
+ 'reader' and 'reader_cls' fields?
+ (should be easy for indexing since we must have a file) */
break;
case 3: /* file-index-with-hash-confirmed */
if ( (GNUNET_OK !=
ret->data.file.do_index = GNUNET_YES;
ret->data.file.have_hash = GNUNET_YES;
ret->data.file.index_start_confirmed = GNUNET_YES;
+ /* FIXME: what's our approach for dealing with the
+ 'reader' and 'reader_cls' fields?
+ (should be easy for indexing since we must have a file) */
break;
case 4: /* directory */
if ( (GNUNET_OK !=