I assume that by values you are referring to field identifiers? The reason this is strongly discouraged is that when field identifiers are omitted, the compiler applies negative field identifiers in descending order to each field, in the order encountered.
If you use negatives for ALL your parameters, you'll be fine, but if you have a parameter in there with no field identifier and then you make some changes to the others, the assigned identifier for the unspecified field might change. Basically, it is always best to specify a positive identifier for every single field. -----Original Message----- From: [email protected] [mailto:[email protected]] Sent: Thursday, March 11, 2010 7:28 AM To: Thrift Developer; Thrift User Subject: Prevalence of -ve values in function parameter/exception ids How often do people use negative values in function parameter/exceptions returns? I get the impression that it is strongly discouraged even though it is permitted by the IDL. Thanks Mayan
