Thanks for the quick replies on both questions! I've written it to be tolerant for inputs (y/1/yes/YES/true/TRUE), spec-compliant for outputs. --Ed On Jul 5, 2006, at 5:16 PM, Daniel Templeton wrote:
Ed,
There is a tracker for that issue. It will probably not be corrected in the near future, so just follow it to the letter for now.
Daniel
Ed Baskerville wrote:
Got another question...
In the C spec 1.0, the two boolean-valued attributes, drmaa_join_files and drmaa_block_email, have different string values for true and false. drmaa_join_files says to use 'y' or 'n'; drmaa_block_email says '1' or '0'.
Should I follow this to the letter as an implementor of the 1.0 C spec?
The Java and IDL specs use actual boolean values rather than strings; will a later version of the C spec begin supporting non- string attribute values? If not, will these be made consistent?
Thanks, Ed