Thou shalt have no other gods before the ANSI C standard 1581

On Thu, 24 Feb 2005 23:16:08 -0500 in alt.folklore.computers, rpl

The in band return code is a NULL pointer, and the usual idiom is:

if (!(p = malloc( sizeof *p ))) { perror( program ); exit( EXITFAILURE ); }

in as few places as possible.

Also great dependence on what kind of defects measured when. Do these numbers include defects in: requirements, design, simple typos on the first compile, all test stages, post-implementation defects found after deployment; are only logical errors considered defects; are defects counted multiple times if they affect multiple stages?

IME it's not unusual to proceed with design and development before every requirement detail has been agreed upon, as long as everything has been reduced to, at most, agreed upon specified alternative possibilities with limited costs. It's more cost and schedule effective to proceed with design, development, and testing, including all the alternatives if no decision has yet been agreed upon, and add runtime parameters for selecting between exclusive alternatives, which may become configuration parameters, once the end users have had time to consider the alternatives, or see the results of processing with different alternatives selected.

-- Thanks. Take care, Brian Inglis Calgary, Alberta, Canada

fake address use address above to reply