Date:         Fri, 3 Oct 2008 14:29:55 -0400
Reply-To:     Lou <lpogoda@HOTMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Lou <lpogoda@HOTMAIL.COM>
Organization: Aioe.org NNTP Server
Subject:      Re: Macro validation must-haves?
Comments: To: sas-l@uga.edu

"RolandRB" <rolandberry@hotmail.com> wrote in message
news:bccd88b5-f2db-4df7-9d12-e4e0c4d59de8@z66g2000hsc.googlegroups.com...
> At the moment I am very interested in peoples views on "Macro
> validation must-haves". But then I see a big difference between
> "Application validation must-haves" and "Macro validation must-haves"
> since many macros might get written in response to a formal
> application requirements document. A single set of requirements might
> spawn many macros as a solution and this might be the result of the
> designer. Does the designer have to go back to the client to justify
> every macro he/she writes in response to an application specification
> document and get them to write a justification for each and every
> macro he/she might choose to use and maybe drop as the coding pans
> out?
>
> This has been stirring around in my head. The only way through I see
> at the moment is to use the SAS OQ approach such that whatever they
> supply has a test pack that goes with it to ensure the user that it is
> doing what it is supposed to do.

As I understood it, the SAS IQ/OQ doesn't assure anyone that "it" is doing
whatever it's supposed to do.  It's aimed at  assuring that every file
copied from the installation media was written to the destination drive
without error.  Every month, maybe every day, someone in the user community
finds something in the system that doesn't work the way it's supposed to.