My Quest GPM/PowerShell/PowerGUI grind continues – I’ve gotten the export working fully, and its pretty well designed. I’m determined to make each script into a usable set of cmdlets. When I finish the entire project, I’ll post the final code here. Right now, everything revolves around Quest’s GPM, but there is some pretty involved code which could be re-purposed for other tasks (perhaps even straight AD GPO manipulation).
But what I haven’t really considered until early this week are the actual requirements. And then, that same consideration needed to be made by the PM and the Dev guys helping me. And that’s because the requirements were not being observed in the original pass of what was developed. Ian D (my “customer” inside Quest) actually did something I hadn’t done in quite a while – he made me review the dev requirements before we did anything else on Tuesday.
And those requirements, simply put, was that the user getting the set of imported files should be able to run a single script/executable and all the ‘magic’ happens unattended. No intervention should be needed, and if there’s a problem, the user is not going to be able to help. So just dump a log file, and let the user send the log back to ‘the mother ship.’ Even if the user is able to help, he will have little to no rights in AD.
That one rule (that the user will have NO rights in AD) caused me to restructure and rethink some things on Tuesday. I’m now 80% done, and hope to have something in a deliverable draft on Monday afternoon. And I’ve taken on board what Ian did to get me focused on Tuesday – I wrote out the requirements and have them in front of me as I’m coding. I forgot how easy it is to lose focus and ‘go dark’ if you don’t keep the goal in sight.
Comments on this entry are closed.