Re. “Having a go at some 2003 predictions” (Jan. 3, 2003 issue, page 3): Cross training is largely a waste of money unless you are currently employed at a company that plans on moving to the technology, or more likely has already moved and is now trying to figure out why things aren’t working as advertised.
Speaking of which, I see ICAA is moving from Novell to 2000 (Jan. 3, 2003 issue, page 3). I expect that they will be in for a surprise when they finish the project. It is possible that they are still using some variant of Netware 3 from 1990, but if they are using 4.11, they had better be prepared to add 15 per cent more staff to cover the increased costs of managing such a large AD network.
This is just an estimate on my part, given the limited information, but AD can easily become an administrative nightmare if stringent controls are not maintained and standards rigourously enforced. The phrase “…believe that the learning curve will be short” does not fill me with confidence; if they try to mimic the NDS tree (I’m assuming) they had previously, it will be a disaster. The OUs in AD are not as robust as in NDS and workarounds will need to be found to provide the same functionality. The complexity of the replication model and site determination will also exceed what they are used to using with NDS. It will be interesting to revisit them in 2004 to see how things turned out.