In 1995, Ken Orr presented a slide which showed his predictions for the probable effects of Y2k for AWARENESS, DENIAL, PANIC, TRIAGE, LITIGATION.
Modified version of Ken Orr's orginal Predictions
Taking his example and projecting forward to the
current time, we are seeing AWARENESS and DENIAL moving
forward until today, and expect to project them forward
as an ongoing exercise. We have not seen any PANIC
yet. TRIAGE is yet to come.
Now that we are emerging from the "Middle Game" and
moving into the "End Game" (to steal Chess terminology)
and have a look at generating a Project Plan, starting
today, a chilling picture emerges.
The Y2k Life Cycle with mid-1998 as Drop Date Date"
There are really two project plans, one for Mainframes (the EBCIDIC scenario) and one for PC's and LANs (The ASCII scenario).
The EBCDIC scenario presupposes total replacement of the Operating System software, Utilities and Middleware. And then retrofitting application software to work in this new environment. July 1, 1998 is in my opinion the "DROP DEAD" date for such systems. The ordering cycle and product lead times would preclude success.
The ASCII scenario presumes application of patches, Setup of appropriate parameters, and also retrofitting of application software. But total replacement, and the corresponding disruption, is not necessarily an issue.
Somewhere in between are those systems where a decision has been made to migrate to a different architecture or more commonly where an organisation has both EBCDIC and ASCII components.
Another assumption is that if people do not start today, then 1 May 1998 seems to be the latest feasible start date.
NATIONAL Y2k PROJECT PLAN
Immediate Start: (1998-03-17) 470 working days
Late start: (1998-05-01) 394 working days
Drop Dead date: 1998-07-01
Support for Y2k Life Cycle: Early Start Late Start Target Milestone 1998-03-17 1998-05-01 2000-01-03 AWARENESS 1998-03-17 1998-05-01 2000-01-03 ASSESSMENT 1998-03-17 1998-07-01 2000-01-03 REMEDIATION 1998-03-17 1998-07-01 2000-01-03 VALIDATION 1998-03-17 1998-07-01 2000-01-03 TESTING 1998-03-17 1998-07-01 2000-01-03 IMPLEMENTATION OPTIONAL MILESTONES 1999-03-17 1998-07-01 1999-01-01 Year End Testing 1998-03-17 1998-07-01 1999-01-01 Euro compatibilityNow this is a very crude model. It does not even attempt to apply estimates of how long each phase should take. However it is a starting point.
The original suggestion that Y2k remediation and testing should be completed by end 1998 is probably no longer feasible. If however, 1997 year end test data is still available, an "asynchronous" test could be performed. Similarly, the 1998 year end data could be used for parallel testing during 1999.
A complicating issue is that organisations differ in their end of financial period reporting dates.
My suggestion, in order to obtain the maximum "parallelism" between phases is to chop the effort into smaller chunks, starting with Obvious Mission Critical Systems, put these into the Remediation and Testing cycle as soon as possible and continue the cycle for other systems.
For example, any systems which would be affected by the impact of the Euro currency should be put on a separate fast track to make the dealine of 1999-01-01.
The function of the National Competence Centre would be to map the schedules of the Major Governmental Infrastructual Players against this base model as a base for the National risk Assessment.
Local Competence Centres could do the same for Industry and Commerce.
There would be great PR Awareness value in making this information public.