|
Section 1 - Introduction
Since 1990 SNZ has been aware of the Year
2000 issue and its possible ramifications for
a national Statistical office. For business
reasons starting in 1993 the SNZ IT programme,
a major redevelopment in Statistics New Zealand's
Information Technology processing infrastructure,
has been in progress and has assisted in addressing
the Year 2000 issue by replacing earlier legacy
systems which were mainframe based with a client
server architecture. During the programme emphasis
was placed on the correct usage of date processing
which would result in date manipulation and
calculation processes which were robust enough
to withstand the rollover from Year 1999 to
Year 2000. Given this background SNZ is in a
more fortunate position than some other organisations
who will not have time to replace their legacy
systems. Our predominant environment at the
time of writing this is LAN based client server
with large scale deployment of Lotus Notes and
some remnants of mainframe systems remaining
due for final retirement on 31 March 1999.
As already outlined SNZ has been in the somewhat
enviable position of having most of it's technically
higher risk systems replaced but there are still
a number of issues requiring our attention.
As progress on our Year 2000 effort continues
we continue to be surprised by the actual scope
of the requirements of this project as new information
and materials come to hand.
This paper is broken down into three sections,
this introductory section, a general background
on our scope of the initial Inventory and Awareness
project, experiences with three specific issues
surrounding year 2000 problem solving which
we did not anticipate initially which is entitled
- experiences.
Section 2 - General
background and precise of our scope of the initial
Year 2000 Inventory and Awareness project
This is essentially the introduction from our
Year 2000 Inventory and awareness scoping document
produced August 1997 where we cover the ground
rules for attacking the Year 2000 Problem. Our
approach has been to decentralise responsibilities
as much as possible at the senior middle management
level which will result in our being able spread
the workload of Year 2000 project into areas
where we could expect to find relevant expertise.
Two evolutionary developments which have evolved
since this original scoping introduction have
been the benefits of using Lotus Notes for awareness
(we have a Year 2000 Problem Links page which
acts like an internal web site directing people
to resources and other materials) and Inventory
processes ( Lotus Notes databases containing
Inventories are now more fully featured making
their use and the logistics of Inventory a great
deal easier).
Commentary
The amount of material now available for Year
2000 from consultants, vendors and industry
organisations is vast. The material referenced
in the production of this scope has been mainly
confined to Gartner group research on the Project
Management issues concerning Year 2000. Gartner
identify that the two first stages in Year 2000
Projects are Awareness and Inventory. This project
is intended to cover these first two stages
only - the complete co-ordination Project's
scope will be derived from the outputs produced
by this project.
1. Awareness campaign
1.1 Goal Overall
The goal of this campaign will be to make Group
and Divisional managers aware of the Year 2000
problem and how to identify possible problem
areas so as they can contribute to the Year
2000 Problem Register. They will be informed
of Year 2000 Problem Areas already identified
during this process to assist them.
1.2 Accountabilities and
Responsibilities.
Divisional managers (senior middle managers)
will be made aware that they are
- Responsible for their
own areas unique entries in the Year 2000
Inventories
- Accountable for ensuring
that they have taken all reasonable steps
to identify all their likely problem areas
into the Year 2000 Inventories. This includes
ensuring that possible Year 2000 problem areas
which would normally be considered as being
of use to them but outside of their control
are identified. This is to cover situations
where corporate records which are used to
populate the Year 2000 problem register initially
such as the IT Tools register are incomplete.
(e.g. If Excel is missing from the Tools register).
- Accountable for ensuring
that any contracts or agreements that they
enter into with external parties cover Year
2000 compliance requirements.
1.3 Guideline, Resource
and Background awareness.
The awareness campaign will also
- Educate as to the Year
2000 problem background of likely causes and
possible effects.
- Educate on legal outcomes
for the organisation
- Keep divisional managers
informed of where to find resources and how
to contribute to the Year 2000 Resource Register.
2. The Year 2000 Resources
Register
This Register will be set up as a Notes Database
and will act as a Knowledge Base or reference
point on where to find Year 2000 Resources.
Its main target audience will be divisional
managers looking for resources to meet responsibilities.
The register will be populated by
- Vendor product offerings
- Industry commentaries
- Where to find particular skills by Tool/Application
The Year 2000 Resources register will be maintained
by the Year 2000 co-ordinator but will receive
entries from any Divisional manager who feels
that they have or know of a useful tool.
3. Inventories.
This section from the original scope is
no longer relevant. Early practical experiences
with conducting these inventories led to the conclusion
that maintaining and enhancing existing Inventories
was a better proposition than creating a specialised
Year 2000 problem register. As a result resources
targeted at this part of the project have actually
gone into supporting our existing Inventory and
register structures and the inventory process
has revolved largely around getting these up to
date and accurate. Two new Inventory classifications
have had to be created due to the fact that we
did not have them previously - they are End user
computing and External Electronic data interfaces.
Both are described in more detail later in this
paper.
4. Concluding Report -
final output from this project as a precursor
to the full Year 2000 Project scope.
This report will be produced of the expected
scope of the Year 2000 Project and presented
to the Group manager Information project. The
report will form the basis of the Year 2000
Project scoping document. It's exact format
will be defined as this Awareness and Inventory
project progresses - results from the Year 2000
Problem Register may dictate a format not expected
at this stage.
Section 3 Experiences
- The Year 2000 Problem solving Labour Market.
Gartner and other sources are predicting an
exponential increasing demand for Year 2000
problem solving skills both at the Project management
and technical skill levels right through to
the Year 2000. If these predictions are correct,
there is anecdotal evidence supporting this
already, as the market demand factor for these
skills increases so will price unless new people
with the right skills can be introduced into
the Year 2000 Problem solving market quickly.
There have been many examples especially in
more open market economies like New Zealand
where dire predictions of shortages have been
overcome by the supply side of the market demonstrating
elasticity well beyond expectations and filling
the gap much more rapidly than would have been
expected. It is the hope of recruiting and senior
management in all organisations with known and
potential Year 2000 problem solving labour market
requirements that this will be the case. To
a degree I think those hopes are well founded
- but only to a degree and care will need be
taken to ensure new entrants on the supply side
are capable or meeting requirements after what
will be of increasing necessity a shorter and
shorter training and breaking in period as Year
2000 approaches.
As most Statistical agencies are Government
owned and managed they often face a higher degree
of centralised wage fixing and scale setting
than their private enterprise competitors in
labour markets generally. In New Zealand, as
I suspect is the case in other countries, this
pressure on their ability to compete in more
competitive labour markets is met in three ways
- By applying resources
into upgrading existing staff skills
- Offering greater security
to staff
- Purchasing skills at the
lower end of the market.
Upgrading skills can be a lengthy process
and has a heavy reliance on acquired knowledge
being available both of which can be seen as
limiting when facing an immovable and rapidly
approaching deadline without precedent. Offering
greater security is now no longer as feasible
as it once due to labour market and state sector
reforms.
Some possible alternative approaches may need
to be investigated if Gartner and others are
correct in their forecasts and supply is not
as elastic as is hoped in some quarters. Some
of these alternatives could include
- Giving your Year 2000
problem solving staff the ability to take
on work for external organisations which allows
them to maximise their returns for the limited
period available.
- Providing Year 2000 problem
solving staff with a temporary 'bounty' element
on their existing contracts or terms of employment.
This could include a 'bonus' element (colloquially
referred to as Golden hand cuffs) which is
available to them once the Year 2000 Problem
solving project is completed and proof of
results is available ( so it would be prudent
to make it available in the last quarter of
the year 2000)
- Outsourcing as much of
Year 2000 problem solving activities as possible.
There are also be some practical things that
can be done to help reduce your own demands
on the Year 2000 problem solving market
- postpone major changes
in organisational restructuring which could
see redundancies and departures amongst staff
experienced in relevant areas (e.g.: IT Staff
versed in older legacy systems which are more
likely to have Year 2000 problems).
- gather as many personnel
into and involve them with the Year 2000 Problem
to help spread your risk.
- Choose capable employees
with a longer than average tenure.
- Choose capable employees
who have a vested interest in remaining with
the organisation such as a pending entitlement
(e.g.: Superannuation, long service leave).
In Summary
A number of steps can be taken now to minimise
the risks associated with the predicted increases
in demand on the Year 2000 Problem solving labour
market. Throwing money at this market - when
and if the problem of supply shortages becomes
acute, may not be an option and even if it is
it may not be an ideal solution as new entrants
to the market may not be able to produce the
required quality of outputs given timing pressures.
Start planning now if you have not already done
so.
Contd/ Section 3 Experiences - External
Electronic data interfaces (EEDI).
Information Technology professionals often
comment on the fact that the results of many
hours work in developing software can be contained
in finished form on a portion of a three and
a half inch diskette which appears insignificant
in its physical surroundings. Statisticians
employed in National Statistical offices must
feel somewhat the same when after processing
vast amounts of information a single figure
such as a price, cost or growth index amounting
to no more than 5 digits is produced. The amount
of information required to produce most of these
indices and Statistics is greatly underestimated
as is the cost of gathering the raw data in
the first place.
Responding to demands to reduce costs and
produce a wider variety of outputs which often
require more raw data Statistical agencies have
often resorted to capturing raw data from external
sources electronically. Much of this Electronic
raw data carries dates supplying timing information
on when the data was gathered and therein lies
this particular areas potential to develop Year
2000 processing problems. If the year portion
of the date is two digits then the party sending
the information and the party receiving the
information must have a common understanding
of how the century portion of the date is to
be 'inferred'. Inference as a term is not to
be confused with Statistical inferencing techniques
but is used to describe inferring a century
value from the year value portion of the date
as described in Rule 3 of the British Standard
of Year 2000 Conformity.
There are some issues surrounding the use
of two digit dates in external electronic data
interfaces.
- Both parties must agree
on the way that the century value will be
inferred from the two digit year value. This
can be as simple as saying any year value
within 50 years of the current year value
will be inferred as belonging to the appropriate
century (also called the 50 year window).
The inferencing techniques will vary from
partner to partner and there must be a repository
set up somewhere for recording the inferencing
technique for future reference.
- Some organisations have
decided that they will convert all dates to
four digits. We have not adopted this approach
using Rule 3 of the British standard on inference
to minimise costs. We will encourage our external
partners in electronic data interfaces to
adopt an inferencing technique rather than
convert the format of the electronic data
interface to four digits. Converting date
formats to four digits would require redevelopment
and re testing which may not, given the nature
of the information being exchanged, be the
most cost effective or necessary approach.
Obviously when working through external electronic
data interfaces early contact with data exchange
partners is desirable. A necessary pre-requisite
to establishing that contact is of course inventorying
what external data interfaces are actually active.
Two key points on building External Electronic
data interfaces inventories from our experience
indicate that
- Do not be surprised
if you have many more interfaces than you
expected. When planning our EEDI Inventory
I was expecting a count of twenty to thirty
actual live interfaces but have now recorded
over sixty interfaces active with both Public
and private partners.
- Make sure when you make
your inventory that those responding have
a clear understanding of what an external
electronic data interface actually is. It
was surprising how many Inventory location
managers in our organisation supplied details
of regular yet unformatted email communications
and one time job request data exchanges.
In summary
By way of external relationships External
electronic data interfaces can pose several
problems which may require working through with
your external data exchange partner. There are
often more interfaces active than at first expected
once an inventory is taken. The inventory process
needs to have very clear definitions supplied
to avoid mis-recording of items that fall outside
of the area you actually need to target.
Contd/ Section 3 - experiences
- End user Computing.
Four factors have contributed to the rapid
growth of End user computing solutions in Statistical
organisations.
- Statisticians and other
specialists employed in National statistical
offices often have a higher degree of computer
literacy ranging right through to computer
applications development.
- Increased deployment
of desk top processing tools capable of running
program code and macro instruction code which
is developed for those environments (eg: SAS,
Excel which has both Macro and full third
generation program support in Visual Basic
for applications).
- Budgetary constraints
in Information Technology service providing
areas have reduced resources available from
those areas encouraging a 'do it yourself
approach'.
- Many professional Statisticians
and Economists have a degree of intellectual
curiosity which motivates them into investigating
and then deploying end user developed computer
techniques and applications.
End user computing provides special problems
for Year 2000 problem solvers. First and foremost
is inventory - many of these end user applications
are scattered around desktops and do not have
identifiers attached to them which facilitate
inventory taking - furthermore they are sometimes
inadvertently hidden from inventory takers if
there is a perception that the end user application
will be reported or forfeited. Secondly their
mission criticality is often under-estimated
- the statement "this doesn't produce a critical
output" can be found to be misleading when an
end user computing application's output is used
as input to an area that does eventually produce
a critical output. Thirdly there is a misconception
that because the program platform (eg: SAS,
Excel) running the macro or applications code
is declared as Year 2000 compliant then the
macro or application code will be year 2000
compliant as well - this is definitely not the
case if two digit dates have been used in the
end user developed code or macro.
End user computing creates special problems
for Year 2000 problem solvers only some of which
have been mentioned here. The taking of Inventory
is extremely difficult and yet critical. Our
approach, still in development, is to use a
Lotus Notes application which can be 'pointed'
at an end user computing application through
a standard dialogue allowing it to be recorded.
With every other Inventory now virtually completed
End user computing is proving to be a real challenge
but I believe one that can be overcome.
Summary:
Do not under estimate the dangers posed by
or the difficulty in inventorying end user computing
applications. End user computing can be a sensitive
subject and a degree of assertiveness will be
required to ensure that a full and proper inventory
is taken.
|
