untitled
End-user Programming
by Silvie Spreeuwenberg
'End-user programming' is a term that has been introduced to me by the editor
of the IEEE magazine SOFTWARE in his column, "The dangers of end-user
programming" (July / August 2004). Warren Harrison defines end-user programmers
as "mechanical engineers, doctors, physicists, teachers, and accountants.
They are marketing assistants, receptionists or commodity brokers. They are,
in short, individuals who taught themselves to program." The idea of having
'end-users' -- people with no background in software testing or development -- directly
manipulate software that is critical for a company frightens the software professional.
Harrison's conclusion is that end-user programming is a threat for mission-critical
systems from the perspective of both correctness and security, and he asks himself:
"Can it be true that software manipulating my credit history could have
been written by an accountant with no concept of software testing or development
processes?"
Observation 1
Given my observations at several companies, I think that the answer to this question
is "yes." Typical end-user programs are spreadsheets and small database
programs. They start small but can evolve into strategic programs for a company. |
What is the motivation for end-users to get involved in software development?
After all, end-users are educated for a different subject matter; they did not choose
an education as programmer, so why would they start building a system, investing
so much (free) time in learning how to code in Microsoft Excel, Access, or Visual
Basic, and overcome all the barriers involved with end-user programming?
These programs are born out of a need: a businessperson has some particular
knowledge that he wants to re-use, automate, or share with colleagues. IT departments
seem not to be able to accommodate such needs for reasons that are unclear.
When IT does build an application to accommodate the needs of the business, all rules
(knowledge) are hard-coded in the programs, business people are not able to inspect
the rules, and changes to the business rules take a very long time.
Observation 2
There is a legitimate source of frustration by business people about the software
development process. |
The business rules approach offers a range of methods and techniques to improve
the collaboration between IT departments and business departments. A business
rules management environment targeted at end-users is a better idea than end-user
programming because this environment should limit the options of an end-user to only
change rules, while security aspects, for example, are handled by the business rules
management environment. But there is much more needed to succeed.
Observation 3
It is very hard to have end-users create valid, consistent, well-defined, thoroughly-tested
business rules. Many business rules end-user maintenance projects failed because
of lack of support for these aspects. This problem is not only a problem of
'end-users' but of humans in general. Humans have a hard time comprehending
more than 20 rules that have numerous complex interdependencies, resulting in rules
with undesired effects or that are difficult to maintain and understand. |
Given the facts that regular software development environments do not have specialized
support for verification and validation of business rules and that communication
between business and IT often involves misunderstanding, I observe:
With proper support for verification and validation of business rules in a business
rules maintenance environment, the correctness of software with respect to business
rules will improve rather than be threatened. The time we gain with serious
end-user programming, supported with a good business rules maintenance environment,
can be spent by software professionals on the security and stability of the systems
supporting the rules of the business. The business rules approach offers a
viable approach to support end-user programming.
|
Drs. Silvie Spreeuwenberg has a background in artificial intelligence and is the
co-founder and director of LibRT. LibRT helps customers assess and improve
the quality of their business rules. Silvie's experience with business rules
modeling has resulted in the development of tools and techniques to increase the
quality of business rules. She writes, "We believe that one should focus
on quality management of business rules to make full profit of the business rules
approach." LibRT is located in the Netherlands; for more information
visit