These Enterprise Reporting Tools attempt to overcome the inherent
limitations in the
database dependent Enterprise Reporting Tools by adding a
Third Generation Language (3 GL). This means
that any report that is not merely a formatting exercise is going to be written using a 3GL. Usually a variant of BASIC or in some cases PASCAL. This is putting a
pretty face on old technology. In fact, if you were going to chose a 3GL for report
writing BASIC and PASCAL would probably not be at the top of the list. COBOL is a better
and more productive 3GL language for writing reports than either BASIC or PASCAL. So
from an enterprise perspective, these languages put a pretty face on reporting languages
that are less productive than COBOL. This is not a direction
that promises significant productivity gains.
The issues with this class of tool are:
- Productivity issues in developing reports. Because
these Enterprise Reporting Tools rely on an inefficient 3GL,
productivity is adversely effected along with responsiveness.
- Training staff. Staff will have to be
trained in the 3GL used by these Enterprise Reporting
Tools. This is an expense and also a
hidden cost, because the skills building phase is expensive.
- Cuts out End-User development. Instead of providing
tools that an end-user can work with, they now are just as
dependent on the IT staff as when most reports were COBOL
based. Sure they can do the easy reports, but most times
they need something more. The skills required to do this
are beyond most users. Also, watch an end-user's eyes
glaze over as you explain the difference between inner and outer
joins.
- May have issues with dataset size depending on what is being attempted.
This not only causes errors, but run time issues as well. If
the Enterprise Reporting Tool requires that a large dataset needs to be
returned to the users workstation, than not only does network
traffic go up, the end-user is often looking at a 'locked'
workstation during the transfer. This is a
productivity killer.
- These Enterprise Reporting Tools are also dependent on SQL so that legacy data is
not accessible. Most large organizations
have much of their data in such systems.