Daily Newspapers each do things in very unique ways. From the workflow they use to the system features they consider critical, they are all different. But customization has a price. DTI has employed an architecture that it believes can address this dichotomy between the cost effectiveness of a standardized system and the requirement that specialized needs be met.
February 2001 - Daily Newspapers each do things in very unique ways. From
the workflow they use to the system features they consider critical, they
are all different. The bigger the newspapers, the more different they are,
it seems.
When it comes to selecting new systems, the selection of vendors often comes
down to “will they give us what we want, in the way that we want it, by the
time that we want it”. Sometimes, selection committees are astounded that a
vendor doesn’t have a feature they consider critical and wonder out loud how
that vendor even manages to be in business without such a feature. Yet many
of these features are not important to other newspapers, and surprisingly
often, are completely contrary to others’ way of producing a newspaper.
Because of this, there is enormous pressure on vendors to customize their
systems to meet the specific needs of their potential clients. But
customization has a price. Of course it costs more up front. But it also
costs more forever. Once a system is customized with special features it can
only be supported well by people who are familiar with those features and
how they might differ from other versions. Support and trouble-shooting is
reduced to a few specific people. It becomes more difficult for a newspaper
to share its experiences or get help from sister newspapers when their
systems have real differences.
Standardized upgrades, which are common in the shrinkwrap world, are
difficult with custom systems because all of the custom features must be
added into the new version on a one-off basis. At some point, new features
can only be added to a customized system with even more one-off
customization. A dead end is eventually designed in by extensive
customization.
DTI has employed an architecture that it believes can address this dichotomy
between the cost effectiveness of a standardized system which uses mass
market software such as Adobe’s InDesign and the requirement that
specialized needs be met that require customizing.
The core of DTI’s system is completely standard, offering all newspapers the
same applications and database architecture. Changes to this core are only
made with the next major upgrade that goes to all customers.
However, there are three major customizing capabilities on top of that core
system.
First there is a plug-in architecture, created by Adobe, that enables
special features to be added that don’t break the core system. (One of the
strengths of Quark was that it had an extension architecture that also
allowed specialized plug-ins. But it was very limited and the extensions
were specific to a version of Quark. They had to be upgraded with each Quark
upgrade.) InDesign plug-ins enable a much broader range of customizing
because they can become a part of the InDesign program on equal footing with
other InDesign components which are themselves simply plug-ins. (85% of
InDesign are plug-ins to a very compact core.)
DTI has yet another level of customizing to its systems through extensive
scriptability. Using Applescript on the Macintosh or Visual Basic on
Windows, DTI’s applications can be customized through scripting which can
automate many tasks and even add special features. Associated with
client-side scripting is server-side customizing through the use of Java
stored procedures that can run within the database and can add considerable
functionality to how information is handled and flows through the system.
This kind of customizing is within the capability of nearly any newspaper
and DTI has a large portion of its user group actively customizing their
systems using scripting.
While plug-ins are powerful, they are very difficult and time-consuming to
create, and they are beyond the scope of most newspapers in-house
development capability. Scripting is much easier and can be used for a great
many specialized features. But it has definite limits.
So DTI has added yet another method of customizing in an attempt to get ever
closer to the ideal of having a standard system that can be stretched into
meeting specialized needs without forcing the core into a one-off version.
DTI has written most of its system beyond InDesign in the Java programming
language. A feature of Java, which is a very object-oriented language, is
the ability to incorporate new objects on the fly. These kinds of objects
are called Java Beans. Through the use of a special designer tool which
works much like a traditional report writer, newspaper personnel can look at
their database and create views of the data which correspond to their
workflow. Deadline tracking, which is handled very differently from
newspaper to newspaper can be customized this way, including sending alerts
to people based upon actual progress. Daily budget reports can be created
which combine stories and photos based upon any user-defined criteria. These
are just two examples of what can be created. But when they are created, the
designer tool will create a Java Bean that incorporates the desired
functionality, and this bean can be added directly into the DTI system where
it becomes just another menu item. Major new features can be added in a
custom way to the DTI system, which operate just like core features. Yet
they can be done at low cost by the newspapers themselves. They can be
supported and maintained without dependence upon DTI.
The holy grail of having a standardized system meet every imaginable need
has been called “mass customization”. DTI hasn’t reached that point. But,
with three major customizing capabilities supported as part of its core
system which extend to both the client applications and the server-side
databases, DTI believes it has gone further than any other supplier to daily
newspapers in providing a standard system that can be customized to meet new
requirements on an ongoing basis without becoming a one-off dead end system.