This chapter is from the book
This chapter is from the book
This chapter is from the book
Do We Need a Grand Unified Tool?
The problem definition is too big for one tool or person to maintain, so there appears to be a dilemma. The full complexity of the problem needs to be embraced, and an understanding is required of everything that's around, including the existing IT and business environments. But all that information needs to be pulled together so that the Views aren't discrete or disconnected.
Many people have argued for tool unification as a means to achieve this, to maintain all these connected Views in a single tool and, thus, enable a single documented version of the truth to be established and maintained. But that is missing a vital point about Views.
As explained in Chapter 2, "The Confusion of Tongues," Views need to be maintained by people in their own way, in their own language. Imposing a single tool will never work. Simply too many preferred perspectives, roles, and prejudices exist within our industry to believe that everyone is going to sit down one day and record and maintain their Views in one specific tool.4 If such combinations of Views into single multipurpose tools were possible, desirable, and usable, then it is arguable that Microsoft® Office user interfaces Word®, PowerPoint®, and Excel® would have merged long ago.
Moreover, these integrated approaches that have been at the heart of traditional tooling are usually pretty poor at dealing with ambiguity or differences of opinion. On large projects with many people working on the same information, it is not unusual to have formal repositories that enable people to check out information, make changes to it, and then check it back in. Such systems prevent two people from updating the same information at the same time, which would result in confusion and conflicts. The upshot of this approach, however, is that the information that is checked into the repository is the information that everyone else is then forced to use. The implications of your changes are not always apparent to you—or perhaps immediately to your colleagues, either. Maintaining a single source of truth when hundreds of people are changing individual overlapping elements is less than straightforward. A change made by one individual can have serious consequences for many other areas of the project, and no mechanism exists for highlighting or resolving ambiguity—whoever checks the information into the repository last wins!
In summary, grand unified tools are to software engineering what grand unified theories are to modern physics—tricky to understand, multidimensional, and elusive, often involving bits of string. No one has created a single tool to maintain the full complexity of a complex IT project. Likewise, no one will do so unless the tool enables people to maintain Views in their own way, in their own language, and to identify and deal with ambiguity cooperatively.