With research tools, this is as it
should be: efforts should be focused on new, innovative ways of de?¬?ning modeling
languages, and evenempirical trials will focus on real users working as metamodelers,
not modelers. For tools intended for actual use in modeling, de?¬?ciencies in this area
are however serious. Unfortunately, the consequences of these failings re?¬‚ect poorly
not only on the tool itself but also on DSM as a whole. As there are many modelers for
WHAT IS NEEDED IN A DSM ENVIRONMENT 385
each metamodeler, the majority of experiences of DSM across the industry will be
based on the modeling tools.
Basic Functionality To be considered for real use, tools must offer at least the
basic functionality expected of any program these days, regardless of whether it is a
graphical editor, email software, or word processor. These functions include multilevel
undo and redo; cut, copy, and paste; direct manipulation of editable elements; and
printing and exporting to various ?¬?le formats.
When copying and pasting, there should also be the possibility to affect whether a
direct reference or newcopy is created.With complex objects, it must also be possible
to specify how deeply contained elements should be copied: objects in properties,
subgraphs, and so on.
Direct manipulation is particularly important for laying out diagrams so that they
are easy to read and work with.
Pages:
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741