If the example models seem to be working well, with the language providing a good
way to model individual applications or features in the domain, we can consider the
next step. If the domain is relatively new, the number of people that will use the
language is relatively small, and your DSM experience relatively thin, our suggestion
would be to move on to the generators and domain framework. In more challenging
cases, and with a little more experience, there is another area it would pay to look at
?¬?rst: DSM use scenarios.
ADSMuse scenario looks further than the issue of howthe modeling language can
be used to build a single application or feature. First, it includes the broader picture of
the usage of the whole DSM solution: the modeler, the tool, the models, their
representation in model ?¬?les, the generated code, and the ?¬?nished products. Second, it
expands the scope to look at multiple instances of each of these: how modelers,
models, and generated ?¬?les for different features relate to each other.
Looking at the broader picture of the whole DSM solution lets us spot possible
sticking points for the introduction of DSM to the organization. Comparing the
various phases, entities, and artifacts to current code-based development also reveals
what organizational change will be necessary to adopt DSM.
Pages:
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646