. Do we have rules that are too strict? Sometimes there is a tendency, perhaps
because rules were easy to identify, to de?¬?ne language so strictly as to
actually prevent modeling actions. You may, for example, correctly de?¬?ne
the rules but they make sense only when the models are ready and remove
freedom during early sketching or when making model modi?¬?cations. A
typical case here is when model elements can??™t be changed but need to be
deleted before a new correct model element can be de?¬?ned. As the
temporary deletion may also delete related properties, relationships, and
even submodels, modelers are forced to model part of the speci?¬?cation
twice. Here, checking the consistency rules only on request instead of at
modeling time usually solves the problem.
(3) Automation with generators normally deals more with the capability of the
generator than with a modeling language. However, sometimes certain
generation actions simply can??™t be implemented since the initial input is not
available or is not easily accessible from the models. Here, you should analyze
examples that combine the original data in models and their implementation
code as produced by the generators.We will discuss generators in detail in the
next chapter.
10.8 MAINTAINING THE LANGUAGES
Languages do not remain static but evolve over time when the domain or generation
needs change or when modelers, after having learned to use the language, see new
opportunities for modeling.
Pages:
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485