The year 2006 saw the release of the largest number of new tools like this,
even exceeding the metaCASE boom of the 1990s.
These prototypes and ???version 1.0??? tools generally support a single user, one
modeling language at a type, simple metamodels focusing on objects with basic
properties and relationships, and symbols with a single graphical element and a label.
Generator facilities will be based on text-to-text transformations or handwriting code
to read models. The resulting modeling tool will normally be missing the majority of
functions that users expect from a graphical editor.
Compared to creating the initial prototype, turning it into something that could be
used in the real world is a much harder task. Building such a system without signi?¬?cant
personal experience of industrial scaleDSMappears to be a near impossibility. This is
sad, but not surprising: if building a metamodel requires the top expert developer in a
domain, building a meta-metamodel for all such top developers requires the very
highest levels of experience as well as intelligence and skill. The only other hope is
dumb luck, but then the current authors probably used most of that up already, leaving
little for newcomers!
We can thus assume the basics: a tool maturity level of at least 3, and the basic
metamodeling and modeling facilities mentioned above.
Pages:
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698