SEARCH
0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Prev | Current Page 170 | Next

Steven Kelly and Juha-Pekka Tolvanen

"Domain-Specific Modeling"

This is needed because it usually does not make sense, or
is not even possible, to put all the rules in the metamodel and check them during each
modeling action. This is especially true when checking partial models, when there are
multiple models, or when integrating models made by different developers.
Generators for model analysis can also be used for guiding modeling work and
informing about actions needed. A typical such scenario is to look if a model, or
models, is incomplete and report possible actions needed to make the model complete.
Such model checking can be run similarly to generators: when needed or after
conducting certain modeling actions.
Metrics When moving toward model-based development, code-based metrics can
still be applied; they are now calculated from the generated code instead of from the
manually written code. As platform code is already available, the metrics may also
cover platform functions or libraries used instead of focusing only on the generated
application code. Use of code metrics based on models is easy as it does not require
much change to earlier practices, but it is not likely to be the most effective use of
metrics.
In DSM, code-based metrics no longer measure the amount of human work
needed. Metrics, like function point analysis (FPA, Albrecht and Gaffney, 1983), to
analyze program size and to estimate required development effort are no longer
relevant: the application is often already ready when we can calculate these metrics.


Pages:
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182
hotel jelenia góra Russian bride Free English grammar and study guid powiekszenia wielkoformatowe counter strike 1.6