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 581 | Next

Steven Kelly and Juha-Pekka Tolvanen

"Domain-Specific Modeling"

We could no longer use text input ?¬?elds to show the time, nor buttons that
the user could click with a mouse. The main application class would no longer
be an Applet, but a Midlet, which worked rather differently. Compilation would no
longer be as simple as ???javac *.java,??? but would require a number of new phases
and con?¬?guration ?¬?les.
Perhaps most challenging, we wanted the same models to be able to generate an
application for either platform. Or rather even more: we did not just want to be able to
get to a state where one set of models could generate both, we wanted the existing
models, built before any idea of MIDP was considered, to work on both platforms. In a
real commercial case, this would have saved the modelers (of whom there would be
many) having to rework each of their models. While achieving this may require more
work in the DSM solution (the metamodel, generators, and framework), the bene?¬?t
would be multiplied over many modelers, each with many models.
Of course, if the claims of DSM are true, and if theWatch DSM solution had been
made according to good DSM principles, this requirement should be feasible. The
watch models would contain only information about what the applications should do,
and no implementation details of how they would do it or on what kind of platform.


Pages:
569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593
hotel jelenia góra Russian bride Free English grammar and study guid powiekszenia wielkoformatowe counter strike 1.6