Comparing the productivity was somewhat straightforward as Nokia had already
been developing mobile phones and their applications and had internal metrics
available. A point of comparison for the study was developing modules from scratch
to a ?¬?nished product. It therefore also included phases after generating code, such as
testing and quality assurance. The comparison was based on developing applications
that were similar to applications and functionalities made into earlier products. As
mobile applications were increasing considerably in size and complexity, we may
well expect that the applications developed with DSM were no simpler than earlier
ones, and most likely had more functionality.
Military Command and Control Systems at USAF In the U.S. Air Force
(USAF) domain-speci?¬?c languages with code generators have been compared against
best-practice code components (Kieburtz et al., 1996). This comparison, covering
both initial development and maintenance tasks, detected productivity improvements
in the ratio 3:1. The study is reported in an academic conference and has a carefully
planned and followed research method. Other researchers can therefore repeat
the study in their application domain. Experiments to compare two alternative
development technologies have been conducted in relation to message speci?¬?cation in
the Air Force C3I systems.
Pages:
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63