Quick questions on Tolvanen and Kelly reading
Usual disclaimer: Being able to answer these questions does not prove that
you remember the reading perfectly! But at least, not being able to answer
them does suggest you should reread it.
There's no automated checking of these: just see if you could answer each
question in a sentence or two. If not, see the page indicated (page numbers
are given in terms of the PDF document, so "p1" is the one with 711 at its
foot, that being the page number in the published volume).
- Broadly speaking, what view do the authors present of the success or
otherwise of (a) UML use (b) use of the OMG MDA approach? (p1-2)
- Roughly what productivity gain do they claim users of their tool get?
(p2)
- What is their tool called? (p2)
- The authors describe another paper which reported on the amounts of
effort to develop a DSL using each of several different tools. The same
paper, the authors say, presented theoretical predictions of which tools
should require more effort than others. How did the empirical findings
compare with the theoretical predictions, and what do the authors say about
this? (p3-4)
- Describe two aspects of scalability that may be important for DSL tools
to support well. (p5-6)
- Why can maintenance be a problem for DSLs? Outline the claims the
authors make for their tool. (p7-8)
- Open-ended challenge question Clearly, these authors are not neutral!
Suggest one important reason why companies continue to use Eclipse-based tools,
despite the claimed superiority of these authors' tool. Reread the paper
with a sceptical mind, looking for places where perhaps they may be
partisan. Then look on the web for other views and reviews of the various
tools available. For example, here's
one short post, and one of the authors of this paper turns up in the
comment section...
This page is maintained by
Perdita Stevens (
perdita@inf.ed.ac.uk
)