As you rightly point out it is not easy to identify a link between
software faults and identifier naming problems. Some have tried. See,
Arnaoudova et al. Physical and Conceptual Identifier Dispersion:
Measures and Relation to Fault Proneness. In Proc. of the Int'l Conf. on
Software Maintenance - ERA Track, pages 1-5, 2010. IEEE Computer Society.
or our own work http://oro.open.ac.uk/view/person/sjb792.html
Other research you might want to look at includes the link between poor
identifier naming and program comprehension problems, which Brad Myers
identifies elsewhere in this thread. As well as issues of overly long
names and esoteric typography, concerns have been expressed about the
semantic content of identifier names. Deissenboeck and Pizka Concise and
Consistent Naming Software Quality Journal, 2006, 14, 261-282 is
probably the best known source.
Einar Host's work on the connection between Java identifier method names
and method functionality may also be of interest to you. The connection
is strong enough for the functionality of methods to be used as an input
to name refactoring. (Høst, E. W. & Østvold, B. M. Debugging Method
Names Proc. of the 23rd European Conf. on Object-Oriented Programming,
Springer-Verlag, 2009, 294-317).
On 03/26/2012 11:08 PM, Raoul Duke wrote:
Check out my paper from PPIG 2006:
Thanks, I just skimmed it. I probably haven't explained well at all
what I had in mind. I searched for "bug" and "error" with no hits. It
so far appears to me to be a review of how naming is used; doesn't get
into the resulting impacts of those uses and choices.
Now that I think about it, I guess the sort of thing I have in mind
is: has anybody somehow managed to study a bug database and the fixes
and figure out %ages of what caused the bugs, and if it is possible to
figure out that naming was a problem, how much was due to problems
with naming? I guess it would be quite hard to tease it out.
Presumably naming hasn't caused Arianne to crash or anything like that
(http://www.sans.org/top25-software-errors/) so I'm really probably
only ranting about my own little pet peeve. Apologies.
"No set of mutually inconsistent observations can exist for which
some human intellect cannot conceive a coherent explanation,
The Open University is incorporated by Royal Charter (RC 000391), an exempt charity
in England & Wales and a charity registered in Scotland (SC 038302).