Mailing-List: contact dev-help@apr.apache.org; run by ezmlm
Precedence: bulk
Message-Id: <m17wlAy-001KSaC@engr.smu.edu>
Date: Wed, 2 Oct 2002 10:12:36 -0500 (CDT)
From: gkoru@engr.smu.edu (Gunes Koru)
To: dev@apr.apache.org
Subject: A survey about handling bugs in open source projects

Hello all APR contributors,

I am conducting a survey about the way defects (or bugs-I use these two words
 interchangeably) are handled in open source software projects. It is
very easy to fill out. It consists of three short sections which can be
completed at once or in different sessions. The survey can be found in the
address:

http://www.seas.smu.edu/~gkoru/surveys/dhsurvey.html

This survey includes questions that can be answered by developers,testers,
bug fixers, bug database owners, and project managers. I would greatly
appreciate if you could visit the above web page and fill out the survey.
I am sure you will find the questions very interesting and thought
provoking.  We need the help of all contributors of APR in the above roles
to understand how we can use bugs data collected in your project for
software engineering research.

Nowadays, there is a huge amount of bug data on the Internet collected
during the development of all open source products. A bug database include
useful information to identify high-risk, problem-prone modules
(components) in the software. It is also possible to measure these
problem-prone components using several complexity metrics (McCabe's
cyclomatic complexity, Halstead's metrics, etc.), since the source code is
available. If a characterization, which is generalizable across many
projects (sub-projects) could be made in terms of complexity, focused
quality improvement would become possible in the future projects. So far,
in the literature, there is quite amount of evidence that 80 percent of
the problems occur from 20 percent of the modules (or software
components), which gives hope toward tremendous quality increase, time
savings, and increased success in the open source projects.

Also a bug database includes useful information about the efficiency of
development. For example, if a lot of bugs related to design of the system
are found after release, when coding bugs are expected more, then this may
indicate that the design or architecture of the system was done cursory.
Of course, these kinds of conclusions would require some kind of
classification of bugs according to their insertion time after they get
fixed.

However, before trying to find an answer to the questions like above, our
first step is to understand if the available data is usable for empirical
research purposes. This means having an idea about the consistency of bug
reports and fixes, accuracy of data, completeness of the data, etc. At
this point, we need your help. By filling out this survey, you will help
us understand availability, representativeness, and characteristics of the
data in the bug database of your project.  I'd like to stress that the
purpose of this survey is only and only research. Open source development
gained a very good momentum in the last decade. We hope that outcomes of
our research will help even increase this momentum. We will acknowledge
the help of all supporting communities when mentioning our results.

In addition to these goals, we also think that filling out this short
survey can lead to some immediate brain-storming in individuals and
groups, and exchange of ideas among the APR community, which will
be useful too. As I said, you will find the survey questions interesting.
If you'd prefer to learn more about our research, I included a link in the
above web page of our survey. This link provides more information about
our research.

Once more, your contribution is very important to us. Please visit

http://www.seas.smu.edu/~gkoru/surveys/dhsurvey.html

and fill out our survey.


We greatly appreciate your helps. Please contact me for any question you
might have.

Thanks

-- 
***************************************************************************

A. Gunes Koru

Research Assistant, Ph.D. Student

Southern Methodist University
Computer Science and Engineering Department
6425 North Ownby Drive
Science and Information Building Room 317
Dallas, TX 75205


Home: 214 691 5633
Work: 214 768 2005
Cell: 214 893 7311
http://www.seas.smu.edu/~gkoru
Email: gkoru@engr.smu.edu

***************************************************************************