We have been running a large user study of the Mylar tool in order to
determine how effective it is at reducing information overload in Eclipse.
Mylar monitors programming activity, actively searches for related
elements, and extends the Eclipse UI to focus on the elements of interest (http://eclipse.org/mylar). A key part
of the study is the initial baseline phase which provides us with statistics
about the views and commands that you use to navigate. It would be great
to have both this and some Mylar usage statistics from people using AJDT, so
that we can feed this data back into AJDT development. If you are
interested in participating in the study please read the invitation below and
make sure to sign
up for the mylar-announce@xxxxxxxxxxx mailing list.
----------------------
If you have any trouble reading this email
please view: http://eclipse.org/mylar/ubc-study.html

|
UBC Department of Computer Science
201-2366 Main Mall, Vancouver, B.C., V6T 1Z4
http://www.cs.ubc.ca
604-822-3061
|
Mylar Preview
Release User Study Invitation *
We
are happy to announce the v0.3 preview release of Mylar (http://eclipse.org/mylar).
Mylar monitors programming activity and extends the Eclipse UI to focus on the
elements of interest. This preview release is available to anyone willing
to participate in a user study. You have been invited to participate
because you previously expressed interest in Mylar. The study will help
us validate, harden and improve the usability of the tool before we make it
generally available from eclipse.org. Participating in this study will
require minimal time from you, as it simply involves using Mylar for your daily
work. You are free to stop using the tool at any time.
To
use Mylar you must running on a Java 5 VM and using Eclipse
3.1, do not install otherwise. As you work Mylar will monitor your
interaction with Eclipse by logging the IDs of views that you use, editors that
you work with, commands that you invoke, and preferences that you change. Mylar
also generates statistics about the program elements that you edit and
navigate, but removes any private information identifying those elements by
obfuscating their names and paths. Approximately once every four days you
will be asked to confirm an upload of this interaction log to a UBC server,
will have the opportunity to review all of the information being sent, and to
answer a short questionnaire. You will also have the option to remain
anonymous throughout the study.
The
study is composed of three phases. After a sufficient amount of
interaction data has been collected for each phase, around 10 days of 'typical'
programming activity, you will be prompted to upgrade to the next phase.
· Baseline: This
will only track plain Eclipse usage to create a baseline (no Mylar features)
· Study: Mylar
features providing degree of interest facilities to standard Eclipse views, and
predicted interest facilities and active versions of some
standard Eclipse views
To
start Phase 0, use Eclipse's update manager to install the Mylar Monitor
feature (Help menu -> Software Updates -> Find and Install -> Search
for new features to install, then add the URL below Mylar as a New Remote Site
called "Mylar phase0", click Finish, and proceed through the
installation wizard):
www.cs.ubc.ca/~mylar/alpha/baseline
You
are free to start the study anytime over the next two weeks, and if you would
like to join later just let us know. If you experience any problems
installing or using Mylar please submit
a report, as we will be releasing frequent updates. To receive
further announcements about the user study and future releases please sign up
for the mylar-announce list. If you participate in the study we
suggest that you sign up for the mylar-users list. If
you have plans to develop or extend Mylar we ask that you wait until you have
completed Phase 2, and sign up for the developer list after that point.
If you have any other questions simply reply this email.
Thanks
and stay in touch,
The
Mylar Team
* UBC Study title: Using Interaction History to
Determine Interesting Information in a Development Environment. There are
no known risks associated with this study. After the study, this
interaction history will be stripped of all information identifying
non-anonymous subjects, and we will use this data to determine what Mylar views
are useful, and to determine whether Mylar makes it easier to work on large
systems in Eclipse. We also plan on making this anonymous usage data
available to other researchers interested in improving the IDE user
experience. Please note that the data collected will be archived for
future use. This data will be anonymous and not contain any
identification information. This study is being conducted by Mik Kersten
(beatmik-at-cs.ubc.ca) as a part of his PhD Thesis and supervised by Dr. Gail
Murphy (murphy-at-cs.ubc.ca). If you have any questions about the data
being collected or about the study in general please email us or call Mik
Kersten at 604-822-0193.