Collaboration in Open Source Domains - Collaborative Technologies and Applications for Interactive Information Design - page 174

Table 3.

Opened by Vinicius Depizzol (reporter)
2006-12-07 14:57
gnome-cups-manager should follow the GNOME Human Interface Guidelines.
Comment #1 from Christian Persch (developer)
2006-12-10 18:47
Thanks for the bug report.
Could you be more specific about where exactly gnome-cups-manager lacks HIG compliance?
Comment #2 from Vinicius Depizzol (reporter)
2006-12-11 14:50
The properties window of the printer (Edit > Properties) don't follow the Layout Specification [1] of GNOME HIG, as well information
alerts (Printer >Print test page) [2].
Thank you.
[1] http://developer.gnome.org/projects/gup/hig/2.0/design-window.html#layout-callouts-figure
[2] http://developer.gnome.org/projects/gup/hig/2.0/windows-alert.html

engineering (UE) concepts and processes with that
of software engineering (SE). It was a matter of
seamless integration of UE within SE so that de-
velopers would be able to progress without waiting
for, or get interrupted from usability engineers.
Given that a graphical user interface comprises
half of the development effort and code in modern
applications (Myers and Rosson, 1992), there have
been mathematical computations regarding costs
of usability development in order to justify the
development efforts. While both SE and UE both
provide a wide range of process models, these have
been seldom exploited mutually in order to build
a systematic model with a focus and approach on
user centered design (UCD). According to Hakiel
(1997), “Although the emergence of human factors
engineering predates that of software engineering,
an integrated approach to the design of ease of use
in human-computer systems is yet not routine in
the software industry.” There are many reasons
behind this, namely i) current UE models differ
much from SE models, ii) usability models are
complex, iii) company managers do not believe
that usability is cost justified, and iv) project
managers, designers and developers see HCI as an
academic subject (Granollers, Lorés, and Perdrix,
2003). The challenges have popped up within the
integration models of SE and UE, and research-
ers have identified different SE process models
with different abilities to create usable products,

assessing their UE absorption level.
This issue, by far, is not limited to uninsti-
tutionalized UE methods. Distributed projects
with a loose development process have far more
issues than those developed under a controlled
mechanism. The issues arise from different facts,
mainly bug reporting systems and hierarchical
decision paths. For example, deciding whether
fixing a usability bug affects the issue itself, or
should the developer work on other parts of the
user interface to keep the coherence and satisfy
compatibility (i.e within-application consistency)
has been an issue. If the application is relatively
small and has a few number of screens, the bug
does not have a rolling effect. On the other hand,
if the bug affects the whole desktop (i.e. changing
a logo, modifying a help page, having a major
update to the basic UI library), its implications
should be analyzed deeper, potentially including
other developers. Moreover, identifying whether
a bug is really a usability bug has been an issue.
Nichols and Twidale (2007) note that some us-
ability bugs have been regarded as unneeded. This
may stem from the fact that the compliance of a
bug with the human interface guidelines (HIG)
of the given product or project cannot be easily
determined due to lack of adequate guidelines.
As an example, Table 3 shows discussions which
have been taken from GNOME bug reporting
tool, Bugzilla.