[kepler-dev] Kepler actor icons

Laura L. Downey ldowney at lternet.edu
Thu Jul 14 08:06:51 PDT 2005

See comments interspersed below.


-----Original Message-----
From: Bertram Ludaescher [mailto:ludaesch at ucdavis.edu] 

We talk about different user groups for sure (as is now acknowledge
generally ;-)

But we also talk about very different kinds of uses cases! (not just
different user groups) 

[LLD>] Agreed

Some "workflow" use cases might very well have some components that
correspond to numerically solving differential equations. Other
workflow use cases might be just darn complicated by their nature,
i.e., inherently. No usability engineering (or software engineering
for that matter) can simplify the inherent complexity of those complex

[LLD>] Again, agreed.  Usability engineering can't magically make complex
problems and issues easy -- but it can help us design the system to better
support the tasks -- whatever they are via a combination of good visual
design, information design and interaction design.

But this seems obvious. Simple linear pipelines can capture certain
linear workflows. But when your problem requires complex data and
control flow, nesting, complex iteration and/or recursion etc. then
one has to bite the bullet, no matter what. Obviously icons don't do
much to help this problem. Higher order functions (map, foldl,
foldr,...) or other workflow patterns as recently adopted by Triana
(from the van der Aalst type of pattern) do. 

We should focus more on these things too. Icons (and the discussion
about them) should focus on those used in toolbars, menus etc. 

For actors they are overrated (IMHO)

[LLD>] Just to clarify here Bertram, we are not going to focus on the visual
design any longer.  I will incorporate the latest feedback and make a final
recommendation and we are moving on to other issues.  I have a consolidated
list of usability issues I've created (from the group testing, observation
and team feedback) and will continue to add to that list, including the
important issues you have raised.


More information about the Kepler-dev mailing list