Jump to content

A Methodology for UI Design Patterns Discovery

EderAndres's Photo Asked by EderAndres Peacock
Posted Apr 22 2010 05:37 AM
3856 Views
A Methodology for UI Design Patterns Discovery

Hi from Colombia,

I'm a university undergraduate student on computer sciences and for getting my bachelor's degree I'm proposing a project about UI Design Patterns for desktop applications plus bloopers. Designing Interfaces: Patterns for Effective Interaction Design by Jenifer Tidwell (O'Reilly) is my main starting line. Also, Pattern Oriented Software Architecture: On Patterns and Pattern Languages, Volume 5 by Frank Buschmann, Kevlin Henney and Douglas C. Schmidt (Wiley) helps me on design patterns specification.

But, I'm getting stuck on one formalism required for proposal approval: I need to specify what research methodology I'll use for the development of my catalog. When seeing the types of research on Research Methodology: a Step-by-Step Guide for Beginners by Ranjit Kumar (Sage) page 9 I think a good candidate would be a Descriptive Research, but I'm unsure.

I'm wondering what process The Gang of Four and Jenifer Tidwell followed for the discovery of OO an UI design patterns respectively. Can you help me?
Tags:
3 Subscribe


9 Replies

Comment by EderAndres Peacock  : Apr 30 2010 07:15 AM
On Twitter @KevlinHenney said me this:

"Descriptive research looks like a good start, but you may find that patterns also encroach on explanatory research. HTH!"
0
  zigurd's Photo Answered by zigurd O'Reilly Author Answers Early Bird Peacock
Posted May 02 2010 10:45 AM
The UI book you are starting from is a good starting point. It has good coverage of direct manipulation, for example, and why direct manipulation is desirable. But, like the GoF book, it is a compendium of already-implemented design patterns. It's about selection, not discovery.

You may also find that some UI design patterns, like good practices in forms design, might have a very long heritage behind them, going all the way back to manual operations, while others, like 3D in a touch UI, might be mostly a green field, or you have to go to high-end systems like 3D CAD to find a body of work to select existing design patterns and inspire new ones.
Comment by EderAndres Peacock  : May 04 2010 07:55 AM
On Twitter, @KevlinHenney said me this why Design Patterns Discovery is not an exploratory research:

"Patterns are based on mining, framing and combining existing knowledge, which sounds less exploratory"
Comment by EderAndres Peacock  : May 13 2010 08:08 AM
This is something confusing: in Pattern Oriented Software Architecture: On Patterns and Pattern Languages, Volume 5 chapter 03 page 101 we find that the elements of a pattern form are:

Quote

Context: Situation giving rise to a problem
Problem: Set of forces repeatedly arising in the context
Solution: Configuration to balance the forces
Identification: Name and classification for identifying the pattern
Consequences: Consequences arising from application of the pattern


which shape the classical description of a pattern. But, in Designing Interfaces: Patterns for Effective Interaction Design and Designing Social Interfaces, both about Interaction Design Patterns, we find these elements:

what:
use when:
why:
how:

I'm trying to map the two patterns forms as follows:

Context <--------> use when

Problem <--------> what

Solution <--------> how

Identification <--------> name

Consequences <--------> why

Two questions arise:

1. Have I mapped correctly the two forms?
2. What pattern form is best? I mean, which pattern do I have to choose for interaction design?
Comment by EderAndres Peacock  : May 13 2010 09:25 AM
On Twitter, @KevlinHenney said that this is the proper relationship between the two pattern forms
(Tidwell vs POSA5) is:

what -> problem
when -> context & problem
why -> problem & solution & consequences
how -> solution

So, I think that extracting the elements of a patterns from Tidwell and re-compose them into the POSA5 or classical requires a good effort.

But now, What form do I have to adopt the classical from OO or from the emerging Interaction Design? Agree with the OO professionals or withe ID ones?
Comment by EderAndres Peacock  : Aug 03 2010 02:49 PM
Ok, the first thing I have to say is that my project wasreduced to work only with .NET Windows Forms for the following three reasons:

  • Since I have to propose a case study to show howthe interaction patterns can be implemented, that implies a duplicated workbecause I should develop the .NET version and the Java version of the sameapplication. Even I can use web services for the interoperability, the user willfell that the case study is huge to be developed in short time, say 3 months orless.
  • That duplicated work consumes the double of timeboth when researching, writing, editing and developing . (And I need to get mydegree in March2011 in order to get a job)
  • That duplicated time consumes personal energy,and of course more money for the light meals.

Comment by EderAndres Peacock  : Aug 03 2010 03:10 PM
Now, for this project it is necessary to write up two documents:

The first is the project degree mini-book that every student has to left at the university containing the description, results, andconclusions of the research.

The second is the guide for developers explaining the casestudy and the interaction patterns. I conceived my project as a fushion betweenthese books:

Designing Interfaces by Jennifer Tidwell, and
GUI Bloopers 2.0 by Jeff Johson


Posted Image


Plus how to implement these interaction patterns and guides in .NET 4.0 Windows Forms through the case study.
Comment by EderAndres Peacock  : Aug 03 2010 03:21 PM
Writing the methodology research process...
Comment by EderAndres Peacock  : Aug 04 2010 05:53 AM
Question: Tthe term "standard user" is valid for referring a person without physical or mental disabilities?

 
Sign up today to receive special discounts,
product alerts, and news from O'Reilly.
Privacy Policy >
View Sample Newsletter >