When user interfaces fail

Some people like to do “designer bashing” from time to time. I was just in the mood to do some “developer bashing” today.

There are a number of reasons why user interfaces of many software packages fail. I assume (slightly unfair and inaccurate), that in many cases there is no interface designer involved with user interface development, but rather the interface is designed by the application developers.

Here is a list of (some) common fallacies of some developers in regard to user interface design:

Fallacy #1: User centered design approach is optional.

Some developers actually have no idea what “user centered” actually means (as many consider the implementation of software that users can interact with already as being “user centered”). Also, the most important aspect of user centrism to developers is the feature list, because the features describe what the user could do. The actual ability is a function of a) the features and b) the capability of the user. What is not agreed on many times is that “capability” is not solely an attribute of the user, but is constituted by user experience level plus interface design. The 80/20 rule is applied as 80% application development and 20% interface development instead of the other way around (and in the order “first interface then application”).

Fallacy #2: Features are more important than usability.

For some developers the loss of features during planning phase seems to be a too big tradeoff in comparison to a “minimal” usability improvement.

Fallacy #3: “Design” is just an emotional and subjective quality.

Some developers think a “design” will just become necessary if the application should not only work, but also please and delight. While “likeability” is an important aspect, it is heavily underestimated how much users dislike software that is hard to use.

Fallacy #4: Functionality is what the user could do.

Some developers consider “functionality” being an aspect of the software. In fact “functionality” is an attribute only present in the usage context – where the user is the most important variable. Functionality is not what the software provides, but what the user is able to use. Microsoft Word may have many thousand functions which most users are unable to use. So the functionality of MS Word for users is what they can actually (and not potentially) achieve.

Fallacy #5: Personal experience is the best advisor.

Some developers often think they are able to do “cognitive walkthroughs” on behalf of unexperienced users. While this is a possible approach, many developers (that usually are power users with deep knowledge about the application) do not go far enough when defining “unexperienced”. When doing cognitive walkthroughs many developers keep the same mindset about what is important inside the application. Really unexperienced may consider completely different things as being important.

Fallacy #6: Good application design is the primary determinator for good interface design.

Most developers are interested in designing the application (that’s why they call themselves “developers”). Interface design is an uncomfortable requirement to be added to the application design. While there is much truth in the idea, that well designed applications often offer cleaner approaches to interface design, it is a false conclusion, that a well designed application will necessarily lead to a good user interface.

Fallacy #7: It’s OK to reject major changes of the application for minimal interface design improvements.

Well, sort of. Most of the time it is basically a matter of a wrong design approach in the first place. There is also a persistent understanding of developers that interface design issues do not interfere with the deeper application design – which is simply ignoring the fact that in most cases it remains to be the case.

Fallacy #8: A bad user interface alone cannot set the seal on the fate of the application.

It seems to be irrational to developers to prefer going with no software and unresolved problems instead of trying to work with a hard-to-use application. Unresolved problems can often be deferred or ignored – or – other workarounds could be tried with an easier path to a less optimal solution.

3D desktop? Not again!

Ingo Hinterding pointed to the SphereXP project and Sun’s Project Looking Glass. Sphere looks quite easy to use – but doesn’t seem to be an incredible enhancement. Sure: you could have open 130 windows and see them all almost at once — if you rotate fast enough! All the 3D approaches I have seen have one major drawback (that Expose has not by the way): the viewpoint. It means that most of the space is outside the view. So there is the same problem as with virtual desktop tools: you need to remember where you have left what.

I do believe, that one day the 3D graphics hardware in almost every consumer PC today will be responsible for the next generation desktop interface. But I believe it won’t happen as long there is no departure from the 30 year old interface paradigm.

I have ideas about that. But I won’t tell anyone unless he/she pays me for it ;-).

Open Source software lacks good interface design

Michelle Levesque contemplates about the role of Interface Design in Open Source Software:

The lack of focus on user interface design causes users to prefer proprietary software’s more intuitive interface. Open Source software tends to lack the complete and accessible documentation that retains users. Developers focus on features in their software, rather than ensuring that they have a solid core. (…) If Open Source software wishes to become widely used and embraced by the general public, all issues will have to be overcome.

I can only second this position. I have evaluated many open source projects and I am amazed that the User Interface Design is seen as a followup problem — if at all.

The animosity of software developers about interface design issues is steady. The problem here is not that they devalue the issue in general, but that there are common misconceptions about the importance. There is usually also minimal or no knowledge about how to ensure usable software at all in many open source projects.

Windows UI critics

Paul Thurrott runs a site that discusses Windows UI (esp. XP and Longhorn):

“Now, because I present this information, I’m somehow labeled a Microsoft lover and/or an Apple basher. That’s silly. But Apple has done very little to make its UI better per se, beyond simple enhancements to what is, again, a classic desktop OS. There’s nothing wrong with that. Frankly, Apple’s crowd is technical enough to deal with it. But saying that such an OS is “easier to use” or “more elegant” than Windows is wrong. Mac OS X is attractive, and arguably “better looking” than Windows XP, though that’s a subjective declaration. But it is most certainly not “easier to use”. And that’s not “Apple bashing,” it’s just the way it is.”

I agree that Apple just brought a standard user interface to UNIX. But it is the first actually quite usable consumer grade user interface in the UNIX world ever. And this is something that some people believed would be impossible to do.

My personal experience about the activity centered approach of Windows XP is that it does disconnect the user from the computer by layering a webpage-like interface over the enourmous amount of control panels – adding more complexity instead of reducing it. It fails to create mental models that work and it prohibits the user to work with the computer creatively.

There is some merit in having more powerful control panels (which show you all mobile devices in one place for instance), but if that does not lead to less dialogs there is no real gain – it just looks more user friendly. There is no doubt that Longhorn will be more user friendly in that sense, but to claim MacOS X is “just attractive” is too simple.

OS X has to be compared to UNIX GUIs some years ago. The Aqua interface made UNIX accessible to consumers. The core OS is open source, many UNIX geeks love OS X because it allows them to work with their old stuff in new ways. That is something Microsoft fails to do.

A much better resource than Paul Thurrott’s site can be found on the X vs. XP site

Decoding visual language elements in news content

A great visual analisys and reflection on how to use images in order to shape people´s minds:

News delivery in this country (USA) is increasingly comprised of carefully crafted displays of visual information. As consumers of information, however, most of us have never been taught to critically read or decode images and other graphic displays of information in the same ways that we have been taught to analyze verbal communication. We are taught reading comprehension and writing skills throughout most of our educational experience, but not visual language comprehension.

[Design & Revolution News]

First Principles

Bruce Tognazzini: »The following principles are fundamental to the design and implementation of effective interfaces, whether for traditional GUI environments or the web. Of late, many web applications have reflected a lack of understanding of many of these principles of design, to their great detriment. Because an application or service appears on the web, the principles do not change. If anything, applying these principles become even more important.« [WebDEV]


There seems to be what Worlds.com or ActiveWorlds (formerly »Alphaworld«) always meant to be.

I was once asked to do a concept for a similar project called Nikoworld. The company hired 80 people that were working on 3D enviroments and high-end machines – no concept. To deliver a concept I asked to set up my own team, my own shedule and a fixed budget I can manage on my own. That seemed to be too much – I never heard of them anymore. Now this startup is history like so many others that tried to make money with technology but without ideas.