From Hand Wringing to Writing
Reading over Google Reader today, a share came up from Dylan McCall. It troubled me deeply as I found the conclusions to be somewhat reckless.
After a bit of teeth grinding, I pondered the posting by Mystilleef a little more. While my initial reaction was quite adversarial, I gave it a little more time to stew. Being Libre software, we need more darn communication and less confrontational approaches, so I thought I'd stick a quill in my closed hand and try to get an alternate side to the discourse started.
Emotional Engineering Vs Software Engineering
The argument goes that software should adapt and bend to users preferences and not the other way round. As such, "good" and "powerful" software should be configurable and littered with plentiful options to accommodate all known and unforeseen needs of users.Here the premise starts. While there is a clear polemical model being established, this was not my issue. Had the post solely focused on the belief that multiple preferences and options were poor design, I likely would have agreed and moved along. However, it did not.
The Bone of Contention
Human preferences are affected by too many variables most of which are subjective, cultural and emotional. In contrast, the most efficient and effective way to accomplish a task is a purely engineering feat. It is measurable, quantifiable and observable. Somehow, many people have missed this important point.And here begins the problem.
As anyone that is familiar with this blog would attest, the subject of audience and context crops up time and time again. Who is the audience? What are their needs? What is their background?
The above quote nicely rounds out a good number of pitfalls associated with Libre design. Even though the quote looks quite simple, there are hidden assumptions and, as Foucault would suggest, buys into a very particular épisteme. It is one that I believe permeates Libre culture.
Human Preferences are Complex
True. One hundred percent agreement.
Sadly however, this is no excuse to cash out and leave design at the door. Do we give up because something is apparently complex to a given person? Is that complexity necessarily insurmountable by another person? Are we to assume that an averaging is an optimal solution?
Try to remember the last time you saw a wonderful magic trick.
Many magic tricks defy explanation. And yet, to the performer, that magic trick is nothing more than the result of solid design work and thought. To you, the audience, it results in a wonderful experience when executed properly. To the magician, it is often more about complex work and execution.
I will give Mystilleef full credit for being one of the few to see that there is in fact a cultural component to this. Aesthetics and visual perception are very deeply rooted in cultural starting points. In fact, the Eurowestern belief that there is a physiological consistency across all humans is in fact challenged by some seminal anthropological tomes on the subject.
So, in short, my first point would be to not avoid the subject of solid design merely because "human preferences are affected by too many variables." Rather, I'd encourage everyone involved in Libre software to begin to examine those variables and pay strict attention to them.
Avoiding the complexities has led us to the precise blight of design we see today in Libre software. Bad guessing. Poor judgement. Awful pursuit of mythical universality.
Most Efficient and Effective
In the second sentence we can immediately see two trap words. "Efficiency" and "effective" must always be evaluated within contextual information, even more so when we add "most" into the mix. If we were going to suggest how a graphical user interface is the "most efficient and effective" interface to control a computer, we have effectively ignored the context and circumstance as design considerations.
Absolutism has no place in design.
So on my second point, the presence of attractive trap words leads us away from the hard nitty-gritty of design. Context. Questions. Audience.
Thus far in our human history, the desire to see mathematical rigor in the fields of aesthetics and design has yet to come to pass. Even if we were able to fully quantify and qualify this complex interaction, we would likely find that the results would be extremely varied based on the complexities of the situations.
This isn't comforting for many. This is likely especially not comforting for a large portion of the brilliant minds that are attracted to Libre software.
At risk of horribly oversimplifying a culture, Libre software culture seems to have taken its current cultural trajectory from the historical minds that founded it; mathematicians, engineers, software programmers.
Those fields are moored in a heavy context of right and wrong. Yin and yang. Compiles or not compiles.
Design, aesthetics, and other such things are nebulous and subjective. They are specific to people, their environments, and their historical contexts. In a world that worships right and wrong, design is an awkward and sinister beast.
Design, aesthetics, emotion, and experience are all valid and powerful traits of a piece. There is no simple measurement or secret formula that deliver them. It isn't always simple to observe nor quantify the machinations and ramifications of such facets. They are not subject to the traditional engineering rules established by the harder sciences. They are the subject of emotional engineering.
On my third point, I'd likely agree with Mystilleef but perhaps in a slightly different way; it is all purely engineering. But let's not be too hasty to dismiss emotion and experience as vital parts of a Libre software project's architecture to be engineered.
Wandering off into Postmodernism and Foucault
However, these reasons should hive around usability and accessibility and again not individual preferences.The attitudes and priorities of a given culture are reflected in their words. The constructs. The accepted and unwritten knowledge. Who wouldn't want something to be "usable"? Another horribly appealing trap word.
Usability, when cast adrift alone, is a blind pig of a term. It presupposes context. It makes grotesque assumptions. And it dissolves in the face of diversity and snaps under the weight of scrutiny.
It is as vacuous as the term "beautiful" or "delicious".
If ever there were a purely mathematical approach to this magical word, it would be leveraged the world over and everyone would agree. But, in practice, this isn't the case.
If we can avoid pretending the word holds value and quit chasing the dragons it paints, we might stand to make some truly compelling software in Libre.
Let's cast out "usability" and let "people" stand where it rested.
Who are they? What are their desires? What are they attempting to accomplish? What are their cultural needs? What excites them? What engages them emotionally? Where and when are they living and what is relevant to them?
For some, there is little passion or fascination for design and the emotion it brings. I'd like to believe that the majority of the audience that reads this blog is of a different temperament.
I'd sum up the answer as to why we should focus on individuals with a number: 18.
Malcom Gladwell makes a case for the number 18 in his presentation Malcom Gladwell on Spaghetti Sauce. I would heavily encourage all readers of this blog to watch it. It is entirely topical to this discussion of "usability" and other universals. I'll leave you with an intriguing quote so that you might go and watch it yourself. The number 18 comes from the difference of a coffee rating among averages and groups. It is tied in with the rejection of universals and the embracing of diversity.
"The difference between coffee at 60 and coffee at 78 is the difference between coffee that makes you wince and coffee that makes you deliriously happy." -- Malcom Gladwell on Spaghetti SauceThe emotional and experiential realm is upon us in technological design. It cannot and should not be avoided simply because the art and craft of tickling those nerves is radically different from the art and craft of the lower level software code.
It is my firmest belief that until we accept and embrace the need for emotional design, we will never get those magical 18 points of difference that Mr. Gladwell references. We will forever repeat the horrible and unfortunate mistake of aiming at the average between two targets and hitting neither.
We must embrace and focus on crafting subjective emotions and reject the notion of idealized universals.
To suggest for a moment that emotions cannot be crafted or engineered of course, is to suggest that a tear or a smile has never been shed over prose, during a movie, or while listening to a song.
It is still engineering.
Just of a different sort.
Once again, thank you immensely for taking the time to read. Thank you to Mystilleef for stirring the brain...
 The Influence of Culture on Visual Perception and The Hidden Dimension might be worthy reads if you are interested in the subject.
 If a person were administering a computer half a world away via a slow bitrate connection such as dialup, it is possible to suggest that a command line interface may be the most efficient and effective. In short, there is never an absolutist definition of "efficient" or "effective" without examining the contextual elements.
 Marc Hassenzahl has done a wonderful job of assigning a numerical model to some of these more nebulous concepts, but even then, it is a model. It may work in some instances, it may break in others.
 If you have issue with this claim or find it interesting, it is perhaps worth reading Michel Foucault's The Order of Things.