terriko: (Default)
As many of you know, my sister and I have been amusing ourselves overmuch with our latest project, Now We Must Fight.

As anyone following my twitter feed or #kernel.org this afternoon knows, my labmates and I were having way too much fun with fonts.

I actually had an hour before bed, so I combined the two to produce a comic, starring Dan.

Now We Must Fight: Papyrus

(Best viewed at original size)
terriko: (Default)
In the course of doing some thesis research, I stumbled across this fascinating paper in aesthetics and usability.

I'm not sure I've ever read a paper where the researcher seems so thoroughly flummoxed by his/her results.

The idea of the study was to test whether objects rated as more beautiful would also be rated as more functional. The author, I suspect, found this idea faintly ridiculous, but previous work in Japan had shown that people did indeed rate prettier banking machine interfaces as more usable. He suspected that perhaps this was just an effect related to Japan and its "culture is known for its aesthetic tradition." He would repeat the study in Israel, where the culture has a stronger emphasis on action over form. Surely, he thought, the practical Israeli people would not be as affected by aesthetics.

But what happened? "Unexpectedly high correlations" The author says, "usability and aesthetics were not expected to correlate in Israel" but they did. Oh, they did.

Even though I'd not read this paper until this week, it's something I'd noticed in doing basic testing of my web projects (back when I made more of a living writing web code rather than deconstructing and mocking... errr... inspecting its security). I used to test designs on clients, friends and invariably, I'd get more positive results (and useful!) feedback if I spent the bit of extra time to make the first draft look clean, if still aesthetically simple. Pretty matters.

It's kinda nice to have a couple of scientific papers to back up one's gut feelings, eh?




Want more than a gut instinct to explain why attractive things work better? Don Norman suggests an answer in his book, Emotional Design: Why We Love (or Hate) Everyday Things. (I noticed it when seeing who had referred to this study and am working my through it. Research is fun!)

The theory goes like this: pretty things change your emotions in a positive way, make you happy, less stressed. Your emotional state changes your perceptions and ability to work. When you are happier, you often find things easier to use. Thus, pretty things are easier to use. And ugly things make you more easily annoyed, stressed out. Stress makes you perform poorly. Thus, ugly things are harder to use.

And in honour of the new Star Trek movie, I'll finish with a single word:

Fascinating.
terriko: (Default)
The story thus far:

Terri, stuck in one of those bits of PhD that seem never-ending, realized that she needed two new sections in her thesis: one on typography & design, to prove a point about web pages and one on security policy, to prove a point about how difficult getting it right can be. But then all of her hardware decided it needed replacing Right Now, thus making it nigh impossible to work, and after spending entirely too long debugging and replacing stuff, she decided to console herself by buying a zombie game to test her new network equipment. That's a perfectly valid response to stress, really.

We now return to her regularly scheduled thesis development...



In the course of working on these two pieces at more or less the same time, I've noticed that security policy shares a bit more with visual page design than I might have initially thought.

Security policy is designed to be both rigid and flexible. The idea is that if you do it right, it should be hardened, unbreakable, no loopholes. But the policy languages have to be sufficiently flexible to accommodate varied types of policy and capture desires from different organizations.

Graphic design is one of places where "the rules are made to be broken." Flexible first, but with a rigid structure to help guide you. And practical constraints regarding readability, screen sizes, printing sizes, etc. also affect design choices. It feels a bit like it's backwards from security policy: in graphic design, the flexibility is stressed first, and the rigid constraints are acknowledged after the fact.

There's a lot more math than one might expect in design. And in security policy. I took the grad security course at Ottawa U, and wanted to smack some of my colleagues as they complained incessantly every time the prof so much as mentioned math. I don't know how they thought they were going to comprehend basic cryptography without at least a few equations... but after reading parts of The Elements of Typographic Style last night, I wonder how many designers expected to learn about the golden mean and regular polygons? I'm a mathematician originally, so I delight in finding such things, but I know that's atypical in general (less so among geeks).

Good security policy is nigh invisible to the legitimate users. If it prevents you from doing your job, it's probably not good policy, right? Ditto for graphic design, in some ways. It seems weird to talk about a visual medium as "invisible" but in a lot of cases, you want the content to be doing the talking -- the design is a way to frame it nicely. It should be quietly doing its job, making the viewer feel better about the content, without the viewer noticing.

Of course, invisibility isn't always the desired thing for either medium: Sometimes you want attackers to see that big impenetrable wall. Sometimes you want someone to be drawn in by the artistry of a design. But a real whiz about either security policy or design is likely to need to be able to cover both ends of the spectrum (and a good chunk in between).

I'm not sure entirely where I'm going with this train of thought, but I thought it was kind of interesting that they're not as dissimilar as one might think.

Profile

terriko: (Default)
terriko

September 2017

S M T W T F S
     12
3456 78 9
1011 121314 1516
17181920212223
24252627282930

Syndicate

RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 24th, 2017 12:02 pm
Powered by Dreamwidth Studios