90 percent of everything : Usability Blog

Written by Harry Brignull

I’ve always been a bit of a fan of Web 2.0 darlings 37signals. I used to read their ‘design not found‘ articles fairly avidly and even bought their excellent book ‘Defensive Design for the Web‘. Their sense of aesthetics is faultless and they put a lot of thought into making simple details of usability perfect.

So I finally tried out Basecamp - the project management web app.

And I must say I feel a little under-impressed. All the parts seem a bit separate. You can’t link messages to to-do list items or milestones or anything else. The only place where everything intertwingles is on the overview page and this appears as nothing much more fancy than a chronological list of changes.

I know the whole philosophy behind their apps is a about simplicity and avoiding featuritus but this harks back to the debate about simplicity that went on a few months back. Simplicity of use doesn’t imply absence of features and some well chosen additions to Basecamp could make it massively more useful.

It also implements one of my pet hate - plain text boxes that allows formatting via simple markup. I hate this in Wiki’s, I hate Textile and Markdown and all the others. Mainly because there *are* several and I can never remember the formatting rules.

Although web-based rich text editing is far from perfect I feel like I’m flung back to the stone age everytime I have to use markup on a site.

.

Another great deal from Three.

After all the beta testing - how did this one slip through the net?

When you browse videos in Vista Media Center it only shows you thumbnails with no text label whatsoever. (you do get a filename displayed when you hover over a particular item).

All my videos fade in from black. So I’m stuck browsing 30 black rectangles!

Microsoft. Reinventing Mystery Meat for the modern living room.

I just paid my tax online (how fascinating) and was asked to take a quick survey. One of the questions was a ‘If so, then…’ type question. Of course I’d answered ‘no’ to the previous question so I skipped it. And of course I couldn’t submit the form without answering it.

I did another survey a few weeks back (I was procrastinating that day as well) and it assumed a few things about my job that weren’t true. As a result half the questions - which were again compulsory - were unanswerable. After a page or two of entering junk just to be allowed to go on to the next page I gave up.

The problem with the tax question was just silly but this is a much more subtle issue.

If you’re designing an online survey you’re obviously assuming respondants are fitting a certain profile and this will be reflected in the design of your questions.
If you get this wrong you’ll either be getting a high drop-out rate or people will fill in junk answers.

I’m apparently the CEO of a company called ‘N/A’. I’m sure there’s a few of us on the board…

If you’re in the UK you may have seen this in The Guardian today, but if not - check out this this article by Charlie Brooker. A great takedown of the Samsung E900… As he puts it: “The whole thing is the visual equivalent of a moronic clip-art jumble sale poster designed in the dark by a myopic divorcee experiencing a freak biorhythmic high.”

Love his choice of words!

> Read the article

In this BusinessWeek article, Jakob Nielsen calls the OLPC UI design approach “reckless” because they have done no user testing so far. Meanwhile, John Maeda bizzarely praises process as “…the Steve Jobs method. […] You don’t use focus groups. You just do it right.”

You just do it right? …You just do it right? … That’s a little overconfident don’t you think? You are talking about children’s lives and huge chunks of the education budgets of governments that don’t have money to spare. And no self-respecting UCD practitioner would use focus groups in this context anyway. Imagine trying to explain the UI to a bunch of kids in a room. The only way to test it is to put it in the target environment, and look at the way it’s adopted.

What exactly would the UI designers have had to loose from doing a bit of field testing every month from the outset? They had everything to gain and nothing to loose. After all, the UI is hugely adventurous:

• There are no windows, all the applications run in “full screen mode”.
• On the “desktop”, you don’t just see icons of your own files, you can see icons of your friends, and you do thinks with them like chat, draw, browse the web or study together.
• There’s a special button on the keyboard that lets you view and (allegedly) edit the sourcecode of the program you are using at any time.

Innovation is a wonderful thing but you have to reality check: are these concepts really the best approach for the user requirements? And have they been implemented in the best possible way? They are making a lot of assumptions - it’s a big gamble.

I’ve yet to hear about the plans they have for releasing UI updates for the OLPC. Since when in the history of computing did version 1 of anything turn out to be the panacea of good design? At least when you buy version 1 of something, you know what you’re in for. The kids wont have a choice.

> Read BusinessWeek Article (”The face of the $100 laptop”)