Friday 29 June 2012

Things change

The French do things differently ... in technology as in many things.

In the 1950s they had an 819-line TV system that was, even by (some) current standards, high definition. Their colour television was so different to those used almost everywhere else that it was nicknamed 'System Essentially Contrary to the American Method'. And they made a roaring success of a telephone-based text system ... called Minitel.

This weekend Minitel will finally be turned off. It was the place the French went to check the weather and news; book tickets and shop; and find porn. It was, as the BBC story puts it, the France-Wide Web: 26 thousand services used by 26 million people from the early 1980s until this Saturday. To those people in the UK (mostly travel agents) who remember Prestel ... BT's equivalent service ... this will seem amazing. But the key factor was probably that every telephone subscriber in France was given a terminal, even though that did apparently strain the telco's finances.

The news about Minitel prompted me to consider the lifetime of systems, and especially how computing and the internet are constantly evolving; even at a structural level with the change to IP addressing that has just started. Do you have any routers, or computers, that may need modifying to work with IP version 6? Will your machines work with new versions of the OS ... especially if you have an older Mac Pro that won't work with the new version of OS X. In my case, with a small amount of equipment used for work, it's easy to keep track of such things. For a large company it must be a nightmare, and means that you need a reliable system to track the state of all your hardware.

At least when my elderly audio amplifiers went 'bang' the other day I know I can still get them fixed and get them working again ... computers don't work that way, sadly.

Thursday 21 June 2012

Usability – at what cost?

This aspect of interactivity has been one that has changed radically from the early days of IT usability. We used to have to have dedicated usability labs, large numbers of testers, cameras, lights and sound, bells and whistles, belts and braces and so on.

Even Jacob Neilsen, acknowledged veteran guru of the field, has had a very different take on the numbers needed for interactive usability application testing. His latest posting on this makes compelling reading as he advocates five people for the great majority of tests except for certain applications and depending on reasons for testing. The time,effort and numbers involved, let alone dedicated resources like usability labs, mean cost of course, so it is good news that Neilsen’s maximum cost-benefit ratio lines up pretty well with just five test subjects.

Underlying this movement from large expensive tests to small focused ones is a shift away from the onus being put on quantitative research in favour of qualitative research. Yes, I hear your groans! This quantitative versus qualitative debate has dogged art versus science since the academic year dot. However, Neilsen puts a refreshingly simple spin on it.
“... the vast majority of your research should be qualitative – that is aimed at collecting insights to drive design, not numbers to impress people in PowerPoint.”
Jacob Neilsen, June 4th

He goes on to cover some weak reasons why developers are pushed into larger studies and agrees that he himself probably uses more like eight people but gives his reasons. He mentions card sorting. We haven't looked at this specific technique before so what is it?

Card sorting? It is a knowledge elicitation technique for assessing navigation category preferences in interactive applications. This is important when you find that your users are not fulfilling their needs in visiting your site – especially when they can't find the item they wanted to buy, for example. You see the issues. It involves giving testers category headings and asking them to sort them into hierarchies so that navigation preferences can be mapped from use. Remember, insights to drive design. See User Focus for short, uncomplicated answers to your card sorting queries; and Knowhow (non profit), How to use card-sorting to structure your web site, 19th June 2012.

Books have been written about this technique so don't be fooled by the over-simplistic approach to it. If you want more detailed info try Donna Spencer, Card Sorting : Designing Usable Categories, pub. Rosenfeld Media, 2009.

Friday 15 June 2012

Can ... can't ... shouldn't

There have been a few newsy stories that caught my eye today.

There's the schoolgirl whose popular blog about her school dinners has been stopped by the local council. The BBC's Rory C-J reckons it's just a publicity stunt and the council explain that they have now stopped photographs being taken in the school canteen. [Later: they've changed their collective mind now!]

What I find interesting about this story, apart from the delightful blog itself, is that it looks as if 'the management' (fans of Hale and Pace should do the accent here of course) only got worried when the mainstream media picked up on the story.

The Register reports on a statement by Sarah Lamb of the group Girl Geek Dinners saying that sexism in TV shows such as The IT Crowd is preventing women from thinking about a career in IT. Reg reporter Anna Leach refers to Ms Lamb as a Girl Geek Dinner lady (you can see why this comes second in the list can't you?) but funny as that is I applaud any attempt to get girls interested in any kind of science or technology. (I once had a bit of a row during a BBC appointment board with the Editor of Blue Peter about this.)

Mind you: turning it off and on again does fix most problems with digital technology. Buffers get clogged, memory overflows, civilisations rise and fall. You know the sort of thing. And you did watch The IT Crowd didn't you?

Which brings us untidily to the subject of browsers. Once upon a time you had to buy them ... my first even came in a box! For the first edition of our book we needed to put a browser on the accompanying CD-ROM (which was built like a web site) and only Microsoft were willing to waive any licence fee. So I do have a bit of a soft spot for Internet Explorer ... up to it's final majestic Mac implementation of IE 5.5.

Since then we all know that it's been a bit of a pain, especially when it ruled the browsing roost. It was famously quirky (including a quirks mode) and had an unsympathetic approach to standards in the any colour as long as it's black style pioneered by Henry Ford.

I've tended to do web builds that worked about the same across the main browsers, but then I don't do nothin' fancy. The acres of conditional CSS that's needed to cope with browsers these days seems crazy. I could almost believe it's a kind of reverse taxation, where the cost burden is shifted from the browser writers to the web designers.

'Enough!' I hear you cry. Or at least I hear the cry from Australia. The online retailer Kogan.com has now implemented an IE7 tax of 0.1% for every month since IE7 was launched (now standing at 6.8%). Since they sell things like televisions this could be no small thing. Their web site even calls it an Internet Explorer Tax and points out that "you or your system administrator has been in a coma for over 5 years". The tax, says the site, is to recoup some of the cost of coping with IE7's quirks: apparently this was as much time as Chrome, Safari and Firefox combined but for only 3% of the customers.

Do you think you could sell that one to your clients?