autocratic for the people.

Let's See Some Fundamentals Out There.

There's been a lot of Nelson directed at Microsoft for its latest Vista and Office delays. Although I do not purposely keep up with its every move, it's tough not to think about Microsoft frequently when its actions are so universally scrutinized. So, with that as an excuse, I was perusing the Slashdot comments last week and came across this:

"(...) All these years we've been giving MS monopoly rent for OS software in the belief that we were paying for an exciting future, and now the company that's been taking our money is going to give us another 'ticking time-bomb of unstable code'. After five years and more than a hundred billion dollars revenue from computer users, Microsoft will revamp Vista at the 11th hour to turn it into a little more than a skin on XP, which was little more than a skin on 2K. Almost all recent innovations in computing have come from organisations with orders of magnitude less revenue than MS. We are simply not getting value for money." – ozmanjusri

Although certainly more than just a skin on XP, publicly available information leads me to agree that nothing in Vista really raises the bar. The list of new end-user features reads like a Cliff's Notes of the last three years in popular computing (tabbed browsing, desktop search, widgets) along with long-awaited solutions to ridiculously dumb problems (default users are administrators, Windows Update runs through Internet Explorer) that aren't features so much as they are corrections.

The real problem, though, is that almost every one of the features customers are anticipating is already available for Windows XP, for free. Yahoo! has a widget engine. Google, Copernic, and even Microsoft itself has a desktop search tool. Firefox eats IE alive. And there's a wide assortment of free antivirus / antispyware apps, which Vista won't even include.

But that's not the point. It's easy to kick Microsoft for not creating innovative features at the same rate as competitors "with orders of magnitude less revenue." As has been said many times, innovation is not its goal. Microsoft's first goal is to keep Windows a stable platform – which is to say, it does everything in its power to maintain backwards compatibility. It's a noble goal, certainly, and it's the one thing it will always have as an advantage over its smaller competitors, who have neither the people nor the money to cover the enormous cost and complexity of all the testing required to ensure a program written to Windows 95 will still run in Vista. But it's reached the point where I have to question the validity of this goal, and it is in that sense that I agree with the Slashdot commenter.

Backwards compatibility simply isn't sustainable, and in the long run it hurts the platform. It's pretty clear that the tradeoff Vista is making for compatibility is to throw out a lot of new features; to claim a refreshed GUI as a great new feature is to ignore the fact that every new Windows release since 95 has had a new GUI, and it hasn't once been of value. In fact, it's been a huge detriment to backwards compatibility, because every single user needs to re-learn Windows in order to do work they were already able to do in the previous version. It's a huge waste of time and money, and now it's being done to Office simultaneously.

I don't mean to say that familiarity with an interface should preclude attempts to redesign it, but it seems to me (from my very limited vantage point) that Microsoft is wrong to revamp the UI if the Windows+Office combo's focus, and advantage, is compatibility. A UI designer for the new version of Office, Jensen Harris, vaguely cites increasing complexity and difficulty with navigation as reasons to combat the perception of a bloated Office UI, but he then later admits, "we didn't end up making the suitcase any bigger (...) we just added more pockets." Although that's well and good (who doesn't love increased efficiency?), I have trouble believing that Office and Windows customers want an upgrade that breaks all their current knowledge but brings little new to the table.

To be blunt: I don't think Microsoft knows who its customers are. Mr. Harris even says as much when he discusses the long tail of Office commands people use: "(...) Beyond the top 10 commands there are a lot of different ways of using the product." When you have no example user, you can't design effectively. But instead of taking that as a sign, they looked at aggregate click data and focused on increasing GUI efficiency. Who wanted increased GUI efficiency? Home users? Commercial users? Anybody? I don't think so.

• Expert home users have already figured out how to do what they want, and novice home users have memorized or written down a set of steps to follow in order to do what they want. Changing the UI fouls everything up.
  
  
• Commercial users have an expensive infrastructure built around their existing installed software. They have training plans, manuals, support staff, and deadlines. Changing the UI breaks every piece of documentation they have, and causes a huge up-front drop in efficiency which isn't necessarily going to be counterbalanced down the road. Changing the UI fouls everything up.

Okay, so, if not a new GUI, then what do Vista customers want? I couldn't tell you. That's what properly conducted user research is for. Since Mr. Harris seems to think traditional methods are hard (thought: maybe letting users self-select at the end of the post is a bad idea), I would suggest starting the research with tech support call centers for Windows-based apps. I imagine front-line support personnel for AOL, Activision, and, well, Microsoft Office have a pretty good idea of what current trouble spots are – and I bet they wouldn't suggest changing button labels. It would probably also be enlightening to take a tour of public libraries around the US, taking screenshots of the Windows-based public computers to see what's happened to them. And finally, sit in on a few "learning to use computers" night classes to see what metaphors are in vogue for teaching novice users.

...

..

.. Okay, yeah, UI-centric posts tend to peter out disappointingly, because they end up advocating proper research instead of suggesting concrete ideas. So with immediate gratification in mind, here's a stab at what might work. First, split up different versions of Office by functionality as well as by bundle. For example, Office 2007 Student + Teacher Edition doesn't need to be able to import pivot tables or Access databases, so remove those features. By doing this, the designers can reduce complexity by reducing functionality, and at the same time tailor the presentation of specific sets of features to specific audiences.

As far as making Vista a compelling proposition, home users would probably find it valuable if Windows Update were able to acquire and update drivers for all of their hardware. Commercial users probably want a universal mute button that works even during boot, so they don't annoy the whole plane or classroom with their start-up sound. And I bet everyone would like it if the GUI stood pat, or maybe even devolved a bit, if it meant their apps would launch faster and battery life would improve.

Oh, and include a free virus scanner. That one's a gimme.

posted by avixe @ 10:07 PM