AskTog: Interaction Design Solutions for the Real World
Interaction Design Section   Living Section   About Bruce Tognazzini
NN/g Home > AskTog > ReaderMail > April, 1999 Ask Tog, April, 1999

AskTog Reader Mail

An article designed to give you Fitts
Conniption Fitts
Achieving the right Fitts
Available/Select Boxes
Text: the long and short of it
Terror of the rails: The French TGV

An article designed to give you Fitts

Hallo Tog

I ... with shock read your piece on Fitts' Law

I gained a number of insights, and many questions immediately drifted to the surface. One insight was the reason for the ease of use of the Hyperjump menu supplied by Logitech for their trackballs. I am not sure
if this is also available for mice, but the menu pops up directly beneath the place where you click, with various options then available in a circle around this point. Now I find it frustrating to work on a mouse that does not have this function.

At the same time, I wondered what happens to the most accessible areas on the screen. Trackball designs vary. How does the trackball design influence the five screenspots that are most easily accessible, seeing as different muscles are used? (Would there be a research project in this? Being in an Anatomy department.... )

The next question is of course how browser design influences web-page design? If the most easily accessable areas on the screen are the sides, how can I place navigational elements on my webpage in these areas? The top area is already taken up by the browser, so that is out. The bottom contains the scrollbar. The right contains a scrollbar. The left? I don't think you can place anything containing a link directly against the screen's edge!

Should this not perhaps be taken into account when developing HTML- and other standards? Should this not perhaps be taken into account when designing browsers?

Marius Loots
Athens, Greece

In the study I did on mice vs. trackballs, trackballs were rather consistently around 50% slower than mice, regardless of whether the ball was large and weighty or small with little or no inertia. Certainly, the combination of a small ball with no inertia and a poorly-designed accelleration algorythm (such that the mouse pointer does not move proportionally farther depending on the accelleration of the device) could doom any benefit of corners and sides.

Trackballs, in general, deserve more study and effort than they have received. I've used a number of commercially-successful units that are just plain bad.

Neither site designers nor, alas, browser designers, can take advantage of corners and edges unless the system software architect will give them access. Unfortunately, many of the biggest increases in usability can only come from the system software houses, and none of them have done much in this area since 1984.

Conniption Fitts

Hi, Tog,

While I agree with most of the answers on your Fitts quiz, I do take issue with the answer to questions 4 and 5.

I actually HATE that the mac has the menu at the top of the page, every page, all the time. It’s not a question of speed however. As someone who does web design for a living, the shortest distance between two points is actually the shortest distance my mouse has to go across the screen. I usually have at least 4 applications open at one time and I constantly toggle between them. The more I have to drag the mouse, the greater the pain in my shoulder and back. The more pain, the less work. Most of the time I actually use keyboard shortcuts for my applications anyway.

I prefer the Windows Toolbar layout because:

  1. I can always see which items/applications are open without having to click and drag on menus
  2. I can set it up to fit with what I’m working on -- on the side, bottom, or hidden -- depends on how large my file is
  3. It gives me access to more information faster -- take a look at Windows 98 and see the icons to switch to the Desktop, open key applications or set up my sound levels.

Yes, there are some problems with the implementation of the “hidden” mode. But overall I’m much happier doing my work on a PC -- its much faster and more efficient for web design.

I’ll have to dig out a message I sent to Apple last summer with some criticisms of their GUI.

One more thing, about icons, they aren’t necessarily the most efficient system of communication by themselves because they can be quite cryptic. If they are graphically similar, they can actually waste time by tricking the user into clicking on the wrong one.

Keep up the great work.

Ann Donovan

Now that I've decided that both the Windows and the Mac interfaces suck, I'm not nearly as anxious to defend the Mac. However, with the Mac, using System 8.5 or later, you can leave the applications menu open whereever you want it on the desktop, so its only drawback is that you can't arrange it horizontally, as you can with Windows. (Or if you can, they've made it one of those infuriating little Macintosh hidden tricks, so only people with connections can find out how to do it.)

To traverse the entirety of my three-screen (two 17" and one 20") display requires me to move my mouse (quickly) the grand sum of four inches. To access the pull-down menu bar usually requires movement of less than 1.5 inches. Still, Apple could learn a few things from Microsoft, and Microsoft could learn just about everything from the Mac.

How any company could spend a decade appropriating the work of someone else and get it so uniformly wrong is beyond me.

Achieving the right Fitts

In the Fitts quiz, I think question 3 is a bit misleading. It says “A right-handed user is known to be within 10 pixels of the exact center...”, which to me implies that we DON’T know the exact coordinates of the pointer (only its general location), so how can we know which pixel it’s on?

Also, while pondering this question, I discovered that Windows NT 4.0 (and presumably 95/98) actually does use the upper corners of the screen. If you maximize a window, then clicking the upper-right corner activates the close-window button, even though the button does not appear to encompass that pixel. Imagine my chagrin when your web page disappeared with no warning! The upper-left corner pixel counts as part of the window-menu button.

Question 3 was intended to be misleading. I wrote it using a technique to which I was introduced in Junior High School. However, when I wrote it, Frank Ludolph had not offered me the solution based on knowing what pixel it is on. I would still argue that neither the user nor application programmer need know what pixel it is on. Only the system programmer need know, offering a path to the “current pixel,” without reference to where the current pixel is.

As a general rule, I’m in favor of having the visual region and “touchable” area of a button share the same border. Microsoft broke the rule presumably for the sake of either aesthetics or speed-of-coding. Neither are strong enough reasons for setting up a situation where the user could experience mysterious system behavior, particularly when it results in windows disappearing entirely.

Available/Select Boxes


I have a question about Available and Selected lists. Let’s say the user can chose to see the daily headlines of numerous major newspapers in the world. To configure this, they have a list of the few hundred newspapers available (the Available list) and they have a list of all the papers they have selected (the Selected list).

Here’s the question: When presenting this to the user, should the Available list come first (above or to the left) or should the Selected list come first?

The example I have picked is similar to what our users will do-initially pick a dozen or so from hundreds, occasionally going back in there to see which ones are chosen and removing some, adding more.

What do you think? I have seen it both ways. Microsoft often does it with Selected on the left. KL Group has an Available/Selected 2-list bean like this with Available on the left. Your speedy input would be highly valued and appreciated!

Thanks in advance!

Tim Dowd

Microsoft often seems to do things backwards just to be different. For example, they elected to have the OK button on the left, instead of the right as it had been traditionally. This kind of perversity tends to come back on you, which it did on Microsoft with the advent of the web with its Next button on the right. So now, half the time you continue by pressing the button on the left and half the time you continue by pressing the buton on the right.

Similarly, the phone company just had to be different by inverting the traditional numeric keypad with the low numbers at the bottom, placing them, instead, at the top, thereby causing mass confusion as phones and computers merge.

The Available list belongs on the left, where God intended (unless you are designing for a Middle East audience that reads right to left).

Text: the long and short of it


Is there a reason that you choose to make your pages incredibly long with basically just text. Doesn’t this seem to break some sort of rule?

Just curious

Alan Ruthazer

Um, I guess it seems better than making them incredibly wide.

Our studies at Healtheon have shown that long pages, rather than myriad tiny pages, are better. Clearly, the world needs better text presentation mechanisms, but what we seem to have available are crude browsers and defective PDF viewers. Not much of a choice.

Terror of the rails: The French TGV

Hi Tog,

Your account (in “the Lighter Side”) of the Bullet train blasting through the station in Japan brought back a similar memory, and made me smile.

For two years (‘95 and ‘96) I was a missionary for the Church of Jesus Christ of Latter-Day Saints serving in the France Bordeaux mission. Missionaries usually labour in a given town for four to six months, and then “transfer” (always by train) to a different area. Then there are conferences to attend, trips to nearby cities for visa paperwork, etc. etc. By the time you’ve been in France six months, you’re pretty used to trains.

Most of my time was spent in the southwest part of the mission, in small towns between Bordeaux and Toulouse. I served two stints in Agen, a small town almost exactly halfway between those two cities. Since there are several TGVs (the French high-speed train; the fastest in the world) running between Bordeaux and Toulouse every day, I was in the Agen station a number of times when a TGV came ripping through at over 200KPH.

As impressive as that was, late in my mission I had a TGV experience I’ll never forget.

My companion and I had gone out to a little village called Aiguillon to visit a member of our church. Aiguillon is about a 20-minute train ride from Agen, in the direction of Bordeaux. Its train station is an old, tiny, rural thing: just a weathered stone building fronted by an almost-ground-level platform. There are just two tracks: one for each direction; and nothing but open sky once you come out of the little building. No public address system to announce the arriving or departing trains, just a little bell/buzzer that sounds when a train from either direction gets close enough. (This is in contrast to the relatively large Agen station, with its six tracks, fully enclosed from the main building to the last track, with underground walkways connecting the tracks and an automated PA system to announce the trains.) Our little episode occurred while waiting for our train to return to Agen.

Trains generally run precisely on time in France unless there’s a strike on. As luck would have it, there’d been perturbations at Bordeaux that day, and so our train was running late -- very late. So there we were, alone on the little rural platform, in the eerie silence of the countryside. It was late November: cold, damp and dark at almost 10:00 at night, and we were waiting for a little regional TER train (known as a “cattle car”) and beginning to wonder if it was coming at all.

We both started (startled?) when the little bell/buzzer went off, announcing our train’s arrival. I thought, “Well, finally...” and walked to the edge of the platform (another step and I’d have been on the tracks) to watch the train come around the bend about a quarter-mile away. It appeared quickly -- much sooner than I was used to. I thought “mmm..... that’s odd, I wonder why the... hey, waitaminnit, that thing’s going WAY too fast for a TER...” It didn’t look like it was stopping, either, so I started to step aside.

WHOOMP! I had moved maybe eight feet back from the edge when the TGV blasted by me, a high-speed juggernaut that shredded the night silence, screaming like a banshee as its cannoned tens of thousands of tons past me almost close enough to touch, moving too quickly even at that distance for me to see the faces in the windows. The wake it kicked up tore at my clothes and lifted dust and paper as far back as the terminal building.

And then it was gone. Half again as long as a normal train, it was past me in less than three seconds, and disappeared around the bend some 150 yards away almost before I could turn my head to catch it.

All my comp and I could do was look at each other in stunned amazement and say “Wow!”

As it turns out, I was lucky. Even the TGV is limited to “only” 224 KPH when running on normal tracks. If I’d been standing that close and it had passed me on high-speed rails at its standard cruising speed of 380 KPH, who knows what would have happened. The funny thing was I didn’t feel danger at the time. Just as you described, it was exhilarating: whoa, cool! It was over before I’d even had a chance to react.

Anyway, thanks for letting me write that out. I enjoy your website and am a regular reader who benefits greatly from the more serious content.


Chuck McKinnon.

Don't miss the next action-packed column!
Receive a brief notice when new columns are posted by sending a blank email to

return to top

Contact Us:  Bruce Tognazzini
Copyright Bruce Tognazzini.  All Rights Reserved