AskTog: Interaction Design Solutions for the Real World
Interaction Design Section   Living Section   About Bruce Tognazzini
NN/g Home > AskTog > ReaderMail > January, 2000 Ask Tog, January, 2000

Ask Tog Reader Mail

 Giving Software Houses Permission to Sell Crap
 On Reputation Managers
 Counterforces to Rip-offs
 Guidelines Ethics Revisited
 Responsibilities of Reviewers
 Tog Soft on Microsoft?
 Fonts: Serif vs. Sans Serif
 Mouse vs. Keyboard under Windows
 Apple and Two-Button Mice
 Vinyl and Vacuum Tubes
 In Defense of Traffic Engineers
 Alcatraz for Systems Engineers
 The 10¢ Light Bulb that could have Saved a Ship

Giving Software Houses Permission to Sell Crap

I read with great amusement your Century of Scams article, particularly the last section, arguing for REAL warranty/fit for use protections.

Sad to say, the world seems to be headed just the opposite direction. When you feel the need to be supremely depressed, do a little reading about UCITA, the Uniform Computer Information Transactions Act. This one makes most of your other scams look like pranks from a not-too-imaginative third-grader.

A good place to start would be

Enjoy your depression!

-Chris Sandberg
(alas, I am an attorney...)

Now, see? That's the real trouble with lawyers. Get one alone and most of the time they turn out to be really nice guys. It's only when they get together that they create great mischief. Makes it hard to do something about it.

The scam to which Chris refers is astounding in its audacity. Essentially, the software houses are attempting to buy enough legislators that they can get passed a bill that will not only enable them to keep marketing dangerously defective software, but will prevent the press and individual users from blowing the whistle when they do so.

While US citizens might assume this latter provision can't be enforced since we enjoy a First Amendment right to free speech, that would be an incorrect assumption: The provisions of the Bill of Rights pertain only to freedom from government restrictions. Our Founding Fathers never envisioned the bloated, amoral multinational corporations that threaten our personal freedom today.

Reputation Managers


What are your opinions regarding “reputation managers” and their affect on the quality of products that we purchase?

Will simple “scams” become less likely to succeed due to our aggregated knowledge?


Yarone Goren

Reputation managers are those feedback systems some web sites employ to enable users to publish their opinions. has the simplest of systems, enabling both users and authors to speak out on individual books.

Ebay has a more proper system, in that buyers' and sellers' opinions of each other are not just reported as a series of letters, but tabulated into a "score" that can easily drive a bad-faith user from the site entirely.

Both these managers share one characteristic: They are reporting on "the little guy." Even though a book might be published by a major publishing house, no one is picking on Random House or Addison-Wesley, they are picking on Joe Blow or Jane MacGillicutty. While Random House or Addison-Wesley might be adversely affected if a book is panned, this country publishes around 10,000 books a year. A few bad reviews aren't going to affect the bottom line.

A much more interesting experiment is's reputation manager. Here, users can directly attack prized brand names of major corporations. This is a really powerful tool for consumers, since we crusty old-timers can find out from real-live kids which toys "rock" and which toys "suck." However, such scores could make a mess of etoys attempt to balance their inventory and it is not clear whether Corporate America is ready to let consumers "tell it like it is."

I found that the incidence of reviews on etoys to be rather uneven. For example, two out of every three toys I checked on their "best sellers" list had reviews (most were glowing). However, I was unable to find even a single review on any of the "Holiday Clearance Sale" toys. (I did not examine each and every one, but I looked at enough to discern the pattern.) Coincidence? It could be, since these poor sellers may not have had many people interested in reviewing them. But it could also be that the reviews had been stripped off. (Nowhere could I find etoys' policy on its reputation manager.) If so, it brings the entire system into serious question.

My prediction on reputation managers is that they will prove honest, forthright, and invaluable on sites that link you to "little guy" products and "little guy" merchants. Conversely, despite what etoys is apparently attempting to do, I expect that as soon as the reputations of brand names of major corporations come into play, the reputation manager will soon become disreputable as sites manipulate the information to their own—and their precious suppliers'—ends.

In the end, consumers may have to continue to depend on independent voices, such as that of Consumer Reports, to get honest, factual information on the products of our increasingly corporate world. I would give you the URL of Consumer Reports' website, but it is a rip-off. Better to buy and save the magazine than to pay for the same information twice. (Reader Erik Neu points out that you can subscribe for $3.00 a month. So you can join, get the info you need, then cancel your membership.)

The one way out would be a reputation manager that was not hosted by any one site, but which lay between you and the sites you visit. Such a manager could truly be independent. It would also help solve the problem of "scams," since the manager would let you know, independent of the site, that either the entire site or a product on it was a scam.

Such a reputation database could be further extended to email, marking or filtering incoming spam that, prior to your logging on, has developed a reputation for being junk.

Counterforces to Rip-offs


As an ongoing look at the scamming of software (not necessarily just PC's - I mean c'mon some UNIX vendors made you pay extra for TCP/IP!) why not take a short glance at the Open Source Community?

Whether or not you believe it to be damned to failure - you have to admit it is a phenomenon directly related to your article. I mean more before it became apropos - it was about quality and for many it still is (myself included).

Granted, I don't offer up a robust UN*X system to grandma as an alternative OS - but I just might write grandma a fetchmail client because hers sucks. I have written a ton of simple little apps for my own and others use using nothing but rock solid OSS.

I really believe there is something to that and it will play a part in the future.

Plus, I would like to kill a lawyer too if I might.

Think about it.

Jason R. Fink
Editor/Author UNIX and Linux Computing Journal

If you must kill a lawyers, please pass over Chris Sandberg, above.

The user interface to Linux sucks since it isn't one; it's two. And two automatically sucks. On the other hand, Linux does have the bastards on the run, and I can't help but applaud that. (Besides, the interface to UNIX, all 27 of them, suck, too.)

I didn't intend, when I sat down to answer these letters, to climb quite so high on my soapbox, but now that I'm up here...

We face a grave and unprecedented situation today when it comes to intellectual property. Over the last 100 years the corporations have made it their business to squeeze all the little guys out and take over. Patent laws were intended to protect inventors; today, independent inventors are a thing of the past. Patents are owned by corporations. (Yes, I read about that guy who fought the auto companies over the delayed-action windshield wipers they just blatantly stole from him. But did you also read that the only way he could pursue them was to sell his rights to another big corporation with pockets deep enough for the fight?)

Copyright used to be the life of the author + 50 years, allowing us writers to pass on our legacy to our children. Now, it's a blanket 135 years, because corporations don't die. And if you want to see your book in a bookstore, you don't publish it yourself, unless you are highly motivated. Instead, you go to a publisher, and they get 90%+ of the proceeds.

Trademarks, of course, go hand in hand with being a corporation, not a little guy.

Computer technology, while helping the corporations strengthen their stranglehold on intellectual property, have also been a fresh breeze of liberation. Jason has mentioned only the latest schism—Linux. Before that, there was the explosion of MP3 audio, which is scaring the bejesus out of the record industry. Before that, and for a long time, we've had the newsgroups, anathema to the corporate plan.

Would it be healthy if all intellectual property rights were to disappear? Probably not. So much of what we enjoy today, from major motion pictures to the computer you are using, is the result of years of effort by hundreds of people, costing tens and sometimes hundreds of millions of dollars. But these bursts of freedom keep industry from grossly overcharging (even more than they are), knowing that if the price gets too high, people are going to download and burn their own copies, resulting in zero revenue.

It's going to get really interesting when we truly have high-bandwidth internet connections and can download a two-hour "liberated" movie in one or two minutes. A couple years ago, the studios were charging $70 for film on VHS tape. Now it's $25 for a high-quaility DVD. Do I hear $10 for an HDTV DVD when I could download and burn it myself for $5?

Guidelines Ethics Revisited

Last month, Jeff Kandt reported in Cursor Keys Cursed that the Macintosh Human Interface Guidelines waffled on how cursor keys should act during a manual selection process. The Guidelines specified the proper selection rules, then suggested that if it acted a different—and decidedly brain-dead—way that somehow that would be all right too. In the ensuing time, we've tracked down the source of this aberrant behavior to which the Guidelines referred, namely, Apple's own "TextEdit." While most developers have had enough sense to not use this particular aspect of the code, at least two have: Eudora and Apple.

I open the issue again because, if you are writing guidelines, you have an ethical obligation to write good guidelines. Writing defective ones just because your company currently has a piece of code with a bug in it doesn't fly. Apple has had several bugs in the Finder ever since 1983, but, until the latest Guidelines, none of us ever thought to make them into features.

For example, the Guidelines, since the early 1980s, have always called for the Shift key to continue a selection, while the Command key (equivalent to Control on Windows machines) was to be used for discontinuous selections. Every application I know of on the Macintosh follows this guideline with the exception of the Finder, which uses the Shift key for discontinuous selections and does nothing with the Command key. For the last 16 years, Mac users have had to carry around two sets of rules for one of the most basic activities on the computer.

16 years! Why not just write it into the guidelines after all that time? Because it is a bug, not a feature. It would be equivalent to having a freeway between San Francisco and Chicago where you drove on the right, all except that 20 mile section near Kansas City where you drove on the left. It wouldn't matter if that freeway was open 500 years, the instructions for that freeway should never be, "drive on the side of the road where you see other cars going your way."

And another thing: Just because it is an old bug doesn't mean it can be ignored. Every time there is a new release of the software, you should put back into the system every single screwup your users have been putting up with. Every one. And label them Priority One, because that's what they are.

Be radical about reporting the bugs, but be gentle on your users in correcting them. For example, the Finder should not suddenly no longer allow users to use the Shift key for discontinous selection. Instead, the Command key should work also. That would overcome the kinds of errors users are making in pressing the Command key in vain, while not screwing up the behavior users have come to know and hate.

Responsibilities of Reviewers


....I think more blame for bad software has to be placed on the shoulders of the consumer and the media. I remember an excellent word processor called Fullwrite for the Mac. After a dodgy start, it shipped with more features and more power than any other word processor at the time. It was also surprisingly easy to use and didn't clutter up the screen with a myriad of panels, panes and UI miscellanies. In essence, it was years ahead of the competition.

However, little work was done on it to add new features. I recall at the time that the press, MacWorld in particular, were already quickly writing obits for the product after barely a year without upgrades. The fact that the rest of the pack hadn't even caught up with it was of no matter. So rather than correctly noting that Fullwrite was still ahead and had features the others should strive for, the media effectively wrote it off and helped kill it off.

Thanks for the great stuff and keep it up.
Jeff Lewis.

And another on the same subject:

In response to your column about scams, particularly the part about the printer with annoying Ethernet problems a reviewer failed to note:

My overall opinion of the computer trade press was already declining several years ago when I heard a remarkable story from the sysop of what was then one of the largest Macintosh-oriented bulletin boards in the world (back in the long lost days of the standalone, modem-based BBS). This individual owned a consulting firm which, among other things, offered emergency bailing out for troubled programming projects.

He once posted the story of a client (he never gave their name...) which developed specialized PC software. While he worked to clean up a mess for them, one of their shipping programs won a PC Magazine Editors' Choice Award. Everyone in the company was beside themselves with glee. But as time passed, he gradually learned from the company's employees that the program in question was so defective it failed to deliver even the functionality promised on the box.

Intrigued, he began his own investigation. Eventually, he found that the reviewer had not actually run any of the software in question. The review had been done by listing all features promised in the manual and naming the program with the longest resulting list to be Editors' Choice.

Over the years this story has helped me understand a great deal about this industry and its products.

-James Austin

Technology reviewers are not, in general, fattened cats surrounded by cool gear in high-tech workplaces, contrary to our image of same. Instead, they tend to be independent writers living from job to job, just trying to get by on a lot less money than most of us.

That's one reason. The other is who they work for. Magazines have the same trouble as etoys and other similar websites will face: Their income comes primarily from the very people they are reviewing. Talk about conflict of interest! Put a copy of Consumer Reports alongside any technology magazine you want and look at the immediate a pervasive difference in tone. Consumer Reports expects the worst, and often finds it. Computer, Audio, and Video magazines expect the very best and are rarely disappointed.

The web at least opens up the possibility of nonprofit, ad-free zine that could report the true facts on technology, one with a testing budget that would eventually rival Consumer Reports, but without the overhead of printing and distribution. Anyone want to put together a business plan?

Tog Soft on Microsoft?


In your "A Century of Scams," did I read this right?

"Rather than going after Microsoft for selling a multi-billion dollar OS for the outrageous sum of less than $100, why not go after the entire industry for the real crime—selling bug-riddled, productivity-sapping junk?"

In the 15 years that I've been working with computers, I've had more problems with Windows than any other OS, application software, computer hardware, or network hardware. Have you tried installing Office or Windows NT lately? Microsoft is criminal #1 when it comes to selling bug-riddle, productivity-sapping junk.

Love the site,

Mark Robillard

I meant in no way to let Microsoft off the hook, only to hang a bunch of other rascals up there with them.

Fonts: Serif vs. Sans serif

Hi Tog,

Have you noticed that MOST, if not ALL, big websites these days use sans serif fonts? Especially the corporate ones. There seems to be some big, unwritten rule among website designers that serif fonts are to be avoided at all costs.

I have always believed that serif fonts are easier to read—not only on paper, but on screen as well. I should think the embellishments to a character, besides making it look prettier, also make it easier to be recognized by the brain since there are more data points per character. Is this not true? Why do you think websites aren't simply designed for the default font?

Why does Microsoft think the new Verdana font designed for the web should be "stripped of features redundant when applied to the screen"?

It just seems there are too many threadbare, ugly chunks of text out there these days.

—Raj Premkumar

While I think the argument that a font or any other interface feature should be "stripped of features redundant" is specious, I have learned since first taking the position that sans serif fonts are bad, that a special case exists when said redundances are so badly distorted as to be unrecognizable.

Such, I am informed, is the case with serifs on today's crude screens. Two font designers who have worked both in the world of newspapers and computers, Vlad V. Golovach and George Olsen, have taken the time to explain to me in exhaustive detail exactly why serifs don't work on low-resolution computer displays whereas they do in newspapers which, while notoriously bad, don't hold a candle to the crudeness of computer screens. (A third writer from UCLA, who shall remain nameless, seemed to share their opinion, but only imparted the direct information that I am a fool. This seemed only tangentially relevant and was, in any case, stale news.)

These arguments have caused me to reverse my previous "serifs at all costs!" position until such time as we have high-resolution displays, at which point I will be back out in front of the serif parade once more.

(Grasping at straws, I was even prepared to put forth the argument that the muddying effect of seris could be offset by anti-aliasing the font. But anti-aliasing softens the edges, and soft edges definitely reduce readability.)

I would really like to see someone to do some serious large-scale studies of relative readability of various fonts, serif and sans, and then I would like to see those results codified in the form of standard, cross-platform, high-readability fonts. In the meantime, I'm concluding my discussion of fonts.

Mouse vs. Keyboard on Windows

Hi, Tog,

I continue to enjoy reading your site! As I go through more and more of the "back issues" I keep discovering new gems. I especially like the insights into the design of the Macintosh user interface (finally I find out why it is supposed to be so good!).

But, enough of this flattery! I wish to dredge up an issue which you have probably heard too much of - keyboard shortcuts versus mouse actions. Now, I know your studies show that keyboards shortcuts are slower than the mouse, but in this day and age of Microsoft Windows, is this really true? Given that in Windows the menu bar is a relatively small target compared to that on the Macintosh (not to mention the perils of trying to click on the toolbar buttons), and given the hideous way MS menus work, and a multitude of other design flaws, is it not quicker to use shortcut keys?

I would have thought that on Windows, reaching a menu item required a similar level of cognitive processing to using a shortcut key; in fact, for the rapid typist who "just knows" where the keys are on the keyboard and can type them without really thinking, maybe the keys are easier and quicker to use?

I would be interested in your comments on this, although it is an old topic.

Keep up the good work!

—James Corrin

Microsoft engineers have always apparently had a hate-hate relationship with the mouse. From the beginning, their software has evidenced a concentration on the keyboard, such a heavy concentration that the mouse interface has sucked. Their prejudice has been accompanied by a total disregard (or, more likely ignorance of) Fitts' Law, resulting in an interface seemingly conceived to slow the mouse user down. With such a defective interface, it is little wonder that the MSDOS/WIndows nerds "discovered" that mice suck. Until Microsoft abandons its keyboard fixation and adds a true mouse interface, the keyboard, under Windows, will continue to be superior.

You would have a similar situation in the real world if Ford had mounted the steering wheel behind the driver, instead of in front. Having to drive with your hand bent around behind you might well have proven so difficult that the other leading contender for steering automobiles, reins, might well have won. I suppose we should be grateful that Charles Simonyi, et. al,  weren't alive back then.

Apple and 2-Button Mice

My friend has recently been forced to use Windows and, while he prefers
Macintosh, he does think that two button mice are pretty cool. I'm a firm
believer in the one button mouse for the simple reason that I never press
the wrong button. I argued that two button mice were slower than one button
mice for two reasons:

  1. The user has to stop and decide which button to use.
  2. The effect, under Windows, of the second button is to cause a pop-up menu to show up. If this is a hierarchical menu, it's slower than using a pull down menu (reference: "Tog on Interface", p. 203, chapter 27, paragraph 5, verse 1). However, since Windows does not have a menu bar like Macintosh (ie, bound the top of the screen), it may be just as fast or faster.

My friend had many of the same arguments that I've seen used for the mouse versus keyboarding, such as "Once you get used to it, it's much faster." and "It feels more responsive." I've also heard lots of nerds bemoaning Apple not shipping two button mice.

Do you know of any learned studies on this subject?

—Peter Merchant

You may add me to the list of nerds who used to bemoan the lack of a two-button mouse on the Mac. Now I have a Kensington two-button mouse on the Mac. Making it standard is long overdue. It's no longer 1981, when people were drawing Etch-A-Sketch art and writing half-page memos. Apple has, of course, always had a multi-button mouse. It's just that the other buttons—Command, Option, and Shift—were on the keyboard. Now, instead of adding a second button for use with pop-up menus, they've squandered the Control key. Foolish.

As for hierarchical menus, it is true that a linear menu on a menu bar anchored to the top or bottom of the display (as is done on the Mac, but not on Windows) is significantly faster. On the other hand, the pop-up menus appearing now tend to be context-sensitive, rather than being simple substitutes for what might be found above. Such menus greatly expand the range of commands that can be offered the user, a necessity with today's complex applications.

Looking forward, both Apple and Wintel should get their square-fingered paws on a Palm Pilot and figure out where pointing-device interfaces should already be. Even though Palm is somehow ignoring a lot of the middle-ground of gesture that lies between simple pointing and writing, the lilting freedom of direct, lightweight pointing is awesome.

Vinyl Records and Vacuum Tubes

Hi Tog,

I thought your latest article [on scams] was great (as always), and I wholeheartedly agree with your views on software. But I must take issue with you over some of your statements about music technology.

You are correct in your statements about wattage ratings on amps, but here in the uk they are given 2 ratings, the one that is usually quoted is Music Power, which is the amplification of audio signal, and is how they quote a 75W amp as giving 200 W of amplification.

However, the use of tubes in amps is not a scam, they do give a better quality of sound, as any serious audiophile would agree. I studied TV & Radio technology at college for 4 years and have seen the evidence with my own eyes (on an oscilloscope). amps with tubes in them have a different harmonic resonance to silicon amps and give a more natural and pleasing sound.

As for CD's versus vinyl, sorry but vinyl is better for sound quality but, yes CD's are more durable. The problem with cd's is as simple as this, CD's have the sound digitized and chopped into bits, and then they throw away all the bits in-between. Its lossy compression, its not natural sound, they can never fully reproduce the sound any more than a jpg can fully reproduce a picture as well as the original photograph can.

This is also why I am still hanging off from going over to DVD, The color separation is great and the picture is sharp, but I have seen some artifacting and scenes with very little movement (especially in backgrounds) look very artificial.


Advanced Group
Peterborough, UK

"Music power," I still maintain, is a scam. Continuous RMS is the only honest measure of an amplifier's output. True, music power does perhaps more accurately reflect what an amp might do in "real life," since normally the capacitors have a chance to charge back up between crescendos. However, there is no way the music trade would be so interested in pushing "music power" vs. RMS if "music power" came out to half the number of watts, instead of twice the number. And I will still insist that if it pulls 75 watts out of the wall, there's no way it's pushing 200 watts into the room.

As for vinyl, while it may not by "lossy," it is most certainly "gainy": it gains wow, it gains flutter, and it gains tens of thousands of defects even before the first play. I would much rather accept the theoretical limitations of CDs than the glaringly obvious limitations of vinyl.

I agree with you that DVDs are also a compromise, but surely you don't favor VHS over them. Perhaps you do favor Laserdisks. Laserdisks, however, are far from perfect and, perhaps more importantly, are dying. They are unlikely to achieve the kind of cult following that vinyl has, and when our current players die, we will be hardpressed to replace them.

Fortunately, both CDs and today's DVDs will soon be supplanted. The two new DVD audio systems both have a high enough bit rate that they will effectively be lossless. On the video front, Pioneer has a new prototype HDTV DVD system using a purple—not blue—laser. It will be capable of storing four hours of HDTV quality video on a single DVD. Unless they screw it up, the disk's total capacity and transfer rate is sufficient to store high-definition video without visible artifacts.

Finally, tube amplifiers: I also like the softness and warmth of tube amplification, but at what cost? Tubes, particularly power tubes, quickly lose their original specs. And these guys are selling their tubes for perhaps ten times the original prices. Am I expected to spend several hundred dollars a year just in normal maintenance on an amplifier in order to keep it up to spec? And what am I getting for my money? Lets take a look at the creme de la creme of tube amplifiers, the new McIntosh Limited Edition MC2000. According to McIntosh's ads, it offers "a monumentally-powerful 130 watt/channels" with a "minimalist .5%" total harmonic distortion. I can show you Radio Shack amplifiers with that much output and 1/10th the total harmonic distortion figure. And you won't have to sell your car to raise money for one.

At the same time, it cannot be denied that tube amps have that "certain something." And, in all fairness, a reasonable amount of that extremely high total harmonic distortion figure is probably the "certain something" we're looking for—distortion isn't necessarily all bad. However, I maintain that you could build a solid-state amp that would mimic that softness and warmth, you could do it for a lot less money, and you could do it in a product with zero maintenance. This need for destructively-hot tube amplifiers is nothing but fashion.

And then again, I must confess I drool every time I see a McIntosh. Always have. And I've had a love affair with tubes ever since I took apart my first radio back in the 1940s. I just don't like having to pay $50.00 for a $5.00 tube just because the other rubes don't know any better. If you know how I can get a McIntosh cheap, along with a couple boxes of tubes, be sure and let me know.

As this reader soon attested, simulating a tube amplifier can be done:

While I was an undergraduate at MIT 25 years ago, my roommate worked on a senior-year research project that did exactly that. They sat people down and had them compare the sound of a tube and a transisitor amp. Then, based on an analysis of the signal output, they built a relatively simple, solid-state signal “conditioner” for the transistor amp. With the conditioner in place no one could tell the difference anymore. Of course, no one would pay large sums of money for such a simple device when they could buy the “real” thing!

-Earl Waldin

In Defense of Traffic Engineers

Dear Tog,

As a traffic engineer, I read with great amusement your column of traffic engineers´ work in USA. There would be a lot to say about what is going on in USA in the field of traffic engineering, but I spare my comment to the basic flaw ( in my opinion) in your thought in your article: a traffic engineer would like to optimize the whole transportation system; your point of view was the user´s or individual´s. This means that the whole system might well take advantage of the work of the traffic engineer, but you as an individual might suffer.

There are also other time savings than the ones concerning expected lifetime; yes, also travel time savings.

The usual way to evaluate the feasibility of a transportation investment project is to calculate the benefit-cost ratio: you compare the effects of the investment to the base scenario (usually “do-nothing”). The calculations are based on forecasts, driving costs, time based costs and accident costs in dollars. It is possible to have costs on environmental impacts and external effects, too, but usually those costs are difficult to understand for the decision maker. The benefits are the difference of the costs of two scenarios. You will get the benefit/cost ratios when dividing the benefits by the investment costs. Now we are talking all the time the total system, what ever it is, the a corridor, a city, a nation etc. Is this the right thing to do, we can argue.

I still think you have very good ideas on your special field, but traffic engineering isn´t quite that simple as you indicated in you column.

Tapani Sarkka
Helsinki, Finland


Dear Tog,

To be sure, traffic engineers have made some bone-headed moves over time, but you neatly ignore Americans’ irrepressible love affair with the car, the (lack of) urban planning, and the neglect or even active sabotage of public transport that led to so many cars on the road in the first place.

Traffic engineers are, for the most part, merely attempting to make the best of a bad situation, and trying to saddle them with the entire blame for overcrowding and auto pollution is just plain silly.

—Christopher Nebel

And when have I ever avoided being just plain silly? Besides, I've only blamed traffic engineers for the mischief traffic engineers have caused. They are hardly responsible for our unbridled love affair with love affairs that have resulted in the massive overcrowding responsible for most things wrong in our countries, including water shortages, energy prices, air pollution, etc. Nonetheless, their bone-headed moves have added to my general discomfort.

Tapani, I think, pointed out the crux of the matter:

A traffic engineer would like to optimize the whole transportation system.... This means that the whole system might well take advantage of the work of the traffic engineer, but you as an individual might suffer.

Such a viewpoint not only ignores the user completely in its focus "on the greater good," it leads to just the sort of patronizing attitude to which I'm objecting.

Overpopulation is putting pressure on everyone to make life more restrictive, but where are the checks and balances? Where are the advocates for removing traffic lights in these planning sessions? Who is there in the engineering meetings to champion the changing of double-lines into broken lines?

Several years ago, I read in the newspaper that, in a small town in the Midwest, they were removing a traffic light. They'd decided it was a great bother and they didn't need it. The story was circulated throughout the USA and probably the world. Why? Because it was a man-bites-dog story. It was unusual. It was unheard-of.

When I visited Moscow a few years ago, I saw hundreds of controlled intersections, complete with lights. The Soviet Union, having to put its entire population to work, had liberally sprinkled such intersections throughout their major cities, but had never automated the lights. Instead, a worker would stand in an elevated phone booth overlooking the traffic and manually switch the lights. By the time I visited, the workers were gone and the lights were out. Strangely enough, the traffic moved just fine, and I experienced a sense of freedom I had almost forgotten.

Traffic engineering has brought us the modern highway system, with its limited access cloverleafs. The roads are safe, fast (except for the effects of overpopulation), and a pleasure to drive. Traffic engineering has also brought us restrictive roadway markings and ridiculously long light cycles. It's time these engineers started thinking a little more about the suffering individual and a little less about "what's good for us."

Alcatraz for Systems Engineers


I have a niece who is twenty and retarded. She plays with her computer while her mother and father are at work. One day recently her mother came home to find my niece screaming. It seems earlier in the day the computer had displayed a message of "Illegal Operation" or something similar. My niece spent the rest of the day in abject fear that the police were coming to get her because she had done something illegal.

Sigh. Any comment?


When San Francisco's notorious Alcatraz Prison closed back in the 1960s, it was suggested a good future use would be as a sort of Ellis Island for migrating New Yorkers. Like New York's Ellis Island, that welcomed so many generations of immigrants, Alcatraz would be a place where hardened New Yorkers could learn to speak English and gradually acclimate to a softer, more polite society. Sadly, the plan never reached fruition.

Perhaps it is time to revive it for a certain subset of systems engineers. Alcatraz could become a safe haven where these introverts could gradually, over time, be introduced to normal human beings, where they could learn in a safe, supportive environment exactly how destructive their cutsy little-boy vocabularies can be. They might be shown videotapes of Carl's niece cowering in fear. Or the woman who, early in the life of the Macintosh, saw the Mac system-failure bomb icon and ran screaming from the house, yelling for the neighbors to call 911.

As for the old prison block, that might be as good place to house the college professors that allowed these individuals to graduate so plug-ignorant.

The 10¢ Light Bulb that could have Saved a Ship

As part of their Modern Marvels series, The History Channel ran a program on "great engineering disasters" (which, oddly, never mentioned Windows). They DID mention a great deal about the sinking of the Andrea Doria, which was broadsided by the freighter Stockholm and sank at a cost of 51 lives.

After some discussion of a theory that the Andrea Doria could have been saved with a better design for the fuel tanks and the ballast control system (largely discredited by the discovery that her keel had broken), they turn to the question of why the accident took place. It seems that the Stockholm thought she was twelve miles away, when she was in fact only two miles away. In changing course to avoid the Andrea Doria, the Stockholm steered directly into her.

Why did the Stockholm's crew think she was six times further away? Well, although radar was by that time already standard equipment on almost all merchant vessels, crews were not always familiar with how to use it properly. The phrase "radar induced collision" was already coming into use.

So what was the crew's error with the radar? At that time, it was not customary to light instruments and controls for use in the dark. Unless you carried a flashlight you would have no way to know how the controls were set, because the only indication would be a knob you could not see. So if you do not remember correctly whether the radar is in short-range mode or long-range mode, you may think you are twelve miles away from something only two miles away.

The story concludes with the comment that "although few experts believe that a different tank design might have saved the Andrea Doria, they almost all agree that a ten-cent light bulb would have."

I might add that some sort of range indication on the display would also have worked, though I do not know if that was technically possible at the time, and a light bulb certainly was.

Any thoughts?

A few. Unlike the Titanic, with its 1200-something lives lost, only 51 people died aboard the Andrea Doria and Stockholm that summer night in 1956. However, I knew two of those people; they were the parents of a good friend. They were killed instantly when the Stockholm plunged into the Andrea Doria's side and his life was changed forever.

People ignore human interface defects, no matter how serious the consequences may be. And when the consequences do occur, they blame whatever person happened to be at the controls at the time. Around fifteen years ago in San Francisco, and armored car driver drove off without locking the back door of the truck. The locals helped themselves to around $150,000 that fell off the truck. Another 10¢ lightbulb, tied to a switch on the back door could have prevented that one. Every single report I read on the event blamed the driver; no one blamed the idiots that didn't provide a door-ajar warning.

The almost meltdown at Three Mile Island was a result of "human error." At least those people "got it"; that event caused a major shift at the Atomic Energy Commission and the realization of the importance of human factors.

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