i bought a Samsung monitor about 6 months ago. Instruction manual showed a bunch of legends across the front bottom of the display. When I got home, I tried pushing on the front next to and on top of the legends. Nothing worked. The instruction book did not say how to turn on that I could find. I accidently grabbed the display along the bottom edge and it turned on. A little experimenting and I found that the touch sensitive switches were along the bottom side under the legends that were on the front. Nowhere that I could find did the instructions say to touch the display on the underside of the display to activate whatever function you were trying to activate.
This effect has been covered in detail in 'The Marching Morons' by C. M. Kornbluth. 60 years after it was published, the story is more pertinent than ever. It's well worth reading.
The fact is that when my Grandfather learned to drive there were a number of operations required to start a car. Put the transmission in neutral. Set the throttle to fast idle. Retard the spark. Apply the choke. Crank slowly until you hit the compression point, then give it a strong pull (DON'T push the crank with the thumb over the handle). Repeat if necessary. Failure to carry out these steps properly resulted in a flooded engine, the engine not starting, or a even a broken arm.
The electric starter, automatic choke, automatic transmission, and computer control module have eliminated the need for such care. Maintenence that has to be performed every 1000 miles is now performed at 10,000 to 100,000 mile intervals.
A similar thing has happened with computers over the last 45 years. In 1967 to run a program I had to enter the program steps on IBM cards using a keypunch. Any mistakes and the program wouldn't run. There were also idiocyncries of the language to be avoided. As the hardware evolved, programs came on a half dozen (or more) floppy disks which we had to load by hand. Configuring a printer could be a nightmare if the appropriate driver wasn't included. Now the operating system is preloaded, software (aps) is installed automagically, and any idiot can use one without understanding anything about how the whole system works
And I remember 20 years ago when I was performing telephone support for OS/2, and some very nice lady called because she couldn't get her new computer to print to her new printer. After some struggling, we discovered the salesman had failed to sell her a printer cable.....
One of the failures of many design engineers. I had similar problems decades ago with test engineers writing test procedures assuming an advanced knowledge of electronics. A previous lead tech would have his best tech vet the test procedures before accepting them. I added a further step - I had the newest tester try to follow the procedure. Any step he did not understand had to be rewritten. It irritated the test engineers, but ensured the procedure actually worked as written.
I've read of a similar test (possibly fictional) being applied to officer candidates in OCS. They were given a task which needed to be performed, and instructed to write up a set of orders to a platoon to carry out the task. If the grunts in the platoon could find a reasonably legitimate way to misinterpret the orders as-written, so that the task wasn't accomplished properly, the officer candidate flunked.
Seems like a good test for design engineers to have to follow... if Joe or Jane Average-Newbie cannot follow your instructions, they need to be rewritten.
--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
I think you may mean the failures of many industrial design, human factors, usability, and marketing. These days, the circuit design engineers have little say in the how a product looks or works.
Guilty as charged, but had a good excuse. Since I was the designer of the product, having me write the test procedures guaranteed that I would make some assumptions as to the operation of the product. When I wrote procedures for my own products, they were universally considered confusing and presumptive. Far too many missing steps. So, I made a deal with one of the other engineers and we traded products. It worked fairly well, and lower management accepted the idea. It also provided an education on other products, as well as generated many original ideas.
Then, upper management found out and announced its displeasure. I don't recall their exact logic but it had something to providing a single point of contact in engineering for the customers. When I pointed out that this also creates a single point of failure, the non-negotiable edict to cease the practice arrived in a memo.
I also had another problem. Some of the test techs were not the brightest on the planet. A few were able to follow even the most boring test procedure, repetitively, unerringly, and without protest. This was great if the procedure worked, but if there were any changes, the test techs would falter. They could tolerate boredom, but not change. So, I grabbed anyone NOT involved in the actual testing of the products, to help me debug the procedure. Just watching someone inexperienced fumble through the procedure was often sufficient to inspire massive reorganizations and rewrites.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
I've told this before but.. The TV tech went on a call for a TV that would not respond to the remote. Tech arrives to find an elderly couple, the gentleman was in his easy chair and proceeds to show the tech the remote doesn't work. Unfamiliar with the remote, he holds it pointing at the ceiling so he can see the buttons he needs to push. The tech told him to point it at the TV, the repair was complete. Mikek
I read it many years ago. Methinks the plot is unrealistic, the assumptions invalid, and predictions defective.
You might be reading a bit too much into this automobile repair story. None of the participants even remotely quality as being below average intelligence. Of the four (including myself), we have 2 electronic engineers, one mechanical engineer, and a machine shop owner. My question revolved around why 3 presumably intelligent people could not figure out how find and replace a fuse.
I used to drive a 1970 Land Rover Series IIa. It had a crank start. The first time I tried it, the crank launched me about a foot into the air and down hard on the front bumper. I was more successful with subsequent starts. After spending literally every Saturday under the car, fixing something different each time, I decided that something more reliable might be a good idea.
This is the good side of automotive maintenance progress. The bad side is that your grandfather could probably have fixed anything on his vehicle with baling wire and bubble gum. These days, it requires a computer program, and factory trained guru to decode the OBD2 codes and results. In addition, todays cars may not need scheduled maintenance for many thousands of miles, they will still break down before the scheduled maintenance. It's not the operating parts of the vehicle that seem to break. It's the myriad of sensors, black boxes, and convenience features that break first. Operation of the vehicle is also far from clear. My Subaru has a seat warmer switch, that took me about 2 hours to find and turn off in order not to get a hot seat. I still don't know where the rear window defroster is hidden, and the function of a mysterious dashboard lighting switch. (Yes, I read the manual).
Yep. I went through all that and suspect I todays computers are genuinely better. However, I'm not quite sure. In 1981, my first IBM PC 5150 took about 5 minutes to boot. 30 years later, it still takes
5 minutes (or more). This is not progress.
I do agree that the level of intelligence required to operate a computer has decreased to the point where most people can theoretically operate one. The various pad computers were a big step forward. However, I again have my doubts, having seen the mess these permanent beginners can make of a machine. Since there are also the largest potential growth market segment, I suspect we will soon seem computers that even the totally clueless can operate. Whether this is a good thing will remain unanswered.
Chuckle. I haven't experienced that one. However, that was fairly typical for the late 1980's and 1990's. Here's my list of horror stories: That begs another unanswered question. Are we making smarter users or just better computers that prevent users from hurting themselves? I suspect the latter.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
There is the biggest problem. I hope y'all don't mind my two cents here but people might be getting stupider. I have a set of beliefs which of course are not adhered to, but they should : Practically all safety devices should be illegal.
I was watching the boob tube a few years ago (can't stand it at all anymore ) and they had this dandy table saw. It had some sort of sensor that made i t stop the blade within microseconds should someone touch it.
Well there are millons of table saws out there without this feature and if you happen to buy one of these new dandy ones, and then for whatever reason have to use a different one, what might happen ? And how do they get someo ne to demonstrate the thing in the first place ? Do they look for salesmen wearing a Jefffoxworthyesque stupid sign ? (two bucks apiece, two for five)
Well in my opinion this should apply to everything. Instead of seatbelts an d airbags in cars they should have pointy, LETHAL things sticking out the d ashboard. On the other hand the doors should be triple reinforced, like wit h railroad tracks or something. net result : If you drive like an (_|_) and hit things, you might die. If you get Tboned, the person who hit you might die but you will live. Within a couple of years there would be no car acci dents. People would actually watch where they're going instead of having a contest to see how many other things they can do while driving.
It pisses me off, all this automatic shit. When I open the car door I want the courtesy light to go on, and I want it to go off immediately when I clo se the door. Period. I will lock and unlock the doors thank you. I know whe n to turn on the headlights. I know the seatbelt is there, and I swear one of these days I am going to start a business doing one thing - getting rid of those damn beepers that bug you to put on your seat belt. I'll probably make a fortune. One thing I will accept is a beep if the door opens and the keys are in the ignition. ONE BEEP, got it ? Can the engineering departmen t count to one ?
And brownwares ? This is getting ridiculous. You want to hear a wierd one a bout a TV ? OK, a Sony XBR direct view, CRT. The CRT had been changed. The set was back becasue the picture would gradually turn all red, that is the green and blue would drop out. The cure ? Adjust the vertical height and ce ntering.
This one required discovering the failure mode to cure, but in some cases y ou might not have to. In this unit, the CRT installed was rebuilt and appar ently there was something a little different about the envelope. I scoped t he AKB pulses and though there was no green or blue on the screen there was plenty of feedback going to the AKB pin of the jungle.
A slight convergence error OFF THE SCRREN at the top was causing the beam t o hit some element in the tube, most likely the frame around the shadow mas k, and it was getting more feedback from that for whatever reason. I suspec t in the CRTs made for the sets with AKB that frame or whatever it was had a coating on it to prevent this occurance.
Once I had to change a chip in a vertical circuit because the set would not switch to video input. That was easy, internal leakage was partially short ing out the data bus. The normal stuff worked, like the picture controls et c., but the option codes had not loaded so it didn't know about any video i nputs.
This is all automation working for us. If you think diagnostic codes are ba d in cars, they are worse in brownwares. Now that they want to detect short s by absence of supply voltages it mucks things up. All hail one shutdown c ircuit. Long time ago a coworker quipped that it's getting to where a bad s peaker can cause no high voltage. Well XBR to the front again. A CRT based RPTV, I detected no vertical as it shutdown. I repaired that and it still s hut down, even though I scoped it and confirmed that the vertical was now u p and running.
Above the chassis was the amp for the subwoofer which was down at the botto m of the cabinet. I had left it unplugged and the set sensed it and shut do wn. So there you have it for sure - a bad speaker can cause no high voltage . I just happened to temporarily introduce high resistance to the voice coi l at the time.
In the end, it is all about control. At one time I was going to build some things, actually advanced diagnostic devices. I thought about how to not on ly protect my ideas, but being on the other side of the fence I wanted capt ive service if possible. That can take some doing when the customers are te chs in the first place. After all none of this stuff is made to be serviced anymore. Look at the service manuals for plasma TVs and find the waveforms that are supposed to be output from the Y and Z sustain boards. They give you clk timing on the main board which you almost never fix but replace, th ey give you the flat lines on the 5 volts, 3.3 and so forth, but what reall y makes the thing work is absent.
Cars may be different but not much. You can't just drop an ECM in a car any more, it has to be programmed for it. One third of brownware problems are n ow due to software issues, many of them not fixable. Sometimes you can flas h the firmware but usually not. They need you to need them.
At this moment I am thinking about my next car. Most of what I want is expe nsive now, and it would probably have ignition points. Stickshift. Power NO THING.
Oh, and us hillbillies have advanced from the baling wire etc. The most mod ern methods now include duct tape and coat hangers. I notice it is getting harder to procure metal coat hangers, an fact I suspect a conspiracy. In fa ct the entire computer revolution was a conspiracy to eliminate paper clips . There was a time I could make almost $100 with a paper clip.
people might be getting stupider. I have a set of beliefs which of course are not adhered to, but they should : Practically all safety devices should be illegal.
and they had this dandy table saw. It had some sort of sensor that made it stop the blade within microseconds should someone touch it.
happen to buy one of these new dandy ones, and then for whatever reason have to use a different one, what might happen ? And how do they get someone to demonstrate the thing in the first place ? Do they look for salesmen wearing a Jefffoxworthyesque stupid sign ? (two bucks apiece, two for five)
airbags in cars they should have pointy, LETHAL things sticking out the dashboard. On the other hand the doors should be triple reinforced, like with railroad tracks or something. net result : If you drive like an (_|_) and hit things, you might die. If you get Tboned, the person who hit you might die but you will live. Within a couple of years there would be no car accidents. People would actually watch where they're going instead of having a contest to see how many other things they can do while driving.
courtesy light to go on, and I want it to go off immediately when I close the door. Period. I will lock and unlock the doors thank you. I know when to turn on the headlights. I know the seatbelt is there, and I swear one of these days I am going to start a business doing one thing - getting rid of those damn beepers that bug you to put on your seat belt. I'll probably make a fortune. One thing I will accept is a beep if the door opens and the keys are in the ignition. ONE BEEP, got it ? Can the engineering department count to one ?
a TV ? OK, a Sony XBR direct view, CRT. The CRT had been changed. The set was back becasue the picture would gradually turn all red, that is the green and blue would drop out. The cure ? Adjust the vertical height and centering.
might not have to. In this unit, the CRT installed was rebuilt and apparently there was something a little different about the envelope. I scoped the AKB pulses and though there was no green or blue on the screen there was plenty of feedback going to the AKB pin of the jungle.
hit some element in the tube, most likely the frame around the shadow mask, and it was getting more feedback from that for whatever reason. I suspect in the CRTs made for the sets with AKB that frame or whatever it was had a coating on it to prevent this occurance.
switch to video input. That was easy, internal leakage was partially shorting out the data bus. The normal stuff worked, like the picture controls etc., but the option codes had not loaded so it didn't know about any video inputs.
cars, they are worse in brownwares. Now that they want to detect shorts by absence of supply voltages it mucks things up. All hail one shutdown circuit. Long time ago a coworker quipped that it's getting to where a bad speaker can cause no high voltage. Well XBR to the front again. A CRT based RPTV, I detected no vertical as it shutdown. I repaired that and it still shut down, even though I scoped it and confirmed that the vertical was now up and running.
the cabinet. I had left it unplugged and the set sensed it and shut down. So there you have it for sure - a bad speaker can cause no high voltage. I just happened to temporarily introduce high resistance to the voice coil at the time.
things, actually advanced diagnostic devices. I thought about how to not only protect my ideas, but being on the other side of the fence I wanted captive service if possible. That can take some doing when the customers are techs in the first place. After all none of this stuff is made to be serviced anymore. Look at the service manuals for plasma TVs and find the waveforms that are supposed to be output from the Y and Z sustain boards. They give you clk timing on the main board which you almost never fix but replace, they give you the flat lines on the 5 volts, 3.3 and so forth, but what really makes the thing work is absent.
anymore, it has to be programmed for it. One third of brownware problems are now due to software issues, many of them not fixable. Sometimes you can flash the firmware but usually not. They need you to need them.
expensive now, and it would probably have ignition points. Stickshift. Power NOTHING.
methods now include duct tape and coat hangers. I notice it is getting harder to procure metal coat hangers, an fact I suspect a conspiracy. In fact the entire computer revolution was a conspiracy to eliminate paper clips. There was a time I could make almost $100 with a paper clip.
I have mixed feelings about the new technology used in automobiles. Overall, I would have to say profound improvements have been made. That does not mean some of the 'improvements' don't drive me nuts. I detest airbags. The idea of driving holding onto a device that will do it's best to smash my arms into my face in case of an accident is crazy. Protecting me in case I forgot to wear my seatbelt is unnecessary - I refuse to go anywhere without my seatbelt fastened.
Yes, some of the power features will go bad before the rest of the vehicle wears out. On the other hand, the power train is more reliable than ever.
I compare two vehicles we have owned - a 1969 Dodge station wagon, and a 2002 Honda Odyssey.
Throughout it's life, the Dodge had to have the oil changed every
1500 miles, spark plugs and points replaced every 15,000 miles, the muffler and tail pipe were both replaced twice, brake shoes had to be replaced, etc. In addition, it had a number of annoying failures. The heater control cable would buckle, diodes in the alternator failed, the heater blower motor failed (requiring pulling the entore heater assembly). And in spite of the best efforts, by teh time it had 100,000 miles it was ready for the junkyard.
The main maintenence item on the Odyssey is the oil change every 3000 miles. And the brake pads had to be replaced twice. No points, no new spark plugs, still has the original exhaust system. Of course, it had it's own set of annoying failures. The power window mechanism fails in a particularly annoying manner - a plastic retainer breaks, releasing the actuator cable. One of the remote 'fobs' failed. The engine mounts failed. The alernator mounting bracket broke!! On the other hand, I didn't have to replace the spark plugs until after
100,000 miles, it still has to original exhaust system, and at 150,000 miles it still is running well.
A YouTube'r I like to watch is mikeselectricstuff. He's a brilliant engineer and is always reviewing and taking stuff apart.
His reviews -- more often than not -- end with this conclusion: the hardware is excellent and the UI firmware/software/design is crap. He points out how potentially fabulous this bit of kit could be if the UI was halfway competent.
Here's a couple of good examples:
formatting link
IMO, user interface is designed by people who probably wouldn't use the product.
When I worked at Apple (tech writer, decades ago) we would do a draft of a manual and then get a prospective customer -- "target audience" -- (someone from HR or such in this example) and put them in a room with a new product, in box, and let them go to it, videotaping the experience. The feedback is what made Apple's documents receive awards on top of awards.
If companies (ANY manufacturer) would do this for their UI, most issues would be resolved before the product hit the shelf. But being "a race to the bottom", I don't hold much hope.
My favorite 'how not to' example is a dustpan, made with strengthening ridges on the bottom sides of the pan, in such a way that its impossible to put the front edge of the pan flat on the floor to pick up the sweepings - they go underneath the raised edge on the front lip created by the ridges. Obviously the designer ( and any subsequent reviewers ) never used a pan.
--
Regards,
Adrian Jansen adrianjansen at internode dot on dot net
Note reply address is invalid, convert address above to machine form.
About a decade ago, we did this, but it was a dismal failure.
I worked with the installation group of a fairly popular circuit simulation program, and someone had the bright idea to have the CEO test out the installation...
Most of my frustrations were unloaded in the original thread:
In a past life, I helped design several marine radios. The user interface was designed by a genuine industrial designer, with input from literally everyone that could possibly render an opinion. That included the janitor and random visitors. Part of my responsibility was to document all these great ideas for the industrial designer so that he could concentrate on the aesthetics and ergonomics. As a result, usability took third place behind feature bloat and artistic packaging. By the time the front panel was presented to engineering, it was obvious that we were in trouble. There weren't a sufficient number of buttons to handle all the features. So, each button had multiple functions, depending on whether it was tapped, held for 1 second, or held for more than 3 seconds. Considerable effort and wine went into crafting said interface, but the inevitable result was a miserable compromise. Marketing insisted that the radio with the most features would sell best, so there was no way to remove features or add buttons. We built a mockup and tried my usability test on various non-technical employees including the marketing manager. Nobody could operate it. There was even some difficulty in turning it on and off. Someone suggested that we have 3 modes (crude, basic and advanced) where some of the obscure functions were simply disabled. Everyone could operate it in crude mode, but the others required both RTFM and practice. We shipped a data logger with several radios, and found that 99% of the wiz-bang functions were never used.
Lesson learned: Features, functions and complexity sell products, but drive the users nuts.
Any product that is worthwhile should be intuitive and not require a manual. These daze, manuals are shipped on CD's and never read until something goes wrong. If the product needs a manual to use, there's something wrong. Besides, today's manuals are mostly legal disclaimers, court ordered warnings, and patent notices.
Lesson learned: Assume that users are NOT going to RTFM.
You're optimizing whale oil and sealing wax. Manuals and static user interfaces are dead, or at least should be dead. A proper user interface adjusts itself to what the user is doing, offering only those choices which are involved in whatever the user is trying to accomplish. A banking machine is a good example. You are presented with a minimum number of selections, all of which are appropriate to the current operation or mode. This results in more menus, but fewer choices.
I had the displeasure of proving the point when my father, the original permanent computer beginner, was trying to learn how to use his shiny new 1981(?) Altos AOS business computer system. I translated the menus from English to Polish, which was a big help, but he was still lost when presented with 20 menu choices per page. When I reduced it down to 5 choices, he was elated.
You can demonstrate how it works by asking someone to add a column of numbers, first with a 4 function calculator, and then with a scientific calculator. The 4 function is easy, but the scientific will usually cause a beginner to panic. Yet, they have the same numeric keypad and arithmetic buttons, usually in similar locations. The difference is the scientific calculator has far too much static clutter in the form of unused buttons, which add confusion.
Lesson learned: The purpose of a user interface is to REDUCE the number of choices, not bury the user in irrelevant over choice and complexity.
About 10 years ago, I designed a user interface for a radio. It was sufficiently unique that the company engineers and marketing people were afraid that users would not intuitively understand how to operate the radio. A few tests confirmed their suspicions. So, I now have to wait until some clueless competitor produces something similar, so my client company doesn't need to take any risks being innovative. If you're wondering why you're seeing the same broken user interfaces repeated ad infinitum, this is why.
Lesson learned: Innovative user interfaces are risky.
OK. End of yet another rant...
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
The best measure of the quality of a product is employee theft. It's not worth stealing something that they know doesn't work or will break.
You may wish to browse the "Made by Monkeys" column found in Design News magazine: Every month, they feature one or two poorly designed or poorly built products. I don't recall reading about your dust pan problem, but there are some that are similar.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
*MY* favorite (actually, most hated...) "how not to" is the spare tire carrier on domestic (Ford, GMC, etc.) full-size trucks. My folks had a car repair and tire place. I've cursed out loud in front of my mother while trying to remove/replace a tire under a truck. A (slightly) better design is the chain hoist system used on some smaller imports. I hope there's a special place in hell for the Mercedes-driving engineer that designed this thing (you know HE never changed a flat in his life!).
Speaking of radios and user interfaces...I had a Kenwood receiver in one of my cars. Great radio, great sound. The one problem was that they used the volume "knob" as a joystick to get multiple functions out of it. When you pressed it in, it switched to the sub-woofer menu. There was no easy way to exit the menu. If you were sitting still, this would seldom be a problem, but in a moving, bouncing, vibrating car, it's quite likely that you will push the knob in while trying to adjust the volume. Then you have to try to manipulate your way through the menus without taking your eyes off of the road for more than a split second. I finally started popping the faceplate off and back on, whenever I got into the menu system accidentally. This turned the receiver off, and let me resume normal operation. Kenwood stopped using this interface with this model.
My former 1983 Dodge D50 pickup had a chain hoist spare tire under the bed. Great idea until I blew a rear tire and high centered the rear axle while driving on a dirt road on the way to a mountain top radio site. In order to lower the tire, a long hand crank was provided. The problem was that I was backed up against a hillside, and could not get the long crank into the hole. I had to dig out part of the hillside for it to fit. In order to remove the spare tire, I had to jack up the pickup bed about 3 ft off the ground, and crawl under the raised bed to release the toggle link holding the tire to the chain. Of course, with the tire lowered, the toggle link is UNDER the tire on the ground. I raised it with a bottle jack and a rather unstable pile of rocks. While replacing the blown tire, the pile of rocks and jack did partially collapse. Perhaps in your parents tire store, it might work, but on a dirt road, it's not easy.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
I carried some scarp pieces of 2x12 in my '79 Dodge pickup. If I had a flat, I would stack them against the flat tire and pull the truck up on top of them to be able to get the factory jack under the tire. Without the extra height, you couldn't get the jack to the jacking points on the frame.
--
Politicians should only get paid if the budget is balanced, and there is
enough left over to pay them.
Sometimes Friday is just the fifth Monday of the week. :(
Rant noted and seconded. I've seen aftermarket auto radios that required you to hold-two-buttons to turn the power off. Now what the hell were they thinking? And this wasn't some cheap unknown brand, it was made by Pioneer. Don't the designers know that complexity in a vehicle environment can lead to dangerously distracted driving? Or do they give a damn? Tom
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.