I was struck, the other day, by how much I had to think when attempting to heat up my sandwich in the microwave. There are so many controls: a clock, a set of food-specific buttons, defrost and timer controls, and of course a full numeric keypad. Quick! What do you press?
Software interaction designers may be amused / interested / inspired / instructed by:
http://www.noideasbutinthings.com/
I wonder if that’s where Jeff found one of today’s photos:
http://www.noideasbutinthings.com/ixd/2007/10/sanyo_maxichef_knob.html
Microwaves, like all technology, have been overcomplicated by the weakest link. Sure, the dial worked fine, but what about people who don’t know how long to pop popcorn? We don’t want to make them have to learn, so we better build a microwave with a ‘popcorn’ button and write algorithms that take the size of the bag into account and produce the correct cooking time. A classic example of technology being over-engineered to suit the absolute laziest of users. What would happen if people had to figure out on their own how to make popcorn? Less people would eat popcorn? Unlikely. Less people would buy microwaves? Even more unlikely.
The same thing happens to me when building information systems, people complain that they need a feature to control some minute business exception that occurs in 1 out of every 1,000 iterations of a process, but do they really need that? No, there are other, more simple ways, like using the provided comment field to mark the exception. However, the user would then actually have to READ the comments (of which there are maybe 2-3), and well that’s just way to much effort.
To make things worse I have a webmaster here who bends right over to adding new buttons and features in ever spare inch of control realty to accommodate such asinine requests. So many things could just as easily be fixed with a minor change to a business process, or a personal education process in the case of microwaves. However, people want technology that does absolutely everything for them, and rarely do they consider the actual complexity of such technology.
They point here is that they put those buttons there because they want to sell more waveovens.
And plain people when want to buy a waveoven and don’t know anything about waveovens doesn’t have a criterion to choose what is ‘best’ for them.
So they usually look for the one that looks more complete, capable of doing complex stuff and, of course, the one that looks more expensive to impress the people that comes to home.
I think that the last sentence is the one that sells more waveovens.
I wonder if that’s where Jeff found one of today’s photos:
Good eye. Indeed it is-- that, plus my perplexing sandwich experience from yesterday’s lunch.
Easy - the popcorn button. 
I just discussed this with a coworker, and ironically when he tried the “popcorn” button on our microwave (it’s the one in the photo), it burned his popcorn. Irony of ironies.
Basically, a UI expert went to Korea
Hi Giles-- I liked that too. I wrote a little bit about that here, with citations to the original posts: http://www.codinghorror.com/blog/archives/000880.html
Has anyone used the “Frozen Pizza” button? I never got that…are you supposed to stick a mini frozen pizza in there or something?
There is a trade off with simpleness in UI and discoverability of features.
If your UI is too simple, people will never know of what your product can do. Then there is the critical mass that Jeff is pointing out. Where so many buttons kill discoverability because its so cluttered.
Code Rush (a productivity tool for .net developers) is a tool that nails simplicity and discoverability. There is a help (training) window that you can keep open while you write code. Instead of one long list of key commands, it dynamically shows only relevant keystrokes given the context of where you are in source code. There is also a skill setting that you can change as you learn shortcuts and key commands. It filters the complexed ones from basic users and basic ones from expert users.
Obviously a dynamic set of buttons on a microwave is fantasy, but i think the point was to draw parallels to software.
I hate to say this but on most consumer devices, the knob and dial interface was much more intuitive, useful and controllable than the flat panel digital crud created by befuddled marketing types and clueless engineers.
I can’t help but notice that when controls are expensive to produce (like knobs and dials), their number is quickly minimized to the most efficient possible arrangement. It forces the engineer to think about what they are doing. Something many managers seem to perceive as “nonproductive.”
As a personal aside, I believe that the designers of television and DVD remotes should be sent to a Hell where they are forced to use them forever.
BTW.
Any true popcorn aficionado knows that you never rely on the popcorn button when using a foreign microwave. You always have to listen for the slow down in the popping interval.
Good post. Just having a rotary control instead of a keypad is not good enough. Seeing the photograph of the old microwave made me realize why the control on my microwave does not work well. I have a Panasonic with a knob (with a digital display) that lets you set the time. I bought it thinking that was a very good idea. But it fails miserably. With the old microwave in the photo you can see how far you must turn ‘before’ you start to turn it. With my microwave I have to watch the changing display to figure out when to stop and the dial itself offers no resistance (or feedback) while turning it. I can never set it right quickly.
The presets are too much. Button overload.
The auto start by pressing a single time button and then adding additional time with another one button press is great. Much better than pressing Start, Time, Go. But the knob was better than this too.
Perhaps a better title for this post would be: “Popcorn Button Considered Harmful”.
What’d work better is a simple alarm-clock interface:
H+10
H-10
H+1
H-1
M+10
M-10
M+1
M-1
These would be expressed, of course, as a series of #9652;/#9662; buttons.
And, of course, the “Start” button.
Simplicity at the cost of function? No thanks!
I agree with you that microwaves are more complicated than they need to be, but I think your single-knob version is trying to go too far in the other direction.
How do you know if you’re nuking something for 1 minute or 55 seconds? The analog control doesn’t give you enough control.
Granted, maybe that doesn’t matter… But what if you need to nuke something for, say, 5 seconds? It’s not exactly easy to set that for 5 seconds.
Plus, there’s no way to change the power settings.
And, finally, the knob in the picture is limited to 10 minutes…
One dish I enjoy, pre-packaged Rice Pilaf, calls for 20 minutes at 50% power… Trust me, 10 minutes at 100% power isn’t the same.
Thanks for this post! I thought it was just me…
I’ve got no problem programming computers. But microwaves…
Why do you think that the intent of the extra buttons was for a better, simpler user interface? Sure most people are going to use the functions (and most buyers know they won’t either). But they have to “know” that they could use those features - so they have to be displayed, after all, they aren’t going to read the manual.
When someone else sees their microwave, do they want to be showing off the plain one (even if it is the easier to use) or that they could afford to get the one with all of the bells and whistles?
I suppose I should clarify that in the case of a microwave oven, they should be M and S buttons rather than H and M buttons. I’m pretty sure that anything will turn to a cinder if you nuke it for one or more hours. #9786;
Clock, Power, Time and a Start/Stop. Plus a up and down to adjust the time and power. Only 6 buttons. That would make a very simple interface.
I don’t like knob in this case because it is more difficult to clean.
But analog knob are awesome in home theater. I remeber my old VCR with fast foward and rewind knob. I miss it so much! Now I have foward and next chapter that are almost indentical and are side by side. It is really not good desing.
To be fair to the designer, they obviously went for accuracy over simplicity. Ever tried to get a 20-minute knob microwave to heat for 1:45 minutes?