On not reinventing the wheel and consistency, this is where all "Electron Desktop" apps are at now, UI tends to be "fancy" and alien to the host operating system.
The best microwave UI has just two dials. Temperature and time. For years my microwave experience involved mostly just twisting the latter and walking away. I've gone out of my way to find microwaves with this control set. Unfortunately the most recent one I bought has buttons and settings, and I hate it. I want the dials back.
An advanced one has extra options that are functionally valid, like blowing more air as opposed to wave radiation/convection heat, or adding more moisture surrounding it to keep that chicken moist, adding some top grill heat to give more caramelisation on the skin etc.
These things requires a brain puzzle if you only use 2 buttons (think oldskool micro devices or configuring digital watches), and adding more buttons for these options offers arguably a more direct and simple user experience if you rtfm.
He really should've told near the end. Anyway, I assume you press the power button, because that is the established standard for microwaves in the last 10 years or so. Probably there might be a timeout and then it starts automatically.
This is indeed something you have to learn though, first time I encountered this kind of microwave it baffled me as well.
Where the article seems to go wrong is with the very idea that this is a microwave. It's not. It's a combination of microwave and grill hence the hinging of the door not matching a typical microwave.
I have a microwave from the Etna brand that has a big flaw in the UI. I believe the software is reused in many microwaves because the cheap ass microwave from the Lidl my parents have has the same interface.
The flaw: In the first step you need to select the amount of heating. Not in absolute numbers but in the percentage of total power (100, 80, 50, etc). So every time you need to calculate how much power each percentage represents.
The article doesn't really go into what the perfect UI would look like. It is more of a buzz feed type article.
That being said, my microwave just broke down and I headed to the store to buy a new one. I was really disappointed at the new models. The UI seems to be getting worse on them. I have a feeling they are not even trying to innovate because it has become a low margin device.
While the StackExchange link is awesome, the rest are just horrible. As with all these wannabe designers they go for the wow effect. It takes about half a second of rational thought to see why.
All these art mockups are just that: art mockups. "How do I make a thing look pretty". A tired mantra of programming smays: make it work, make it fast, and only then make it pretty.
Touch screens in the kitchen. In the effing kitchen! When you are often up to your eyebrows in grease, oil, flour, vegetable juices etc.
Oh, look, you have to turn the knob to set the time. To set the power you have to touch some other area and then turn the nob again. To set the time, you have to touch that area again to switch modes, and then set the time. BUT WHY? This is solved with just having two knobs! https://johnlewis.scene7.com/is/image/JohnLewis/233182566?$p...
Oh, look. As it starts, it shows you time remaining, and a stop button. Because you never ever in your entire life would want to, say, adjust time. Nope? The thought never even occured to the designer.
Oh, look, there are still multiple small buttons (that are already awkward on a vertical pane), but let's make them touch buttons! Oh, and let's leave them exactly the same as they are now, because the only reason why people don't understand how they work is because they are not touch buttons. Make them touch buttons! They will immediately become intuitive!
Oh, look, the iOS time scroller! Because that's exactly what I need: to be able to set 23h 29m 31s by scrolling through infinite lists of numbers. I wonder why is it that the vast majority of microwave uses are around 30s to 5min range on max power?
And in the car, where you should keep your eyes on the road. Knobs work really well for adjusting the stereo and heater; replacing them with a huge touch screen is a terrible idea, and trying to make it less terrible with haptic feedback is a poor solution. Hopefully the current touch screen fad will pass.
All of those examples show large, high DPI (really thin fonts) colour touch screens, in non standard sizes.
These are completely impractical from a cost perspective, requiring expensive tooling (>$100K) and easily $50 per screen (e.g. iPhone screen https://technology.ihs.com/595738/), plus the massive firmware development cost increase from a small 8 bit micro to some giant ARM core, probably with a GPU as well.
You wouldn't design a phone app using the non standard widget toolkit unless you had a good reason, why do designers dabbling in appliance design not use the standard toolkit of physical switches and knobs to start with and see what they can come up with?
In my experience almost every microwave oven has terrible usability. It’s as if the design problem is so simple and the “correct” design (two independent, labeled, mechanical dials for time and power) is so obvious that everyone has to make their own version of it to look cool or sophisticated or something. I guess they never user test these things, because that doesn’t help sell more of them.
There really ought to be a usability certification for products like these, a verification by an independent organization that a typical user is able to operate the most important functionality with few or no errors.
The button I use the most on my Microwave is "add 30 seconds", I use it for almost everything, hitting it repeatedly to get more than 30 seconds.
I rarely want more than a few minutes of heating time, and I rarely need more than 30 seconds of granularity so that one button handles most of my microwaving needs.
Though I agree that dials would be better -- I wish cars weren't moving toward touchscreens, in my old car I could adjust pretty much everything in the car - radio, climate control, etc without looking. In my current car, there's a touch screen for pretty much everything except radio volume which is on the steering wheel.
Yeah, I have a gazillion buttons on my microwave, but the only button I really use is the Start/add 30 sec button.
If I want to adjust power I can adjust from 150 to 600, and then there's a "Jet" function, but I have no idea what that is. More than 600? 750 maybe? Who knows.
Dial is a bad design for time, because most of the time people want to heat something up for a short time: 15, 30 or 60 seconds. On many microwaves this involves slightly turning a dial and then counting the required number of seconds in your head (because the exact angle is too small to select right away). A log scale dial would solve this problem, but then it makes adding another 30 seconds a non-trivial operation, because the angle is no longer constant.
Most microwaves I've seen now seem to have the same (annoying) interface/software where you have to first select the power (always 100%), press start (it doesn't start), then select then time and press start (actually starts then).
What ever happened to the microwaves where you could key in a time and just press start?
Pressing any of those would auto start immediately with the corresponding time. If you want an exact time like 34 seconds, hit the cook time button and now the ten buttons act as digits. Then hit start.
Pressing any of these while the microwave is running would add that corresponding amout of time.
I guess you'd also need a power level button and a clock set button. There, all done.
It works, and it's fast to use. But it's for commercial/professional usage, where doing a set of standard common tasks with fewer number of actions is best.
Having home/office-use microwaves work like this is basically akin to forcing an average office user expecting something like Microsoft Word, to write Markdown in Vim instead...
> Having home/office-use microwaves work like this is basically akin to forcing an average office user expecting something like Microsoft Word, to write Markdown in Vim instead...
If that average office user is going to write something (or use a microwave) more than a few times in their lives, it's worth to force them, because the (very small) cost of additional learning is almost immediately offset by gained efficiency.
Alas, what matters is short-time thinking, so UIs are designed for sellability instead of helping people do their tasks correctly and efficiently.
You're ignoring the "cognitive load" and "attention requirements" of the mythical "correct and efficient" interface. Sometimes I want to use 99.999% of my brain for something else and I want to mindlessly use that microwave with the 0.0001% left. My desired thought process is "woo, woo, monkey-wuga, up-turny-power, down-turny-time, woo, woop, warmy my banana woop, woop!" or something simpler than that. And I also don't use it in a regular/predictable pattern, so auto-pressing one button for one of the "regular" time intervals might be useless for me since what I do is always irregular and sometimes irrational ("hmm, I wanna see some fireworks to light up my mood, let's microwave my colleague's USB stick labeled "bitcoins" for the lulz!").
The problem is not designing for sellability. Is not acknowledging that (1) some people don't want to RTFM! and (2) other people are OK with RTFM but want serious efficiency and correctness gains for it, "I pain, now you gimme gain!".
So you figure out who your users are and design for your users.
Now, the absolute worst is the parent article's example, where someone managed to "design for no possible user" by both forcing people to RTFM and providing zero efficiency and correctness benefits in return. But designing for the wrong user is also bad.
...because you've just describes one of the most horribly unintuitive UI/Xs one could imagine. Some of the unforgivable sins it commits are:
- One button can have more than one meaning depending on context. Aka "modal interface" - this sucks in all cases except 2: (1) touch screens (where the buttons / labels displayed can actually change on screen) and (2) Vim
- You have 10 Buttons instead of one simple Knob: on top of making UI elements do different stuff depending on previous action, you also have a ~5x excess of UI elements. Instead of 2 Knobs, one for time and one for power. Or one know, and a buttons above it to select time or power (I actually had to use a microwave with this setup btw, and hated that pos, I was always in a terrible hurry and made mistakes selecting wither the power or time and ended up with cold or burn food... but I'd still pick this over the 10-keys + extra buttons)
Seriously, don't ever design a UI/X for other people, if you ever need this done hire someone else! Also, if you're and UI/X-designer, please consider changing careers, seriously (I'm sure you're not, but just to be safe :P).
I remember a Linuxconf AU presentation on building a better microwave - the idea being to use a low-res automotive thermal sensor array to watch the turntable, and be able to set it to "heat this to 100 degrees" and have it actually do exactly that.
There were other things you got "for free" after that - like defrost-without-cooking by ensuring all pixels remained below a target temp, but got above some other temp.
The problem is that this forces you to make a confusing decision right at the start. You almost never know what temperature to heat something to -- you just want to put it in the oven for 30 seconds, see if it's good enough, and then repeat.
Typical UX design by an engineer as it is, with wrong assumptions about user expectations. The users of microwaves do not want to heat something to 100 degrees. They either want warm food or they want to cook. Warming means, that not only surface temperature measured by sensors has to be comfortable, but the temperature of the whole meal, which depends on thermal conductivity and penetration depth of microwaves for specific food and cannot be controlled by a sensor array alone. People will just hit "start" again, if they feel it's not ready.
For cooking people will either use recipe from a book provided with microwave and thus they need just a selection of a program, or they will use more advanced microwave with oven and will want to set temperature of the oven to something like 200 degrees for 40 minutes. In this case sensor would be useful, but it's already not a microwave story.
Isn't this just you projecting the limitations of current technology on an effort to improve it though?
Users of a microwave currently expect extremely poor temperature control, and so don't even think about temperature because you get what you get. An interface which lets me heat something to exactly "drinkable but hot" is a revolutionary change in how I use the appliance though.
Same story with warming food: sure, you want it warm all the way through - if you actually know the surface temperature though, then you can model that - food that is cold inside is going to have different surface temperature decay rates to food which is warm inside, and some modeling of the observed thermodynamics means the appliance can know this.
So now we'd really be talking: "warm to safe temperature all the way through" and we can give a probability / % complete read out of how close we think we are to it. No more guessing, no more user interaction, also no burning yourself because some parts went superheated.
Theoretically speaking, you are right, and at some level of technology it might be possible to implement this. However, "drinkable but hot" setting is quite far from "surface temperature at 100 degrees", right?
It's not "per gram", it's an dual function dial, used for both time and weight. If the lower dial is switched to defrost mode (2nd position) then the upper dial is switched to weight input. There is no scale inside.
The label means "time or grams", where "grams" only applies for defrosting.
Defrosting x grams typically takes y seconds, which is just a table lookup.
As defrosting most typically involves leftovers (rather than pre-packaged meals or ingredients), you will probably only know their weight, not how long it will take to defrost off the top of your head.
I think a small improvement to the design would be to use words (COOK TIME / DEFROST GRAMS), with the "DEFROST GRAMS" text colored to match the defrost option on the bottom (which should also use the word, "DEFROST"). This would let you figure out what the dial does without having to think as much about it.
You could also change the screen backlight color in DEFROST mode to reinforce the mode change.
> As defrosting most typically involves leftovers (rather than pre-packaged meals or ingredients), you will probably only know their weight, not how long it will take to defrost off the top of your head.
How would one know that? Do people routinely put leftovers on a scale before putting them into a microwave?
This is almost a good design, but there's a potentially dangerous flaw. Although it's not obvious in the picture, the dials have skeuomorphic fake machining marks. These marks can make reflected light look like lines radiating from the center. And the indicator mark is only a shallow depression, which is low contrast under some lighting conditions. The reflections on the fake machining marks can camouflage the indicator mark and make it hard to see. If you mostly use the microwave on full power, that's what you'll expect to see, so it's easy to miss it if the power setting is actually lower. This could result in food being heated less than you expect. It might be warmed instead of cooked, resulting in fast bacteria growth.
This can easily be fixed by coloring the indicator mark with a black permanent marker.
oh c'mon... there's an LCD/OLED display showing the time clearly. This is so much better than ANY alternative that any critique is nitpicking. (Though, to be honest, I have a very old white with black marks microwave that avoids the problems you mention... but it also had an ugly design that doesn't fit most modern kitchens :P)
Translated to web, you're basically saying "all websites should use a high contrast serif font at >16pt. to avoid people committing dangerous mistakes when reading instructions online" :)
The display shows only the time. The power setting is shown by the indentation on the lower dial. This can be hard to see under some lighting conditions, specifically because it's designed to trick people into thinking it's made of metal.
The analogy is more like web pages using scanned images of hand writing, which I'd also say is bad design.
Agreed. The newer Samsung is an inferior cost-reduced version of this design. The old version is better because the time can be set purely by muscle memory. The new version uses a rotary encoder, where you need multiple turns to set a time. The required back-and-forth movement of the hand is too complex for muscle memory. The rotary encoder skips steps if you move it too fast. The time dial does not move as it counts down, so you're forced to read the display to check the remaining time instead of just glancing at the dial.
The only real improvement is the replacement of the door opening button with a more intuitive door handle.
> The only real improvement is the replacement of the door opening button with a more intuitive door handle.
Actually, it's not an improvement. A door opening button is easy to operate with one hand; a handle needs two in order not to shift the microwave. At least all the microwaves I've seen have doors requiring noticeable amount of force to open. My guess is it's to avoid the situation in which a large dish rotating inside pushes on the doors and opens it.
The last time I bought a Microwave (about four years ago) I remember specifically looking for two knobs just like this. I remember the first microwave I used in the nineties and it worked similarly - just like an oven (also two knobs normally) everyone knew how to operate it almost instantly.
The one I eventually found was also the cheapest one and almost the only one. The more expensive models were usually designed around a button interface and had plenty of features like thawing, grill, and so on. Of course all these extra features amount to little more than switching the magnetron on and off according to a programme. Go figure...
My mom (who has since taken over it) was very confused about the one second delay between adjusting the time and the microwave starting (just like the linked one there was no start button). She had obviously been accustomed to pressing buttons to get it to operate how she wanted. Which just maybe goes to prove that what is "simple design" is not necessarily the simplest interface but the interface most aligned to the users' previous expectations on how this particular item should work.
And how do you choose if you want the integrated heating coil (after all, it also doubles as oven) to run?
And at what temperature?
And if you want to cook in stages?
We have at home a programmable one from the 1980s. We constantly use it for stuff like "unfreeze 2 hours, then microwave 1 minute at 800W, then heat at 230C for 20 minutes, then keep warm at 40C"
This is not bad, but has one problem. How many watts I have to use to heat up my dinner? Watts are not a natural thing that any human knows how to relate to any experience. Of course, with some trial or guidance, I could figure it out. But even then watts are useless for me - I just know 600w means "dinner", but it could as well say "37 zzq" and be as useful as "600 w".
I would put up with a hieroglyphic based UI ten times weirder than the one depicted if it meant the microwave didn't beep.
Seriously, what is it with embedded digital electronics designers and beeping? Nothing else has a gratuitous noisemaker. I swear if the home refrigerator was invented after the cheap piezo buzzer it would emit an ear-piercing bleep every time the door opened, or maybe when the compressor cycled on or off.
Is there some sort of strange subsidy or tax consequence so that adding a beeper has negative net cost?
I feel you. My microwave oven makes 4 lousy beeps once it's done... And will be more than happy to repeat those 4 beeps every 30 seconds until you open the damn door
I'm sure that without that feedback mistakes can happen that might be blamed on UI (didn't press the start/reheat/&c. button all the way through). Similar thought process to most touchscreen keyboards which have feedback (sound or haptic) by default because it's probably going to be a hurdle for users. Reminds me of the first time I installed a Linux distro which didn't disable the system beep (Debian w/o X)… surprising how this vastly and immediately changes UX. It's actually an interesting way to practice making fewer mistakes but it's a good thing that no UI relies on system beep any more.
A big ol' beep/nobeep switch on ovens would be a nice thing.
That Microwave UX is certainly bad. But the design of traditional microwaves is also not defensible. If you had never seen microwave before and tried to use it, you would be quite dumbfounded. Just pressing 1 key starts heating for 1 minute but how do you do 30 seconds? Just press the start key! That's quite unintutive if you ask me. The reason people don't feel that way because most people have learned how to use microwave from their parents or friends and knowledge keeps getting passed on. On those lines, I know of no one who knows how to use other "advanced" features that most microwaves have today like thawing and food type specific heating.
I think the key thing that makes my microwave easier to use is that it has a big dial to choose the function. No clicking through menus, you just turn the dial to the function you want, then the second dial to adjust the settings, or press start to start and add time.
I don't understand the "You are Not the User" mantra. In the past I've worked in teams where we were actively using the product we were developing. Dogfooding is a pretty established practice these days. For most purposes, I was The User - the user finding bugs, performance problems, etc. - but not user interface problems. For the UI I was Not The User. The Users for UI were these hypothetical beings only UI experts could came up with.
Developers actively using their product always have expert bias and may not represent the target audience of the product. Also, in most cases it's a wrong assumption that you use something exactly in the same ways as others, so decisions based on developer or designer own experience are usually not optimal and may actually harm the users that would purchase the product or service.
Part of the "expert bias" is that you want the tool let you do the job as efficiently as possible, which is also something the users want. It's, however, not what the marketing department wants.
There are two wrong assumptions you are making here.
First of all, "Doing the job as efficiently as possible" is not a user's desire. Users always want concrete things and developer's opinion on what that could be or how that will be achieved by a particular user is very often quite far from reality.
Second, it's not a developer's job to create a tool that will help users to solve some problem in the way he thinks it is right. It is not what developer is paid for. His job is to work together with marketing department to deliver a product that is always a compromise between the business and the user, a product, that is just good enough, in a deal, which is good only then, when noone is happy with it, but everyone can accept it.
> a product, that is just good enough, in a deal, which is good only then, when noone is happy with it, but everyone can accept it.
I refuse to accept that is right, even though it is what seems to be happening everywhere. That's probably the root of my opinion. I'm a believer in the concept of making money through providing value to the customer.
I think you have all valid points, but I think one should take a step backwards.
If you look at a plane's cockpit controls, not being a pilot one has absolutely 100% zero clue to what each button does.
They should place a monitor screen, and every time u press a button it should do a pseudo unix man command and output all the options on screen.
We don't think this is bad UI, because we expect someone (a pilot in this case) to have implicit knowledge beforehand .
If the picture in the article is the actual oven, well.. Like a piano not having each actual note name written on it, aesthetics often minimize straightforward information.
While ambiguous perhaps, I think this oven is meant for someone who is desigining his own Swedish getaway and is meant to be minimalist to fit in with the rest off the kitchen design. 99% off the time when buying appliances they come with a manual in a dozen languages. This is because one buys it for consumer usage, which as an office (random) user you can't expect guaranteed access to the manual.
What if this manual has 8 pseudo chapters, 1 to describe each button. With some graphical examples of press sequences of common dishes. It might be that simple (and I can tell you that manuals do come simple like this owning a hybrid myself).
Its 10 serious minutes in the manual for 10 years not-even-think-what-you-press experience.
UI designer here. This guy's advice is OK-ish as far as it goes. However, he fails to exhibit the one thing that sets an actual UI designer apart from somebody with just some opinions about the practice: observational research.
The FIRST thing you do is observe. Watch people in relation to the problem you are trying to solve. Listen to them. Ask them WHY they do what you observe. If at all possible, work out what motivates them and what problems they really have (which they often will not articulate to you). Only then start designing solutions.
I'm not saying his ideas about the design of the oven are invalid. But the fact that so few people in design roles even attempt the work of exploratory research is fundamentally the reason why so much of stuff is shit.
This isn't surprising though. Observational research is often frustrating, confusing or difficult to interpret. But above all it takes a lot of time. This is also the reason why, if you have another job like coding or marketing or whatever then you will not have enough hours in the day to do effective design.
79 comments
[ 3.1 ms ] story [ 145 ms ] threadAn advanced one has extra options that are functionally valid, like blowing more air as opposed to wave radiation/convection heat, or adding more moisture surrounding it to keep that chicken moist, adding some top grill heat to give more caramelisation on the skin etc.
These things requires a brain puzzle if you only use 2 buttons (think oldskool micro devices or configuring digital watches), and adding more buttons for these options offers arguably a more direct and simple user experience if you rtfm.
This is indeed something you have to learn though, first time I encountered this kind of microwave it baffled me as well.
Sometimes I think users are getting too spoiled and simply refuse to RTFM.
Not to mention the article complains about the UI for an oven, confuses the oven with a microwave and then gives advice for designing iphone apps.
The flaw: In the first step you need to select the amount of heating. Not in absolute numbers but in the percentage of total power (100, 80, 50, etc). So every time you need to calculate how much power each percentage represents.
That being said, my microwave just broke down and I headed to the store to buy a new one. I was really disappointed at the new models. The UI seems to be getting worse on them. I have a feeling they are not even trying to innovate because it has become a low margin device.
People are still trying if you are interested.
Articles:
https://ux.stackexchange.com/questions/28006/why-do-microwav...
https://medium.com/@bn/this-is-how-a-microwave-should-work-4...
Art Mockups:
https://www.behance.net/gallery/15102955/Microwave-Redesign-...
https://dribbble.com/shots/1309476-Microwave-UX-Concept
https://dribbble.com/shots/1449210-Microwave-Concept-PSD
https://dribbble.com/shots/3249355-Smart-microwave-UI-concep...
If you have others, feel free to comment! Appliance UX/UI is a very neglected area.
All these art mockups are just that: art mockups. "How do I make a thing look pretty". A tired mantra of programming smays: make it work, make it fast, and only then make it pretty.
Touch screens in the kitchen. In the effing kitchen! When you are often up to your eyebrows in grease, oil, flour, vegetable juices etc.
Oh, look, you have to turn the knob to set the time. To set the power you have to touch some other area and then turn the nob again. To set the time, you have to touch that area again to switch modes, and then set the time. BUT WHY? This is solved with just having two knobs! https://johnlewis.scene7.com/is/image/JohnLewis/233182566?$p...
Oh, look. As it starts, it shows you time remaining, and a stop button. Because you never ever in your entire life would want to, say, adjust time. Nope? The thought never even occured to the designer.
Oh, look, there are still multiple small buttons (that are already awkward on a vertical pane), but let's make them touch buttons! Oh, and let's leave them exactly the same as they are now, because the only reason why people don't understand how they work is because they are not touch buttons. Make them touch buttons! They will immediately become intuitive!
Oh, look, the iOS time scroller! Because that's exactly what I need: to be able to set 23h 29m 31s by scrolling through infinite lists of numbers. I wonder why is it that the vast majority of microwave uses are around 30s to 5min range on max power?
I could go on
And in the car, where you should keep your eyes on the road. Knobs work really well for adjusting the stereo and heater; replacing them with a huge touch screen is a terrible idea, and trying to make it less terrible with haptic feedback is a poor solution. Hopefully the current touch screen fad will pass.
These are completely impractical from a cost perspective, requiring expensive tooling (>$100K) and easily $50 per screen (e.g. iPhone screen https://technology.ihs.com/595738/), plus the massive firmware development cost increase from a small 8 bit micro to some giant ARM core, probably with a GPU as well.
You wouldn't design a phone app using the non standard widget toolkit unless you had a good reason, why do designers dabbling in appliance design not use the standard toolkit of physical switches and knobs to start with and see what they can come up with?
A big frustration is not being able to turn off the damn reminder tone though so I have to wait around to clear it when done.
I prefer 0-9 buttons, and lots of memories, mine has 3.
If you wanted a minimalist microwave though, you could have it reset each time the door is opened or after a minute or two of inactivity.
There really ought to be a usability certification for products like these, a verification by an independent organization that a typical user is able to operate the most important functionality with few or no errors.
I rarely want more than a few minutes of heating time, and I rarely need more than 30 seconds of granularity so that one button handles most of my microwaving needs.
Though I agree that dials would be better -- I wish cars weren't moving toward touchscreens, in my old car I could adjust pretty much everything in the car - radio, climate control, etc without looking. In my current car, there's a touch screen for pretty much everything except radio volume which is on the steering wheel.
If I want to adjust power I can adjust from 150 to 600, and then there's a "Jet" function, but I have no idea what that is. More than 600? 750 maybe? Who knows.
What ever happened to the microwaves where you could key in a time and just press start?
Ten number buttons, 0-9, but each also corresponds to a useful time.
Pressing any of those would auto start immediately with the corresponding time. If you want an exact time like 34 seconds, hit the cook time button and now the ten buttons act as digits. Then hit start.Pressing any of these while the microwave is running would add that corresponding amout of time.
I guess you'd also need a power level button and a clock set button. There, all done.
https://www.webstaurantstore.com/panasonic-ne-17521-stainles...
Having home/office-use microwaves work like this is basically akin to forcing an average office user expecting something like Microsoft Word, to write Markdown in Vim instead...
If that average office user is going to write something (or use a microwave) more than a few times in their lives, it's worth to force them, because the (very small) cost of additional learning is almost immediately offset by gained efficiency.
Alas, what matters is short-time thinking, so UIs are designed for sellability instead of helping people do their tasks correctly and efficiently.
The problem is not designing for sellability. Is not acknowledging that (1) some people don't want to RTFM! and (2) other people are OK with RTFM but want serious efficiency and correctness gains for it, "I pain, now you gimme gain!".
So you figure out who your users are and design for your users.
Now, the absolute worst is the parent article's example, where someone managed to "design for no possible user" by both forcing people to RTFM and providing zero efficiency and correctness benefits in return. But designing for the wrong user is also bad.
...because you've just describes one of the most horribly unintuitive UI/Xs one could imagine. Some of the unforgivable sins it commits are:
- One button can have more than one meaning depending on context. Aka "modal interface" - this sucks in all cases except 2: (1) touch screens (where the buttons / labels displayed can actually change on screen) and (2) Vim
- You have 10 Buttons instead of one simple Knob: on top of making UI elements do different stuff depending on previous action, you also have a ~5x excess of UI elements. Instead of 2 Knobs, one for time and one for power. Or one know, and a buttons above it to select time or power (I actually had to use a microwave with this setup btw, and hated that pos, I was always in a terrible hurry and made mistakes selecting wither the power or time and ended up with cold or burn food... but I'd still pick this over the 10-keys + extra buttons)
Seriously, don't ever design a UI/X for other people, if you ever need this done hire someone else! Also, if you're and UI/X-designer, please consider changing careers, seriously (I'm sure you're not, but just to be safe :P).
There were other things you got "for free" after that - like defrost-without-cooking by ensuring all pixels remained below a target temp, but got above some other temp.
Hope that guy builds it.
Put that option as the heating default, with literally what I just wrote explaining it next to it, and problem solved.
For cooking people will either use recipe from a book provided with microwave and thus they need just a selection of a program, or they will use more advanced microwave with oven and will want to set temperature of the oven to something like 200 degrees for 40 minutes. In this case sensor would be useful, but it's already not a microwave story.
Users of a microwave currently expect extremely poor temperature control, and so don't even think about temperature because you get what you get. An interface which lets me heat something to exactly "drinkable but hot" is a revolutionary change in how I use the appliance though.
Same story with warming food: sure, you want it warm all the way through - if you actually know the surface temperature though, then you can model that - food that is cold inside is going to have different surface temperature decay rates to food which is warm inside, and some modeling of the observed thermodynamics means the appliance can know this.
So now we'd really be talking: "warm to safe temperature all the way through" and we can give a probability / % complete read out of how close we think we are to it. No more guessing, no more user interaction, also no burning yourself because some parts went superheated.
Combined with the easy-to-understand knobs, this does sound much better.
Defrosting x grams typically takes y seconds, which is just a table lookup.
As defrosting most typically involves leftovers (rather than pre-packaged meals or ingredients), you will probably only know their weight, not how long it will take to defrost off the top of your head.
You could also change the screen backlight color in DEFROST mode to reinforce the mode change.
How would one know that? Do people routinely put leftovers on a scale before putting them into a microwave?
The Microwave starts when you close the door.
This can easily be fixed by coloring the indicator mark with a black permanent marker.
Translated to web, you're basically saying "all websites should use a high contrast serif font at >16pt. to avoid people committing dangerous mistakes when reading instructions online" :)
The analogy is more like web pages using scanned images of hand writing, which I'd also say is bad design.
The only real improvement is the replacement of the door opening button with a more intuitive door handle.
Actually, it's not an improvement. A door opening button is easy to operate with one hand; a handle needs two in order not to shift the microwave. At least all the microwaves I've seen have doors requiring noticeable amount of force to open. My guess is it's to avoid the situation in which a large dish rotating inside pushes on the doors and opens it.
The one I eventually found was also the cheapest one and almost the only one. The more expensive models were usually designed around a button interface and had plenty of features like thawing, grill, and so on. Of course all these extra features amount to little more than switching the magnetron on and off according to a programme. Go figure...
My mom (who has since taken over it) was very confused about the one second delay between adjusting the time and the microwave starting (just like the linked one there was no start button). She had obviously been accustomed to pressing buttons to get it to operate how she wanted. Which just maybe goes to prove that what is "simple design" is not necessarily the simplest interface but the interface most aligned to the users' previous expectations on how this particular item should work.
And at what temperature?
And if you want to cook in stages?
We have at home a programmable one from the 1980s. We constantly use it for stuff like "unfreeze 2 hours, then microwave 1 minute at 800W, then heat at 230C for 20 minutes, then keep warm at 40C"
Though perhaps a metastory might be "how not to write a UI essay".
(I actually rather liked this one. The off-the-rails discussion suggests either I'm an exception or the point isn't being clearly communicated.)
Seriously, what is it with embedded digital electronics designers and beeping? Nothing else has a gratuitous noisemaker. I swear if the home refrigerator was invented after the cheap piezo buzzer it would emit an ear-piercing bleep every time the door opened, or maybe when the compressor cycled on or off.
Is there some sort of strange subsidy or tax consequence so that adding a beeper has negative net cost?
A big ol' beep/nobeep switch on ovens would be a nice thing.
There's basically just one microwave I'm aware of that is actually proper design, mentioned here: https://news.ycombinator.com/item?id=16090081
First of all, "Doing the job as efficiently as possible" is not a user's desire. Users always want concrete things and developer's opinion on what that could be or how that will be achieved by a particular user is very often quite far from reality.
Second, it's not a developer's job to create a tool that will help users to solve some problem in the way he thinks it is right. It is not what developer is paid for. His job is to work together with marketing department to deliver a product that is always a compromise between the business and the user, a product, that is just good enough, in a deal, which is good only then, when noone is happy with it, but everyone can accept it.
I refuse to accept that is right, even though it is what seems to be happening everywhere. That's probably the root of my opinion. I'm a believer in the concept of making money through providing value to the customer.
If you look at a plane's cockpit controls, not being a pilot one has absolutely 100% zero clue to what each button does.
They should place a monitor screen, and every time u press a button it should do a pseudo unix man command and output all the options on screen.
We don't think this is bad UI, because we expect someone (a pilot in this case) to have implicit knowledge beforehand .
If the picture in the article is the actual oven, well.. Like a piano not having each actual note name written on it, aesthetics often minimize straightforward information.
While ambiguous perhaps, I think this oven is meant for someone who is desigining his own Swedish getaway and is meant to be minimalist to fit in with the rest off the kitchen design. 99% off the time when buying appliances they come with a manual in a dozen languages. This is because one buys it for consumer usage, which as an office (random) user you can't expect guaranteed access to the manual.
What if this manual has 8 pseudo chapters, 1 to describe each button. With some graphical examples of press sequences of common dishes. It might be that simple (and I can tell you that manuals do come simple like this owning a hybrid myself).
Its 10 serious minutes in the manual for 10 years not-even-think-what-you-press experience.
TLDR
One who never made coffee in his life, present him this: https://media-cdn.tripadvisor.com/media/photo-s/0e/7f/0a/6b/...
And one likely wouldn't get coffee in 30 seconds.
Present them this: http://outwestcoffee.com.au/wp-content/uploads/2015/10/Necta...
And my granny can do it in 10.
Different instances off the same appliance for different environments.
The FIRST thing you do is observe. Watch people in relation to the problem you are trying to solve. Listen to them. Ask them WHY they do what you observe. If at all possible, work out what motivates them and what problems they really have (which they often will not articulate to you). Only then start designing solutions.
I'm not saying his ideas about the design of the oven are invalid. But the fact that so few people in design roles even attempt the work of exploratory research is fundamentally the reason why so much of stuff is shit.
This isn't surprising though. Observational research is often frustrating, confusing or difficult to interpret. But above all it takes a lot of time. This is also the reason why, if you have another job like coding or marketing or whatever then you will not have enough hours in the day to do effective design.
I'd give both kidneys if it could then also ping me on my phone when it's done without involving external servers.