Hacker News Reader

> This is because they, one, don't listen to anyone ever; and two, never use the products they design. Oh, sure, there might be a user study or two. But they never actually have to live with or work with the damn things.

I wonder the percentage of consumer (and business - have you ever used a POS in a pub?!) electronics this applies to? It's gotta be high 90s. I've lost track of the number of times I've lamented "did anyone actually use this?"

A friend of a friend retired from the tech industry to become a joint-investor in a small chain of pubs. He ended up writing his own Linux-based POS software for the pub tills, AIUI...

> I've lost track of the number of times I've lamented "did anyone actually use this?"

We have a combination oven and microwave (yes...) that seems to have been designed with the primary purpose of enraging the owner.

The microwave can only be set in 15 second increments. But you can fine-tune it, in five-second increments, by spinning a dial. Except that you can't do that while the oven is on.

The microwave makes a noise while it's operating. When it's finished, it will beep unobtrusively and then continue to make the operating noise. You can stop it from making the operating noise by opening the door.

The microwave can't be replaced, because it's part of the oven.

> The microwave can only be set in 15 second increments. But you can fine-tune it, in five-second increments, by spinning a dial

Mine is similar, also a combi. You spin the dial and it will go up in 10 second increments to a point, then in minutes, then in 5 minutes, then ten minutes, etc. Fine, I get it, it's like floating point. But it starts doing 5 minutes at the 10 minute mark! So you can't do 11-14 minutes. Which is actually about right for some things in an oven.

You also can't add more than 10 minutes to a short time without stopping the oven (this, I assume, is intended to be a safety thing to avoid being able to accidentally add hours of cooking without positive user intent).

And I do wish it would just stop the fan once it stops cooking, but I suppose the electronics next to the hot oven don't suddenly need less cooling just because the hot metal box doesn't have food in it.

I do like it though because it's much more efficient for small things than a big oven and I don't have space for a toaster oven and a microwave. And it has a much better UI for setting a delay start than the 6-button interface of the big oven.

> Mine is similar, also a combi.

> I do like it though because it's much more efficient for small things than a big oven and I don't have space for a toaster oven and a microwave.

You may have gotten the wrong idea. This isn't a combination where the same box functions as an oven and a microwave. This is a microwave that is for no reason physically attached to a separate oven - it takes up all the space of a big oven plus a normally sized microwave. The only thing they share is the display panel. And yet that didn't stop anyone from disabling microwave functionality while the oven is on.

There is absolutely no advantage to this compared to having an oven and a separate microwave, and many, many, many disadvantages. It continually boggles my mind that anyone ever designed, manufactured, or purchased this thing.

If "the operating noise" is the cooling fan for the magnetron, it keeps going when the time is up to cool it down. Mine does the same, but actually says "COOL" on the screen when it does so.

The problem with this is that it prevents you from knowing when the microwave has finished.

The cooling fan isn't supposed to be an audible "microwave is on" alert. That's not its design function.

Perhaps the actual audible alert on yours should be louder.

My old microwave oven has a fan that is on only when the microwaves are on. If that was possible in 1985 it can be made the same way now.

No doubt it was some combination of lower power, physically larger or simply more prone to failure. Do you suppose the designers go to the trouble of a cool-down timer just for the hell of it?

I always use Reverse Hanlon's Razor, so I suppose they do it in bad faith. /s

Actually I guess they both shave off weight/cost and design for built-in model first (that are naturally more heat-constrained) and make standalone variant as an afterthought.

It's more reliable if you connect the cooling fan to a thermostat rather than the power switch. Otherwise the hottest time is actually after you turn it off (as heat from the inside leaks to the outside but is no longer being cooled by the fan.)

That must be frustrating, but was quite funny to read.

I would happily invest in kitchen appliances that have a silent mode for apartments and shared living spaces.

My microwave beeps at the end, and then beeps again 10 minutes later once the food has been removed. My old toaster beeped 5 times when the toast popped, which is unnecessary when the popping action is loud on it's own.

You might be able to remove the “buzzer” or speaker in the device.

If it's a regular piezo buzzer, just stick a small rare earth magnet to it. Instant silence, and doesn't damage anything!

This is brilliant - I bet there's not too many sizes of piezo buzzer so someone could make little kits to sell on amazon et al ...

Have you looked to see if your microwave has a volume control menu? After years of listening to my microwave blare across the house I saw a tweet that informed me that most microwaves have a menu setting somewhere to control the volume and sure enough, my Frigidaire microwave actually does. I had just never thought to look.

> by spinning a dial

When we moved into our current house, it had a microwave like that. When we replaced it, my primary requirements were 1. multiple fan speeds, and 2. a nice, simple touch pad for entering the time; preferably one where you didn't need to press a button before starting to enter the cooking time.

The dial thing is... infuriating. It never once found a situation where it made using the device better, and in most situations it made it worse.

> The dial thing is... infuriating. It never once found a situation where it made using the device better, and in most situations it made it worse.

Digital dials can be nice in some cases.

My parents had an oven with a dial for the temperature. Spin it either way to increase or decrease the temperature by extra 5 degrees. I loved it. It was so much better than the touchpad design. The range of actual temperatures in an oven is pretty small and I found this much more friendly than punching in the temperature, getting it wrong, cancelling, and punching it in again. It was a “finger tip” dial so it was quick to spin. Probably horrible for people with bad arthritis or other hand mobility issues, though.

I wouldn’t mind a microwave with a similar dial for time. Again, the actual range of times in typical use (for me) is not that high. The design described with a two tier system for adjusting time seems moronic though.

> The range of actual temperatures in an oven is pretty small

The range of temperatures you use is about 200 degrees?

Maybe 250. 300-550. I rarely go up to 550 and never below 300 that I can recall. With 5 degree increments, that’s only 50 clicks, fewer if the system is speed sensitive.

For temperature, I can see it working... though I don't see how it would actually be _better_ than tapping in the number.

For the microwave time, it was a horrible experience pretty much every time I interacted with the device. A keypad is simple, intuitive, and easy to use.

Keypads should be simple and easy to use, but they often are not. The buttons are inconsistently responsive and there’s always at least one other key you have to press, often two. And those other two are named whatever the designers decided and placed wherever the thought would be most confusing.

“Okay, I want to cook for 3 minutes.”

‘300’

”Ok, the numbers don’t work until I press something else. Hmmm”

‘Time Cook’ ‘400’

“Shit. I hit the 4. Where’s the cancel?

‘Cancel’ ‘Time Cook’ ‘300’

“Where’s the start?”

‘Start’

High quality dials with adaptive increments are amazing and much faster & easier to use than a touch pad.

My microwave oven uses a dial, and it works great. It uses increasing intervals between dial ticks (5 seconds for cooking times under 1 minute, 10 seconds for cooking times under 3 minutes or so, than 30 seconds, than 1 minute, than 5 minutes; I don't know the cut-of points by heart). The intervals are chosen in a way that they're precise where they need to be, and large where they don't need to be precise. It's really easy and fast to set the time you need.

It's a bit limited, in the sense that you can't set the time to e.g. 17:27. But who needs that?

When my old mechanical kitchen timer died on me, I bought a digital one with a dial, in the hope that it work the same way as the one on my microwave oven. Boy was I wrong. It kinda increases the interval when you spin it faster, but in a completely unpredictable and impractical way.

So I guess it's easy to make impractical dials, but it certainly is possible to make very well functioning dials as well.

> The microwave makes a noise while it's operating. When it's finished, it will beep unobtrusively and then continue to make the operating noise. You can stop it from making the operating noise by opening the door.

Check out Chesterton's Fence.

Tell us why you think it’s relevant.

Sure. Sibling comments to my comment explain that the continued "operating noise" is a cooling mechanism that's required by the microwave oven.

So, it wouldn't be a good idea to just rip that part out of the design. Even if it may be annoying. Perhaps a better explanation is needed, like some models showing "cool" in the display. But it's per se not a design oversight.

Chesterton's Fence is about not changing things before understanding what led to the current state of affairs. I think the principle easily extends to complaining as well.

No, absolutely not. This is not a problem that other microwaves have. It seriously impairs the functioning of the microwave. Chesterton's Fence tells us that there is no cost to eliminating this ridiculously awful functionality, because other microwaves already don't do it. An explanation doesn't help in any way. The solution is to stop misbehaving.

Maybe those other microwaves have a different design that doesn't make it necessary. So, sure, you can re-design the microwave oven, but that may entail significantly more work than to just deactivate the fan.

I could criticize cars with combustion engines for having one and being so noisy/stinky/... Then I could go on claiming that they should just turn them off, cause there are cars without combustion engines which are less noisy/stinky/... I'd be missing that there is more to it than just turning off the engine. I'd need to replace it with something else, like an EV. Maybe that did not exist yet when that car was build. So, claiming that this old Ford from 1970 having a combustion engine is stupid because "this is not a problem that other cars have" is missing the point. I could go rip the engine out because I don't like that it's noisy/stinky/... But I'm in for a surprise here. The vehicle won't move anymore. Something I'd have understood had I followed the principle put forward by Chesterton's Fence.

> Chesterton's Fence tells us that there is no cost to eliminating this ridiculously awful functionality

No. It tells you to first understand what you are doing. Blindly claiming something misses that part.

> did anyone actually use this?

Are you me? Lol. Sometimes I use software where I wonder "am I the only one that bought this and tried to use it?" I think part of the issue is that requirements get written by people who have no experience. They've never done the job the software will be used to augment and they don't know anything about software development. Then the requirements are used to outsource the actual design and development.

Using POS as an example (in a country with tipping culture). Imagine if a requirements document for the payment system said "must include an option to tip", but the design and development was done in a country where tipping culture isn't a thing. You'd likely end up with the "tip" option hidden away in a hard to reach place while everyone who uses the software knows it should be prominent and easy to use and could immediately tell you the design is poor.

> Yeah I wish we could get those prices.

You might enjoy [1] about a microcontroller that's ~3 cents, in quantity 10.

Of course it's only one-time-programmable, and generally pretty basic. And that's a pre-supply-chain-crisis price - these days prices have risen to 9 cents [2]. Still very cheap though!

[1] https://www.youtube.com/watch?v=VYhAGnsnO7w [2] https://www.lcsc.com/product-detail/Microcontroller-Units-MC...

Have you developed anything with that microcontroller, or tried to use that datasheet?

An ATTiny10 or PIC16F is so much better documented, so much better supported, so much more featureful, and so much better in every way that the average small-volume product design house will never make up the 40-cent price difference per part. How many development hours or support hours can you bill before you're penny-wise and pound foolish?

Even at a higher level than microcontrollers, i've worked with arm chips whose only documentation was random posts in chinese characters on sites i had never heard about before. They were cheap, but for the small volume devices i was working on dev time became significant.

Because the chips are so simple, I actually found the datasheets for Padauks to be sufficient. In particular, I don't think it's any more confusing than say STM32 or even NRF52 datasheets, mostly because the chips are so much easier and thus the datasheets are sooooo much shorter. Granted, I have not worked with 8-bit MCUs other than Padauks.

Their compiler/macro-assembler also seems to be working just fine, once you get past the UI from early 2000s. The error messages that must have been auto-translated from Traditional Chinese, but you squint a little bit and get the idea.

There is an SDCC backend as well.

Make the battery removable, then you have an off-switch and a repairable device. For bonus points, use a battery in a standardized form factor so I can buy spares.

I know, I know, the industry doesn't want me to keep my device working longer than the lifespan of a Li-ion cell.

the real problem is the design team doesn't give two shits about neither users nor developers of the product... developers aren't treated as customers by design teams and then everybody is confused why the thing doesn't work. (hint: maybe it's stupid hard to make it work if you are actively prevented to be able to work on it?)

the real problem is that you can't keep all those little features you'd wish for in your head, so while you are making a buying decision they don't factor in as much as the lower price achievable without those features.

That varies a lot based on product type, price point, and target market.

For something long-lived and expensive, I will pay extra and accept a less sleek design for standardized removable batteries. I research most purchases of durable goods. Of course, I may not be the most profitable customer since I don't plan to buy a new widget every six months.

This is the real problem. Otherwise the markets would have solved the crappy products problem a long time ago.

When I use something at my home, I usually have a constant stream of minor annoyances that could easily be fixed by the manufacturer. However, when I'm shopping for a new product, I dont remember any of them.

The odd part is that for many people that's even the case if it's software _and_ they are programmers themselves _and_ it's open source. But they go complain instead of just fixing the minor annoyance.

Perhaps we need government labels like we have on food - little boxes for common things like “battery: removable” or “on/off”.

I've advocated similar as a possible way of countering other consumer-hostile practices in the past but for something like batteries I think we're already past that point.

E-waste is a huge but so far mostly overlooked problem. Built-in obsolescence is great for exploitative businesses and bad for basically everyone and everything else. Unnecessarily restricting hardware repair or replacement might be the largest contributory factor in this problem.

Time to regulate.

I'd love this - simultaneously beneficial to consumers, improving the market (reducing information asymmetry), and not something that laissez-faire-market-types can complain about (because you're not restricting what sellers can do, technically - you're just forcing them to reveal more information about their product, and then it's market forces that actually deal the killing blow).

More ideas: "online account required for use", "collects location data", "phone call required to cancel subscription", "memory use: 350 MB" (just like calorie counts!), and so many others...

We'd need some regulation for the regulations. My son's laptop battery died last week. Today, it is considered non-removable. But, he did remove it, after removing about 20 screws, carefully prying open the case, removing a few more screws and layers of insulation inside. Eventually, he was able to remove it and test that the battery was the fault and order a new one.

If we require labels, without the necessary strictly defined definitions, today's non-removable battery becomes tomorrow's "removable" battery. Because, hey, it is possible to remove it. Good luck getting everything back together if you aren't an engineer or extremely careful in how you disassembled things.

If a battery is "removable", that means replacing the battery is a normal, reasonable use of the device. Ergo, if the device breaks while the customer is replacing the battery, the manufacturer would need to replace it under warranty.

"Field replaceable" might be better way to describe it. If my flashlight runs out of power, I unscrew the tailcap, take out the 14500 cell, and put another one in it. I don't need tools and I can do it in the dark; it would be a major flaw for a flashlight if I could not.

ikr? They have these things now, they call them "AAA" batteries...holy heck, you can take them out, and they even make rechargeable ones. Like, if your thing doesn't work, you can just swap the batteries instead of waiting N hours for it to charge itself....what a concept.

Rechargeable AAAs are crap though. Power density sucks. Voltage level sucks (and voltage regulators will suck the battery dry before your micro does). And the form factor sucks.

I'd sooner have a 18650-family (e.g. 14240 or 14500) become standard than a bunch of AAAs.

> Rechargeable AAAs are crap though.

Not compared to alkalines, not since about 2005. Voltage regulation (i.e. a boost converter) plays well with NiMH rechargeables too, though not with alkalines.

That's assuming one of them is an appropriate size and voltage for the application. If more than one starts to seem like a good idea, consider a larger cell, higher voltage chemistry, or both.

> Rechargeable AAAs are crap though. Power density sucks. Voltage level sucks (and voltage regulators will suck the battery dry before your micro does). And the form factor sucks.

Not in my experience: our main TV is connected to a 2008 computer running Linux Mint that we use for netflix and amz prime. I've been using it in this way since 2016. The remote is an off-the-shelf wireless pointer thingy with a few buttons on it (works like a Wii controller) that takes 2x AAA batteries.

I've been recharging those 2x AAA batteries since 1996, and it still works fine enough that I don't feel compelled to replace them.

Panasonic Eneloops and similar modern low-self-discharge NiMH batteries are really nice. Their capacity is not quite as high as alkalines, but their internal resistance is an order of magnitude lower. At loads over 100mA, they're a far superior battery in both ability to deliver power and effective capacity. They have low self-discharge, so you can charge them after using them instead of discovering you need to charge them before using them. And I've never had one leak, but I assume they can.

I use a lot of rechargeable AAAs and AAs. For AAA, 1100mAh are pretty good, which almost up to the ~1250mAh of an Energizer alkaline. For AA, there are 2800mAh rechargeables available, which is on par with an Energizer alkaline.

Battery hipsterism. Oh HN, you keep giving.

It was a valid point. There's a significant performance/weight reason most of the devices with non-removable batteries and built-in charging are using Li-ion instead of NiMH.

Getting a bunch of consumers used to buying spare Li-ion cells and using external chargers might be a non-starter, but it's possible to have both onboard charging and removable cells so the upgrade path is there for those who want it.

I've had a series of Canon powershot cameras that have all used more-or-less the same Li-ion cell design[1] and I've been able to swap batteries with them. They're big enough that third parties have even offered replacement batteries.

There is probably room for standardization of a compact, rectangular, rechargeable battery of roughly those dimensions. Is someone pushing for it?

[1] https://www.apotelyt.com/camera-power/canon-g7-x-mark-iii-ba...

I don't think anyone is pushing for that, though sometimes a company will design a device around another company's battery design. There's quite a bit of small photo/video lighting that uses Sony NP series camera batteries. Those are just a pair of 18650s in series inside a plastic box.

It seems to me the size of the G7X could have been designed around a 14500 cell instead.

Thank goodness I haven't heard a hipster sing the praises of carbon-zinc "heavy duty" batteries from Harbor Freight and dollar stores... yet.

> and we only use microcontrollers we can get and read datasheets for, so we pay at least 20x that price.

Yeah, I see what you mean.

And even if google translate can help, there are of course other problems

20x might be the actual cheaper option

Sure, but you have a manager who thinks... against better judgement... again... that they are gonna sell 20 million units of the thing, so their math suggests that you should... yep... prematurely optimize.

> This is because they… never use the products they design. Oh, sure, there might be a user study or two. But they never actually have to live with or work with the damn things.

Yeah, I do software and Ive found the same thing: if I’m building it, then I’m actually the first user of the software, and I have all sorts of feedback for the designers. Luckily I’ve worked with some great designers who were usually more than happy to adjust the designs though.

> (Our design team will never listen. This is because they, one, don't listen to anyone ever; and two, never use the products they design. Oh, sure, there might be a user study or two. But they never actually have to live with or work with the damn things.)

If you don't mind me asking .. if it's that bad, why do you still work there? Is there a shortage in suitable jobs? (Honest question..)

I hope you don't mind this sentiment but you seem like one of the "good ones." A lot of designers seemingly do things like this for some insular reasons rather than for market reasons (better for consumers or it gives them an edge). Just another example how some developments in tech are top-down rather than bottom up from market demand.

> I shall use this post as ammunition next time I am arguing with our stupid design team

So many red flags here.

Can you tell whether you work in a product design company, or an engineering company?

Does that come from clients?

I used the same 3-in-1 latch circuit for power on/off, pairing, and software power off for 12 years.

A good enhancement mode NMOS FET can be as good as a physical switch at low voltages.

If you really need nano-amp level leakage current AND very low drop, you can consider turning to beefy depletion mode devices, which gate the user will discharge with the button press. A nice benefit is that you can directly sense the button press with an MCU.

I once even did a bluetooth mouse with this "hardware" switch, and a graceful shutdown: we sense voltage on the buck capacitor, and detect when the battery disconnect triggers.

The few microfarads in the capacitor are enough to send the last update with the 0% battery level, and a disconnect command.

> Can you tell whether you work in a product design company, or an engineering company?

Both! There aren't many of those out there, so I probably just outed myself. Oh well. Keep it quiet please?

> A good enhancement mode NMOS FET can be as good as a physical switch at low voltages.

Yes, it can work very well. You are correct that it need not be a physical switch. But it still does nothing if it is never brought out to the device's interface. And physical switches are still preferable if the software is broken, as it often happens to be during development.

> you can consider turning to beefy depletion mode devices

Depletion mode devices are not cost effective. If you think they are, point me to one cheaper than the LND150. (And, yes, I do know how to use them. Many designers do not.)

I would fire the whole team immediately if they really dont listen. Noone wants another e-waste.

I am pretty sure the market disagrees with you.

Environment disagrees with "market". "Market" takes whats available. Not producing ewaste means "market" cannot consume it. Simple.

Sure, but the point is that you can keep making these things and people will keep buying them, so "noone wants another e-waste" is clearly not true. Discussions about "the market", its effects on the environment, etc. are all important and worth having, but don't really have much bearing on whether or not "people want another e-waste".

I mean, if you look at it that way then people really want corrupt politicians and exploitative businesses because they keep electing the former and paying the latter.

I think people really don't want e-waste. They don't want oceans (and now bodies) full of plastic. The industries have just done a great job of convincing them nothing can be done about the problem.

> I think people really don't want e-waste. They don't want oceans (and now bodies) full of plastic. The industries have just done a great job of convincing them nothing can be done about the problem.

I think it's a bit more complex than that; a lot of properties that distinguish a "solid product" from "e-waste" are not at all obvious when you buy it, and it's not illegal and it is cheaper, and people got to make rent too, so... yeah. I agree: in the abstract most people don't "want e-waste", but they also look at costs and other factors, and they do "want e-waste" once you factor that in.

It probably shouldn't be strictly illegal too, but now you have a situation where:

1. Consumers blame industry and say that government should do something.

2. Industry says they're just producing what consumers want and that it's not illegal.

3. Government says they want consumers and industry to have free choice and call for different consumer actions and "self-regulation".

So basically, everyone is pointing at everyone else, and nothing gets done. And the arguments from the various sides aren't necessarily bad or malicious either: regulation does come with a cost, and cheap stuff does have its use.

Personally I think factoring in external environmental costs in to the product price would be good, but it's not so easy as it will make things more expensive, and already enough people are struggling to make ends meet (as the current high inflation rates show) so it all ties in to a lot of other issues as well.

> Noone wants another e-waste.

> I am pretty sure the market disagrees with you.

Not applicable - efficient market hypothesis requires a host of conditions to be true, and all of the following are unmet: information symmetry, sufficient number of suppliers, lack of monopoly effects, and rational consumers.