Wow, this is definitely the worst example. At least it's not in a security sensitive context.

Makes you wonder if anyone at the company ever even tested it with food in the oven

I dont mind the fact that they havent tested it with food, but I cant understand how they never recalled every single unit after noticing it for the first time.

Its like they see it, and be like "Ah, everyone who bought it got screwed over, and it will hurt our brand, but its still cheaper to quietly ignore it". Despisable

Usually what happens is that it's tested under ideal "lab conditions", so this never happens. In real life ovens get a bit grimy and produce more smoke. Stuff like that. Still shoddy engineering of course.

It's the same with designers doing their light-grey text on a white background with their 8K colour-perfect screen in optimal lighting conditions, and then when you point out this is difficult to read they go "I don't see the problem!"

Do you have a recent example of a professional designer doing “light-grey text on a white background”?

Because the designers I’ve worked with would never ship that. But maybe I’ve just worked with competent designers.

Recent examples I encountered include (not necessarily the best examples, just that I recall off-hand):

- Neo4j docs: https://neo4j.com/docs/cypher-manual/current/introduction/cy...

- RabbitMQ docs (esp. that sidebar): https://www.rabbitmq.com/docs/reliability

- Argo Workflows UI (no link, as you need to login).

- CV from "senior UX engineer" I received yesterday in response to a job ad I posted.

- Just now I found https://www.nngroup.com/articles/low-contrast/ when searching for something else – the quote at the top is nigh-unreadable due to the "font-weight: 250" which has the same kind of effect as low-contrast grey text.

- I've also had some discussions with designers over the years. Some view their work as "art" and get incredibly defensive about even minor changes done for real pragmatic reasons. Of course, there are also plenty other more pragmatic and competent designers out there.

- HN does it for downvoted/dead comments and "text posts" such as ask/show HN. Dang said it's a feature. Many disagree.

It's not as prevalent as it once was – it was even worse 10 years ago – but it's still encountered fairly regularly.

Semantics, but the #5e636a (39% lightness) text of Neo4j and #1c1e21 (12% lightness) of RabbitMQ aren't what I would consider _light_ grey. That would be up in the #bbb-eee range, or 75%+ lightness (black 0%, dark grey 25%, grey 50%, light grey 75%, white 100%). And I would be surprised if designers were involved in those 2 documentation sites.

Font weight is a crucial factor of readability, and it depends on screen specifics. On my 2020 M1 mbp at ~40% screen brightness, the NNGroup link quote is quite readable. As it is on my phone.

I don't rate any designer or developer very highly if they're too precious about their "art".

The HN dead/downvoted comments is contentious for sure. I don't agree with the choice fwiw.

Ah you are one of those people who believe nobody in the whole world has worse eyesight or screen or lighting conditions than what you presently have.

Nope. And not sure how you made the leap there from me giving context into my environment and equipment and how I was perceiving some text on the screen.

I perceive it as "light grey", so shrug.

It's all "readable" in the sense of "I can read it", but not in the sense of "I can read it effortlessly". I have a bit of CSS in Stylus to fix it, and it takes noticeably less effort to read it with a "normal" font. The RabbitMQ menu is just so much easier to scan as well with a more normal colour.

Good distinction between light grey setting and perception (I was speaking to the former). Maybe on your screen/OS/browser for a given font, a setting of #000 would start at (perceived) “grey”.

And agreed that readability is a scale and it’s best to be on the “easy” end of that scale.

I work with such designers.

They also love to disable the mouse scroll and the scrollbars, so the page has 300 more settings but you have no way of knowing that (this also happened to me on windows 11 btw).

It reached the point where I implemented my own script to bypass the GUI at work.

Oh yes… never buy a Miele oven with touch buttons

Same with Bosch.

Two problems:

Buttons stopped working after warranty expired so had to pay for a service call to have it fixed. Luckily no parts were needed. I don't recall the reason right now.

It has a spinny disc, so like a potentiometer but not. It is a flat removable ring and behind it it uses a touch button of sorts

You have to pull it off amd clean it before every use for it to work and when it does work it is very fiddly to use.

That pseudo-potentiometer is sick

And ridiculously expensive to replace for what is a glorified magnet if you happen to accidentally burn the plastic a little bit which is enough for it to becoe unusable.

Same for Smeg.

We have a Smeg oven, not with touchscreen controls, but with two pushable knobs that are easily pressed (thus starting the oven) by brushing past them. This oven has the worst user experience of anything, hardware or software, I've even used.

In what conditions do you live that brushing past an oven is an often reoccuring situation?

> In what conditions do you live that brushing past an oven is an often reoccurring situation?

I live with my 5 adult sons in a house with a small kitchen. Every hip-level surface gets smacked regularly.

I used one at a relative's house and I agree - worst appliance user experience that I've come across.

I have that same control panel, but on an oven/microwave combo. It's truly horrible.

Our new house came with a new Samsung dishwasher that had touch controls along the top lip of the door, and the door popped open at the end of every wash to let steam out. Imagine heated clouds of water passing over the panel every time. The panel started acting strangely/inconsistently within 3 years, and then by year 4 it was dead.

Some of these flaws are crazy. Do companies release products without actually trying them out?

Best believe they're optimized for maximal replacement sales.

Reminds me of some rental cars I had over the years (Buck being my favorite). I couldn't imagine anybody actually trying to drive the car before releasing. They were so bad.

It's very simple: what sells on the showroom floor and what works at home aren't necessarily the same thing.

not sure about that. How many people have bought an oven _because_ it had a touchscreen?

Probably not many explicitly, but I bet it contributes to a lot of shoppers thinking the appliance is "sleek" or "modern". Touch screens have been a big design fad in recent years and design fads sell products.

I'm hopeful that the tide is turning on these designs as more people have to use them day to day and realize that touchscreens are categorically worse than tactile controls in a number of scenarios.

Often times there’s simply no alternative. When we purchased our current stovetop we didn’t find any with non-touch controls.

i used to have one of those too. the only thing i liked about that that they were easy to clean.

I think they do, but too late and then it would be too expensive to go back and redesign the product.

Same experience with an induction cooktop with touch controls. At least once a week this is what that looks like:

1. Place pot of water on element to boil

2. Enable boost mode

3. Water reaches boil as I'm distracted with other prep / child / HN post, and overflows

4. All controls (including ability to disable boost, reduce heat, or turn off element) rendered completely inoperable due to liquid on glass surface impacting pcap sensing

5. Dry, repeat.

Not only that, but if you don't dry it quick enough the whole hob shuts down

I moved into a new apartment with an induction cooktop and radiators with the same type of "buttonless button" controls and those continually give me problems. You can't just touch them, but you have to slide your finger over the controls in just the right way, and hope that (for the radiator) you've hit the 1 in 3 chance of it actually working.

My next cooktop will probably still be induction, but it will definitely have knobs.

To my enormous relief, the Samsung stove that came with our house gets this right - even covered in water the controls still work. I was previously dead-against induction just because I couldn't find a good quality stove-top with physical buttons, to the point where I'd already budgeted replacing it with gas when we bought the house. I don't know how they've done it, but it's the only one I've ever used that manages. (For comparison, the high-end Miele and cheap-and-cheerful non-branded ones I've used in various rental houses over the last 20 years have all had exactly the problem you describe above).

A few of the new home models (in Europe at least) come with physical knobs.

My two gripes about induction are the touch controls they typically ship with and the inability to roast peppers over an open flame. But the incredible temperature response makes up for both IMO.

My Whirlpool stove does the opposite thing: after I've finished cooking and it's cooling down, it will activate the Warm oven setting on its own. I find that the Warm touch button is especially sensitive on its own, so there must be some small flexing happening somewhere and activating it.

Solution: Immediately after cooking I walk over to the junction box and turn off the breaker to the stovetop.

My oven doesn't have a touchscreen but it does have touch controls and it works perfectly, it's never disturbed by the steam from the oven even when it quite visibly condenses on the surface.

My point is that it isn't always a design failure to use touch controls, sometimes it is an implementation failure that makes them unusable.

It's still a design failure, just not for that particular reason.

What makes it a design failure? It seems good to me. The surface is completely flat glass and easily cleaned, there is no possibility for dirt to get into the switches or accumulate in crevices.