Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. An Empirical Study of Harness Design for Coding Agents(arxiv.org ↗)
    15comments
  2. I Vibed a Proof of Conway's Conjecture(overreacted.io ↗)
    12comments
  3. Bend 2 and the Vibe-Coding Trap(liampwll.com ↗)
    176comments
  4. I don't like passkeys(hawksley.dev ↗)
    298comments
  5. OpenJev(openjev.com ↗)
    182comments
  6. The Shadows Lurking in the Equations – Underwater Islands(gods.art ↗)
    3comments
  7. Subnormal floating-point numbers are expensive on Intel processors(lemire.me ↗)
    34comments
  8. Jemalloc 5.4.0(github.com/jemalloc ↗)
    61comments
  9. Warren Buffett Steps Down as Berkshire Chairman, Names Son to Replace Him(nytimes.com ↗)
    104comments
  10. The scourge of x86 emulation(fex-emu.com ↗)
    57comments
  11. Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint(prismml.com ↗)
    168comments
  12. Cekura (YC F24) Is Hiring(ycombinator.com ↗)
    discuss
  13. Microsoft exec called AI scraping 'the largest theft of labor in human history'(techcrunch.com ↗)
    477comments
  14. Astra for Law(openai.com ↗)
    644comments
  15. NATS publishes preliminary report on technical incident of 8 September(nats.aero ↗)
    2comments
  16. Bend – A language that blocks AI mistakes via proof, on CPU and GPU(bend-lang.com ↗)
    272comments
  17. Second Circuit Allows Government to Search Electronic Devices at the Border(knightcolumbia.org ↗)
    6comments
  18. Hister: A private search engine for the pages you visit and the files you keep(github.com/asciimoo ↗)
    178comments
  19. Replacing Pull Requests with Delta(zed.dev ↗)
    49comments
  20. Qwen 3.8 Omni Flash(qwen.ai ↗)
    101comments
  21. C++26: Trivial infinite loops are no longer undefined behaviour(sandordargo.com ↗)
    1comments
  22. Wax motor(wikipedia.org ↗)
    89comments
  23. ZCode, the GLM coding agent, silently uploads your Git history(tokenstead.ai ↗)
    47comments
  24. Rabbit Hole: Minimum L-seams(fractalkitty.com ↗)
    discuss
  25. When the fractional part of a float fixes your shader(crocidb.com ↗)
    13comments
  26. Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA(global.fujitsu ↗)
    242comments
  27. Pre-Greek: The lost language hidden within Ancient Greek(linguisticdiscovery.com ↗)
    57comments
  28. How to Write with an LLM(sockpuppet.org ↗)
    168comments
  29. Show HN: Navier-Stokes Visualized as 1kB i386 demos(juandecos.github.io ↗)
    4comments
  30. New wild cat species discovered – with only one known living member(bbc.com ↗)
    4comments

1,000 (bad) ideas

125 pointsby 16y agoblog.jgc.org
65 comments
16y agoHN ↗

Brake lights that show intensity: check. I have read of at least 2 proof-of-concept implementations - one is two red bars plus a white bar (white bar should be instantly recognizable as "reverse") another goes circle->triangle (warning) -> large hexagon (stop!)

Update: I believe that some Mercedes models actually have brake lights that flash during strong braking. Google "adaptive brake lights" for more info.

16y agoHN ↗

Many recent european cars automatically flash warnings when braking strongly.

16y agoHN ↗

If I remember correctly BMWs have a high intensity brake light under emergency braking.

16y agoHN ↗

I'd be astonished if blinking braking lights were allowed in a country like Germany. I once read that a truck driver was fined in Germany because he had a small illuminated Christmas tree in his cockpit, by which he was breaking the specific part of their traffic laws that regulates vehicle lights. That must truly have felt like the most wonderful time of the year for that poor guy.

16y agoHN ↗

In Europe braking lights and turning lights are distinct and of different color (red vs orange). Many cars now flash the warnings (both turning lights together) when strongly braking, following a common usage among european motorists.

16y agoHN ↗

Yeah, at least here in Italy it's pretty common that if you have to slow down suddenly because the road is blocked (due to a traffic jam or whatever), you turn on the emergency blinking lights so the people behind you will be aware of the need to slow down and stop.

16y agoHN ↗

It's the same here in Slovenia and pretty much every other European country I've ever driven in.

16y agoHN ↗

Not so here in Norway. I've observed this behavior on several occasions when visiting Spain though.

16y agoHN ↗

I do this (Ontario, Canada), but I don't know if I've ever actually seen it done by anyone else. Our 400-series highways tend to have very good forward visibility, though—there isn't really such thing as a jam just over the crest of a hill.

16y agoHN ↗

401 exits can backfill right on to the actual highway though.

16y agoHN ↗

I do this in Ontario too, someone saved my life by doing this as I was entering the fog from around a corner and didn't have time to adjust my speed.

16y agoHN ↗

Well, it's even in France's law: if there is a strong slow down, you _must_ turn on blinking lights until the car following you does the same.

16y agoHN ↗

Strong braking normally == hazard lights + brake lights on new cars.

16y agoHN ↗

And will switch on their emergency blinkers after any sudden stops > 50 km / h to < 10.

16y agoHN ↗

My '05 BMW has 2 stage rear brake lights for regular braking and when ABS is triggered from hard braking or slippery conditions.

16y agoHN ↗

Why wouldn't a series of lights be the best/simplest pattern to implement this idea: [O O O ] == medium breaking intensity. etc..

16y agoHN ↗

I first saw motorcycle brake lights that flashed faster with the amount of braking at least 20 years ago in the US aftermarket.

I've seen them on the road maybe 20 times.

16y agoHN ↗

I've had a similar idea to the brakes, but in my version the intensity of light was linked to the intensity of the braking. That way if you're just cutting a few mph off your speed, it doesn't really matter and the light is pretty dim. If you're braking moderately, the light comes on bright enough to indicate that you're braking, but if you slam on the brakes then the light illuminates strongly and brightly.

16y agoHN ↗

Some braking strength indication standard should be established and mandated on new cars. This sort of thing is definitely needed in Houston. The frequency with which I have to panic stop in the middle of the freeway or other kind of high speed limited-access road is startling.

16y agoHN ↗

In-ear headphones that automatically pause music when removed. The details mention using either a proximity detector to discover when the headphones are in the ear, or a strain gauge to detect the pressure of the ear canal.

The general idea is good but the implementation is odd. It should be enough to do like the ipod/iphone, that pauses the music if you unplug the earphones. I don't know if it's a original idea by Apple, but I love it.

16y agoHN ↗

Yeah. I often take them out momentarily to talk to someone or do something and having to pull out the player and restart the track would be annoying.

Unless of course it detected when you put them back in and resumed.

16y agoHN ↗

well yeah, I'd say of course it would have to resume on re-insertion.

It's a great idea too, imo.

16y agoHN ↗

It's good but doesn't work exactly how I would like it to - if you are listening to something on headphones, plug the device into a docking station you get the audio out through the speakers but when you unplug the headphones the audio stops.

16y agoHN ↗

If you're going to talk to someone, the natural action would be to pull the buds out of your ear, rather than pull the lead out. This is why John's idea is different to (and probably better than) Apple's implementation (which is great).

It would even work for giving someone else the buds to hear a song you're listening to. Pauses while you hand them over.

16y agoHN ↗

Yeah, but I can't think of two similar actions where it would be difficult to define the expected behavior: sometimes you just take one bud out momentarily to be able to listen to something external and you don't really want the music to stop. The other one is common in teens, when they share one bud with a friend to listen to some track together and the music should be stopped while that bud is not put.

16y agoHN ↗

Both are solved by only pausing when both buds are removed (which is the most polite thing to do anyway).

16y agoHN ↗

There are companies out there making money selling paper and water. Having a great idea helps, but just creating something is the hardest part.

16y agoHN ↗

I think having ideas is actually linked to writing them down.

If you write an idea down and think about it, you show your brain interest in ideas so you will get more ideas over time.

I think it's an essential habit, to write and think about your ideas, when you want to be more creative and have many ideas.

Then the value in actually realizing all your ideas, is much less than picking the good ones and realize them.

16y agoHN ↗

Ideas sparking ideas is the whole premise behind IDEO style brain storming. You get a handful of people to start thinking of ideas for whatever problem you want to solve and one idea always inspires more. Crazy ideas are encouraged since they might also lead to good, tangible ones.

16y agoHN ↗

Great post - but I kinda disagree that an idea can be bad.

A system that pauses music when you remove headphones would solve a problem, thus it's a good idea - it's just a small issue of the implementation.

Probably most of those 1000 ideas are not bad, but might be difficult or infeasible to implement.

RE: headphones - there could be many possible solutions to this - I think Casio makes a watch that activates the backlight when you raise you rist to look at the time - same technology.

16y agoHN ↗

RE: variable brake-lights, Didn't Donald Norman mention this in "The design of everyday things" ?

16y agoHN ↗

I have (and used) one of these wrist watches. Power hungry, may I say...

16y agoHN ↗

It may solve a problem, but it creates a new one. Before listening to my mp3 player on a train, I leave the buds out of my ears and hold them in front of me, to gauge volume so as to not annoy my fellow passengers.

This practice is rendered impossible by this proposed technology.

That said, as an idea, it isn't bad. However, I contend that the statement that ideas can't be bad, is wrong. The idea of eliminating all Jews, gays and Gypsies to fix Germany's economic problems strikes me as a pretty bad idea.

16y agoHN ↗

This practice is rendered impossible by this proposed technology.

Put a fifteen second delay on it. Problem solved.

I have that habit too, actually, in the workplace for instance.

16y agoHN ↗

Point taken! Ok, ideas definately can be bad!

However I'm pretty sure the idea of ethnic cleansing and mass murder was conceived quite a lot earlier than the second world war...

16y agoHN ↗

I set my desktop to pause music when I lock the screen (and play again on unlock), it's really useful.

16y agoHN ↗

The number of ideas one has is a moot point really. More important is ability to discern quality. Most important is to bring the good ones to fruition!

16y agoHN ↗

How do you make it so neat? As soon as I write an idea, I start to iterate (generalizing, specifizing, implementing, MVP, MVImplementation, niches, copy, trying to find the essence of the idea, and how to express it). I'm not saying my way is better, since it's very difficult-impossible to later read as a reference. The only useful result is in my mind.

16y agoHN ↗

Actually, every one of hose ideas is a good one since they all speak to an underlying problem- that taking off your headphones means you miss your song, or that sudden braking leads to crashes. These ideas are only a quick proposed solution to the problem, but as anyone who has ever done a startup knows, it takes a long time to refine your product. So don't label these ideas as bad- I bet most of the problems they intended to solve still exist today.

16y agoHN ↗

I like not writing my ideas down, because it acts as a simple filter.

If an idea is actually compelling, then it will stay with me even if I don't write it down (and I'll continue to think about it, expand upon it, tinker with it, etc.). If I don't remember it, then it probably wasn't that interesting to begin with.

16y agoHN ↗

A "bad", or perhaps more accurately "marginal" or "nifty but unprofitable" can still help the creative process leading to "good" ideas. The marginal one may end up as a feature of the killer app you build later, or the unprofitable one may see some future profit potential as other factors change.

I track all mine, good or bad, and when I'm feeling uninspired or burned out on what I'm working on, I like to go back and at least have a laugh at all the old ideas.

16y agoHN ↗

i write my ideas down, but mostly to get them out of my head. or, more accurately, less in my head. i am often 'distracted' / overwhelmed by them. putting them on paper (well, textedit to be precise) makes it easier for me to let go of them.

16y agoHN ↗

Like the post a lot. Don't think they are "bad" ideas. Worse ideas are successful every day.

Not exciting to say, but ideas are a small part of success. Or said another way, to implement one idea, you need 1000 good ideas.

16y agoHN ↗

For a couple of years, my lab managers would rely on my idea rate in setting quarterly goals for number of invention disclosures filed. Yes, we did MBO and OKR even in research labs.

Fortunately, I was able to keep my filing to disclosure ratio quite high (21/36) (even if it was never tracked officially). Agree that just because the company decides to file a patent does not automatically make it a good idea.

Eventually I got back to coding and went away from ideas. The gulf between ideas and implementation is not just time, its also mindset. I am learning how to balance the two.

16y agoHN ↗

The starter-less engine.

If the ECU knows the position of the crank (because of the encoder on the crank these days) it also knows which piston is past top-dead-point.

Inject some fuel in to the air already present in the cylinder, spark and go!

No starter motor required.

Regarding the ideas, I think the way you are organized and how you have this set up is at least as impressive as the fact that you've had 1000 (not so bad) ideas. My way of 'filing' an idea is usually to register a domain name in the hope that when the time to fish for a new project comes around that I still remember what it was about.

16y agoHN ↗

I agree with the theme in a lot of these comments: It's a good idea if you have a plan of attack. I've had lots of ideas (eg battery powered heated wetsuit! damn, it's been done) and am now working on one where I have the unique combination of training, opportunity, luck, etc.

16y agoHN ↗

What does the HN community think of having a public ideas website? Sort of like ASK HN: Rate my startup. Users can post their ideas from inception and receive creative input, reassurance that someone would buy that, or no-way don't waste your time building that.

16y agoHN ↗

I've seen a few sites like that over the years but they die fairly quickly. This is not to say it's a waste of time, but to point out some basic problems such a site needs to overcome.

One problem is that a general forum is not attractive to specialists (who might have to invest time just to explain why something would be useful in the first place), with the result that suggestions trend towards a blend of trivial and naive. Another is the nagging sense that the more well-specified an idea is, the more likely it is to be be exploited for private gain, and since one can't patent or copyright an idea per se, and building a prototype plus filing an application for a patent is quite expensive, there's not much economic incentive to develop ideas to that point, even if the goal is to make the patent public domain (so as not to allow private monopolies).

16y agoHN ↗

"one can't patent... an idea per se"

Actually, this is precisely what patents are for. You don't need to implement an invention in order to obtain a patent, you just need to describe it in sufficient detail "as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use [the invention]" [1]

[1] http://www.bitlaw.com/source/35usc/index.html

16y agoHN ↗

This is true, though the patent office can require a model (s.114 via the same link), so in practice you'd probably want that on hand. I was just trying to distinguish between the 'cool' basic idea that's not patentable and the expensive detailed recipe that is.

16y agoHN ↗

I'm still wondering if this idea is good: http://techneur.com/post/635311152/announcing-mindspread Call it 'Pay-It-Forward for Books'

It's essentially a webapp where you give someone a book with the promise that when they are done reading it, they will then give it someone else, and so on and so forth. You could then visualize your 'impact' upon the world by seeing how many books you've spread and where they've gone.

Any input?

16y agoHN ↗

Thanks for pointing this out. There is one major difference, Mindspread is suppose to be completely altruistic.

16y agoHN ↗

Something similar is bookcrossing.com.

You read a book, then 'release it in to the wild' by leaving it in a public place with a sticker and unique book-crossing ID. Other people can then pick it up and pass it on, leaving a trail online of it's travel across the globe.

It can be entirely free, although there is a store where you can order the stickers, and unlock 'pro' features on the website.

16y agoHN ↗

Does anyone else generate ideas like this? And if so, what percentage turn out to be good?

Yes, though regrettably without your level of discipline and patience, so reviewing my old notebooks makes me feel scatterbrained. One thing I'm going to steal from you is the practice of numbering concepts: quantifying them is a great way to mitigate over- and under-confidence.

Of course, a lot of 'bad' ideas only seem so because it's not obvious at the time how the cost of implementation would justify the benefit. For example, I like your idea of pausing the music when the headphones are off. Downside: the requirement for sensors in the headphones, signal infrastructure going back to the sound source, and so on. Possible workaround: ability to measure micro-variations in power consumption consistent with worn/unworn pressure differentials.