Marvin Pipkin, inventor of the frosted light bulb | Hacker News Reader

Marvin Pipkin, inventor of the frosted light bulb

en.wikipedia.org

86 pointseverbody29 comments

> Not being aware his assignment was a type of joke, he went about the task as if it was something that could be done and had been done before.

I've read stories similar to this before, iirc assignments given as homework. Is there a name for this method, or comprehensive list of examples?

You're probably thinking of George Dantzig: https://en.wikipedia.org/wiki/George_Dantzig

It's not unbelievable that this happens. Ramanujan presumably solved some important problems without knowing their context in contemporary mathematics: https://en.wikipedia.org/wiki/Srinivasa_Ramanujan

Nobel prize winning physicist Abdus Salam (https://en.wikipedia.org/wiki/Abdus_Salam)

> During his doctoral studies, his mentors challenged him to solve within one year an intractable problem which had defied such great minds as Dirac and Feynman.[29] Within six months, Salam had found a solution for the renormalisation of meson theory.

I was not previously aware of the origin of the ubiquitous frosted light bulb. Thanks.

> The public felt that the glare from clear transparent incandescent lamps gave off a harshness that was unwanted by most people. Another side effect was that they made it hard to see objects in immediate vicinity of the lamp. They desired a lamp that had a softer light with no glare.

The first half of the 20th century was spent figuring out how to use artificial lights safely and effectively. In the last decade we've forgotten almost everything our predecessors figured out.

For example, my father had some great hanging lights over his kitchen stove. These provided just enough light to get around the kitchen - if you wanted more light, you could turn on the other lights.

His wife redecorated and replaced the red-colored globes with clear globes. Now there is no diffusion effect, and visitors without baseball hats get the 'ice picks to the eyes' ("glare") effect from direct exposure to the filaments.

They also installed terrible 3000K LED bulbs in their new fixtures. I switched them out for proper incandescents, so I don't get headaches when I visit. The only LED bulbs I find to be acceptable are some of the 2700K bulbs. If the 3000k bulbs are shielded by a lamp shade they're tolerable, but still slightly too blue.

Old street lights were fine. LED streetlights put out more more lumens per watt, but the quality is terrible. My brother's neighborhood has been LED-ified, and now has a terrible night experience. My city has mostly orange low pressure sodium [LPS] [0] lights still, but the state highway department is gradually switching out their LPS lights for "fake white" LEDs.

Flagstaff has Lowell Observatory (astronomers). They have wonderful night lighting. A few business flaunt Flagstaff's dark sky ordinances with non-shielded floodlights, but most of the city lights and businesses' lights are a pleasant orange (or yellow), and only over-stimulate my eyes when I look directly at them.

The "lighting resource center" has information about good lighting practices: https://www.lrc.rpi.edu/

[0] https://en.wikipedia.org/wiki/Sodium-vapor_lamp#Low-pressure...

I think some people must be more sensitive than others to the color temperature of light, myself included. My parents (who are otherwise wonderful people) use the 3500-4000K fluorescents in their living room instead of the much more warm and pleasant table lamps. When I walk my dog at night, I often see the blue-green glow of what must be 5000K bulbs coming from many of my neighbors houses. And my office mis-ordered bulbs for their track lights and ended up with a random mix of 3000K and 5000K bulbs; I seriously considered finding a new job until I was able to convince them to change the bulbs, but nobody else really seemed to mind.

By the way, Philips makes a line of dimmable LED bulbs that are 2700K at full brightness but go down to 2200K as they dim. Installing those in my dining room fixture was a revelatory experience.

Some people may be more sensitive than others, but there's a reason why computer glasses filter out blue light, and it's not simply personal.

Had a similar experience to yours with an office, except the blinding 5000k hangers also bothered a lot of other people, to the point where normally many of them were unscrewed/turned off... except when we had clients we were trying to land coming to the office. Then management decided that we needed to have them all burning full brightness in order to make a good impression for key visiting clients.

Lighting sensitivities are yet another thing that are easy to personalize for in sane office plans, but can be a big drain in an open floor.

I assume many have bought a light bulb that emits white light when they want a warm light. Everyone look like a zombie, it’s truely awful. Who wants those bulbs and why?

My SO uses 5000K lights for drawing, otherwise the colors are off in sunlight and in photos.

I also prefer outdoor lighting to be on the hotter side - low temperature lights to me look too much like sodium lamps.

I'm also hyper-sensitive to this. Our office has a stairwell light that has to be straight up 5000K fluorescent. It's awful. I found that Hyperikon's LEDs are all extremely accurate in their color temperatures, well matched amongst themselves, and support dimming. I use their 2700Ks in the office, and made sure everything matched.

I've been hypothesizing that people from countries that are closer to the equator prefer higher color temperatures ("colder", white light) than people from northern countries, who I think prefer warm light. I could be wrong, of course.

> LED streetlights put out more more lumens per watt, but the quality is terrible. My brother's neighborhood has been LED-ified, and now has a terrible night experience.

They did this to my street about a year or two ago. It was (and still is) horrible.

I take my dog on a walk every night. Suddenly, on one of these trips, I walked out my front door and into a horrible blue-white neighborhood. It looked like a car dealership out there. Not only too bright, but also blinding because they didn't put any kind of shroud over the bulbs.

I'm acclimated to it now, but I really wish they had done some better design on these retrofits. I may not get the comforting sodium yellow back, but there's no reason to crank them up and spread the light into pedestrians' and drivers' eyes.

> I take my dog on a walk every night. Suddenly, on one of these trips, I walked out my front door and into a horrible blue-white neighborhood

That reminds me... Almost 2 years ago I responded to HN user jacobolus' comment on bad LED lights, which also included his strategy for dealing with LED streetlights while walking his dog:

"[...] After San Francisco installed some awful LED lamps on my street, I’ve taken to wearing orange safety glasses when I walk my dog, and if I had to drive with any regularity I would consider wearing orange glasses to drive at night. [...]" - https://news.ycombinator.com/item?id=14170344

I got some orange safety glasses, but they cut out too much light to be safe while driving (easy to miss pedestrians). I eventually found some yellow glasses that cut out enough blue to eliminate the shock of sudden exposure.

Orange safety classes (mine are called "computer glasses" but I suspect they are basically the same--they filter out blue light) can also help to prevent eye strain from extended computer screen time, and in my case at least, ocular migrations.

I use flux to do the same. I cut the color temp to 5800K during the day, which is just enough to cut out those blues that cause headaches, and I go all the way to 1850K at night.

"I switched them out for proper incandescents, so I don't get headaches when I visit"

Why not just get 2700k LEDs? If you didn't like the Ferrari, because it wont carry the kids, don't go back to a horse and cart.

Specifically, stop buying really shitty LEDs and get Cree 2700ks.

Out of every LED I tested when I finally abandoned incandescent (manufacturers started shipping intentionally short-lived bulbs that would die in less than a year, when they used to last many many years before), Cree was the only brand that didn't feel like I was stabbing myself in the eyes, and also didn't have any weird hues or casts or anything (especially not the goddamned green-cast problem that so many LEDs suffered from).

GE and Phillips are distant seconds, and I don't even know why Sylvania even exists.

The only way to beat a Cree bulb is to start looking into specialty LEDs meant for photography usage (>95 CRI, >95 in R9 and R13, etc).

AKA, yujiintl and others. Still have a nice 4000K "remote phosphor" >95CRI, >80 R9 BC series lamp in my room. I considered their 1900K VTC (normally >98 CRI, but at that extreme iirc "only" >95 CRI) series LED's for true candlelight effect lighting, but they don't stock them (anymore). If someone does a kickstarter for a batch of them, I'd back it though.

LED bulbs are the worst thing. They use less energy but are just awful compared to incandescent light. I've seriously considered switching to candles or oil lamps for home at night.

I got the C by GE bulbs. They are great. You can adjust their temperature all the way down to 1800K (which is why I got them) and you can dim them, all via bluetooth, so they can go into any socket. I put them in the bathrooms so that we can have an almost candlelight experience when going pee in the middle of the night.

I also got a bunch of night lights from Costco that can also do 1800K. Basically, I replaced all the nighttime lighting in my house with 1800K LED, and try to avoid white light after the kids go to bed. It's made a world of difference.

If you find them that problematic, stick with halogen lamps. There's no shortage of supply, they're slightly more efficient than plain tungsten filaments and they give better light.

FWIW if you're looking for nice non-electric evening lights I'd recommend gas over candles. Some nice camping lanterns that put out a lovely light. Cripplingly inefficient, though.

>This idea of having a second treatment to smooth out the fine-grained texture to dimples came about by an accident.

From penicillin to vulcanized rubber to frosted glass. The crucial moment for so many inventions and discoveries was accident. But I remind myself that in nearly all cases a dedicated professional working diligently, iteratively for long periods of time is the perfect person to a) notice the accident in the first place and b) realize it’s significance and put it to use.

And c) overwork themselves even to get tired and have accidents.

Well, I ruined my own joke by typing "even" instead of "enough". And I didn't even invent anything in the process of having this accident.

"One day, while he was pouring the weaker solution into a bulb, the phone rang. In the process of answering the phone, he accidentally tipped the bulb over before it had enough time to finish cleaning out the previous etching. When he returned to his work, he accidentally knocked the glass bulb off the workbench and onto the floor. To his surprise it did not shatter, as etched bulbs normally did, but bounced a few times and then rolled under the workbench"

I'm always intrigued by the cognitive dissonance in rewarding inventors for doing difficult things, then lauding them for accidentally stumbling upon something.

Yes I get that you make your own luck etc. But luck doesn't seem to be accounted for in our present IP system.

The process of inventing is basically controlled stumbling.

Seems like you've higher chances on stumbling on something accidentally useful when you're trying to do difficult things rather than when not

Well yes, that's what I was talking about with making your own luck.

My point was that any discussion of IP revolves around work put in, and skill demonstrated. But a lot of the folklore is based on chance discovery.

I'm not sure what you're suggesting - that we only award patents in cases where clear evidence of hard work can be provided?

That isn't what I was suggesting no.

But...

It is a fun thought experiment though. All those 'on a computer' patents become harder to justify. Whereas billions of dollars and man decades of effort to marginally improve a process becomes patentable.

I guess it would tend to reward actual 'research' and investment rather than dashing of a patent on some half formed idea.

Reply on news.ycombinator.com