Hacker News new | past | comments | ask | show | jobs | submit login
Rotary Cellphone (justine-haupt.com)
571 points by nallerooth 6 days ago | hide | past | web | favorite | 164 comments

Looks like this got hugged to death. Archive link: https://web.archive.org/web/20200211231236/http://justine-ha...

European Union recommended emergency number is "112". Because you can morse it with the hook or even by short-circuitin the wires. This is also why the pulse-dialing receiver is operational in most EU countries.

Anecdote: A user at work was assigned extension 112. Works fine on the Cisco handsets we mainly use, but when users were trying to dial it directly from their VoIP apps on their cell phones, they often got dumped out to the phone emergency menu (depending on which VoIP app they're using).

We're in NA so it was a huge mystery as to why this was happening just with their extension, took a while before we realized what was going on.

For someone not familiar with morse or pulse dialing, how does one do that?

You click the handset down, like you are hanging up the phone, but just very briefly.

So think of the cradle switches - what "hangs up" the phone - as a button, and you just click that button (quickly) for '1' and then you pause and hit it twice in very rapid succession for '2' ...

You need to be on a POTS line to do this, of course, but it will work even today. Even if you have no dialing mechanism, you can dial the phone this way ...

> You click the handset down, like you are hanging up the phone, but just very briefly.

This brings back memories, this is a trick people my age group mastered as teenagers in a 3rd world country. All calls were billed - including local ones, so parents would commonly get lockable cages that would cover the buttons/rotary dial to prevent unauthorised calls in their absence. The cages would leave the receiver accessible even when locked in order to be able to receive calls, so we would use the hook to do pulse tone dialing; dialing a number with many zeroes was always a challenge since it required 10 rapid taps as part of a longer sequence, but I built up muscle memory.

The practice was colloquially named 'tap-dialing' and it was handed down from generation to generation. Sadly it came to an end in the smartphone world.

The old GTE payphones up in Georgetown, Tx (suburb of Austin, now) when I was a kid would let you dial with the switchhook - no coins required. Agreed, zeroes were a pain, but this was partially compensated for by the fact that 5-digit dialing worked for any local number, e.g. 3-1234 got you 863-1234. Interestingly, most VoIP adapters support the pulse dialing of an old phone, but very few have the cojones to actually ring one that has a physical bell rather than an electronic ringer...

Er, I'm not sure where you got this information from, but I can tell you categorically that ATAs that support pulse dialing are few and far between. Source: I have a vintage phone collection and wish bitterly the reality were different. (I use a Avaya PBX as a bridge to asterisk/IP telephony as a practical alternative)

That’s because the traditional physical bell needs a high voltage (relative to the digital circuits in modern hardware).


”The voltage at the subscribers end depends upon loop length and number of ringers attached to the line; it could be between 40 and 150 Volts.”

I think they also (relatively to the hardware in a VoIP adapter) need a large amount of power.

So, to ring that bell, you need to transform your 5V/3V/…, and you need a larger power circuit. That isn’t worth it for the few who still have an old-style telephone.

I think you need the high voltage anyway, even for a fake bell. The power is probably the problem.

I've seen some grandstreams that'll run a ringer.

Second this, I've had no trouble ringing an old GPO 746 telephone (and very loudly) with a Grandstream HT802

It was also an old school way of phreaking and getting free calls in a phone box.

I was very surprised a couple of weeks ago, making a call on a fixed-wireless 4G modem with a phone jack on it, that it supported pulse dialing. (The handset plugged into it happened to be set for pulse, you see.)

I wonder if something like that would get implemented to address a historical regulatory requirement.

Ah, thanks! I grew up using a rotary dial phone but of course was too young to think that much about how it actually works. Learned something new today!

It's been awhile since I tried it, but isn't it n+1 presses to dial n? So you can dial 0 with one click.

No; if you look at an old rotary phone, you're counting the number of finger holes in order to determine pulses.

To dial a 0, pulse the line 10 times.

Unless you are in New Zealand - the dial is numbered "backwards" there so 0 is 1 pulse etc.

Oh wow, thank you for the memories :) I grew up in New Zealand when everything was pulse dialing and had an American modem that my father got us, so in order to dial any phone number I had to rewrite it by subtracting every digit from 10.

It seems I remembered correctly. I'm from Sweden and we use the system I recalled. All of you correcting me are right with regards to most other countries.

"0" was coded with 10 pulses

No, 0 is ten pulses.

As I said, not in Sweden.

Tangent: many old Western Electric touch tone phones have a switch between "pulse" and "dial" for backwards compatibility.

Many digital cordless phones in the 2000's would still allow you to select between pulse and tone dialing. Pulse was fun to use just for kicks, except it added several seconds to the start of your call and most phones were not smart enough to switch the keypad back to tone dialing once the call was established (e.g. to operate an IVR).

Thanks for the information. I didn't know. A nitpick: this is not "morse" [code].

Yes it is. Spells "EEI". Cry for help or "expression of disapproval” in Chinese.

Note: You should NEVER say that you hate the phone company on an open land line. And please, be nice to the operator.


Thank you! Knowing that story is a great mnemonic to remember it, even if you have a push button phone!

I grew up in North America (where 911 is the emergency number), so when I moved to Germany and wanted to remember 112, I noticed that it spells out:

1 + 1 = 2

That would foil me if it was my only memory device. A+B=C is true of ~55 (napkin math) different 3-digit numbers. Plus another 40-ish for A×B=C and 50-ish for A-B=C...

But it’s A+A=B

The good news is, there’s not that many of those. 112, 224, 336, 448, 550 (its a stretch), and you’re likely to start at the beginning, therefore getting it on the first try.

Actually, I’ve heard that on cell phones in EU, you can still dial 911, and it’ll connect. I have had occasions to call 112 before, but it’s always been in big enough emergencies that I didn’t have time to experiment. I’m eagerly awaiting the day when I come across an emergency that’s urgent enough to call 112, but not urgent enough that it can’t wait a little bit for me to try something. It’s a delicate balance

In another recent HN thread, someone posted that due to the way emergency dialing is implemented in the cellular stack, dialing "911" will dial the local emergency number because "911", "112", etc. aren't treated as actual phone numbers, they just trigger the cellular network to bypass normal dialing sequence and connect you to the local emergency system.

(assuming it's A + A = B, there are still 4 sequences that could be if every number is an integer between 0 and 9, but if they remembered one of the numbers was a 1 it would drop the possibilities down to only 1 + 1 = 2)

Why not 111? That would have been even easier to manually-pulse dial.

In the good old days, they looked at 111 at an emergency number but it was rejected because it was too easy for the digit 1 to be tapped out by in high winds by wires rubbing together.

That's why the UK went with 999 as something memorable, but slow to dial.

An odd side effect of that was that in New Zealand, where phone dials are arranged 9-8-7-6-5-4-3-2-1-0 instead of 1-2-3-4-5-6-7-8-9-0, the emergency number was indeed 111, with nine pulses for the 1 digit. British-made exchange equipment was extensively used there, so no modification was needed to make 111 the emergency number.

The reverse dial was a result of the design of early Rotary exchanges built in Antwerp around 1921. Similar Rotary exchanges were installed in Oslo as well in 1914, and Oslo also used the reverse dial (but nowhere else in Norway).

AFAIK they changed it to 0118 999 881 999 119 725 3.

Sorry. Someone had to.

Try calling that number on your Android phone. It still "works" on my Essential phone with stock Android 10.

In the U.K., if the emergency number is a prefix, the call is routed to the emergency services.

* http://jdebp.uk./FGA/truths-about-telephone-numbers.html#Eme...

just take my upvote

Or for a line between telegraph poles to tap it out in the wind.

That's NHS non-emergency issues in the UK, and I presume that would have been a factor.

The NHS 111 number dates from 2009.

Because cats.

Because in the US, an initial “1” unambiguously means “this is the start of a long-distance number; a 3-digit area code follows”.

If I remember, some old payphones locked the rotary dial after 1 and 2.

There are so many of them:


In Europe many of them are variations of 1XX, but since there are so many and sometimes they change too, it's not trivial to remember which one.

The American 911 is more memorable, even in Europe, simply due to it appearing in popular culture so often, I hope that if someone dials that one in Europe it'll also get forwarded to a real emergency line.

>In Europe many of them are variations of 1XX, but since there are so many and sometimes they change too, it's just ridiculous to remember them.

Eh? 112 works in the whole EU. As your link says. Of course old legacy numbers also work in many countries.

It's almost as if Europe is not a country.

Just like the US, every country in has its own emergency numbers that you should probably know if you're in that country, but for Europe as a whole, 112 is the universal emergency number.

And while 911 might be known to many outside the US, as a number that requires actually navigating the dialing pad, it's not all that great. The UK's 999 or Europe's 112 are far better at being something you can dial while under duress (with 999, as a repeated number, being by far the superior choice here).

It is a trade off though. I’m sure “911” is no accident either. Because it is slightly harder to dial, it probably has far less accidental dialings than some other countries emergency numbers.

I bet little kids would constantly pick up a phone and dial “999999999”.... harder to mistakenly / playfully dial “911”


>Up until 1968, the United States was still using hundreds of seven-digit emergency telephone numbers across the country. In fact, some states, like Nebraska, had more than 180 different ambulance service numbers alone.

>The code 911 was chosen because it best fit the needs of all parties involved,” according to the National Emergency Number Association (NENA). “First, and most important, it met public requirements because it is brief, easily remembered, and can be dialed quickly. Second, because it is a unique number, never having been authorized as an office code, area code or service code, it best met the long range numbering plans and switching configurations of the telephone industry.”

911 is faster to dial, but 999 is easier to find in darkness /smoke

Not on a rotary phone: spinning to the first position is easier to find than the 9 (especially if trying to not go too far and hit the 0). On a traditional rotary 9-1-1 was one hard to dial number followed by the two easiest.

112 is the GSM standard, and works in countries all over the world: https://en.wikipedia.org/wiki/112_(emergency_telephone_numbe...

It forwards to 911 in the states? Good to know.

The problem with the "retro" telephone ring sound that so many modern phones have built in is that it just doesn't resonate and echo and slowly decay over time like a real phone ring. It would be cool to code up a physical ring simulator that reproduces that sound properly, and responds to the accelerometer, especially when you drop it or hang up on somebody by slamming it down.


Not being able to slam hang up on someone is a critically missing feature and due to such, I believe the entire smartphone revolution is a failure.

The biggest problem with telephones that still persists even today, is that you can only hang up on somebody once. Anybody who can solve that problem is bound to revolutionize communication as we know it.

Or the phone company doesn't immediately disconnect the caller, so you could pick the phone back up and slam it down again, at least a couple of more times.


Imagine that there was a phone receiver that delivered a zero to nominally painful electric current based on how abruptly the audio from the other speaker cut off... All sorts of reasons that this isn't doable, but the psychological satisfaction of knowing that spam caller just got zapped...

I'd love for Simone Giertz to design a crappy robot Hang Up Machine phone receiver, as effective as her Wake Up Machine alarm clock!


Yes, that is only slightly offset by being able to unfriend someone multiple times.

Just call 'em back and do it again!

When they bring a new Razr out you can snap it shut like a high powered executive looking out over Manhattan's skyline and maybe that will be cool enough.

When cellphones first came out, they eschewed classic ringers to sound modern and fresh. As soon as they started offering custom ringtones, I loaded up sounds from a classic TV and Movie sound effects library onto my Nokia. It included a Rotary Phone sound that became my standard ringer up until I started leaving my phone permanently on silent.

I can't find the sound fx library online anymore but it had all sorts of sound effects used in TV and movies from the 50s to today. I've heard the effects in Looney Toons cartoons, classic Scooby Doo, Mission Impossible, and Knight Rider just to name a few.

Unfortunately I didn't think to keep the entire library and so I only have a few sound files from it converted to MP3.

I always wondered if Apple deliberately tried to kill wacky ringtones by having every default ringtone suck except for the standard simulated landline phone ring so that everyone would just choose that one.

But these days everyone just keeps their phone on silent anyway...

I feel like that wasn't an option initially and that I was relieved when but also offended when everyone started using the fake phone sound.

Not EXACTLY what you're describing, but check out http://projectmf.org/

It needs a real bell.

For those interested, here's a Sparkfun project from 2005 on converting an actual rotary phone into a portable cellular phone:


You used to be able to buy them too: https://www.sparkfun.com/products/retired/286

> the battery lasts almost 24 hours

It seems like for hobby projects it's really hard to get good battery run times. I wonder why that is. Is is difficult to get everything into the appropriate sleep modes, or are hobbyist parts just that much more power hungry?

Commercial feature phones have stand by times measured in weeks.

Most performance optimizations take more time to implement than the actual goals. Hobby projects focus on fun proof of concepts, performance optimizations are ot very attractive (unless optimization itself is the goal).

The battery looks like it might be this one: https://www.adafruit.com/product/258 which is about half that of most smartphones. At 24h there's a best case power budget

The module appears to be https://simcom.ee/modules/wcdma-hspa/sim5320/ hardware data at https://simcom.ee/documents/SIM5320/SIM5320_Hardware_Design_...

... which claims a standby current of about 2.5ma. So either it's not correctly entering standby, or most of the power is going somewhere else. If the LED bargraph runs continuously that would explain it, but I assume she's thought of that and it's triggered by a button. Or doesn't mind the 24 hour battery life.

It takes a combination of software and hardware to get good power consumption. In this case, assuming the author has implemented sleep modes on the microcontroller, it's probably the 3g module being not very power efficient.

Assuming it has a comparable-size battery to other "feature phones", I suspect this could get better battery life with a bit of time spent on making sure everything goes to sleep whenever possible.

Having said that, hitting "more than a day of usage" is crucial, anything more than that is much less important for most users.

Yes, the hobbyist boards tend to include features that make the low power modes difficult to get into. For example, the nRF52 eval board datasheet mentions that they have to waste 20uA of power to provide a reset button. Things like reverse-polarity protection diodes use power. You want those in your hobbyist board, but you probably want the extra year of life on your CR2032 in production.

In other cases, the boards aim to be small instead of complete, so just use a simple LDO to power the device from USB. This uses a ton of power even if the MCU is off, and that's where the battery drain goes. Designs for real products tend to use a DC-DC converter that can be tuned for the desired sleep state.

Finally, the software that hobbyists use is basically designed to be easy, not efficient. You don't write any MCU init code, it just turns all power-sucking peripherals on by default. That way when you go to use them, they work! Accessing MCU-specific features to save power tends to be difficult, because the hobbyist stuff is designed to support every MCU ever designed from decades-old AVR chips to the latest ARM cores. If you're willing to include a bunch of headers and hack around what they do for you, you likely break core things like "read a GPIO pin" or delay_ms(). All in all, it seems unlikely to me that anyone is going to buy an Adafruit board, program it with Arduino, and get a low power device. I tried and failed. Adafruit themselves recommend using an external device to turn the board on and off at a scheduled interval to save power.

(I made an interesting nRF52 clock / temperature sensor with an e-ink display a few months ago. Every minute, the RTC interrupts the MCU to wake up, refresh the display, and send the temperature/humidity data out over BLE. I used Arduino and ... never got the power usage down at all. It lasted for A DAY with a 100mAh Li-ion battery, even with the radio disabled completely. The voltage regulator just burns power while the MCU is asleep waiting for that minute interrupt. As it happens, it is super jarring when the e-ink display refreshes itself every minute, so I ultimately abandoned the product.

Also somewhere in there I realized that using Arduino is a giant waste of time. When you just write your own C code, all of the manufacturer's documentation suddenly becomes applicable, you can use fancy things like an RTOS, and you even get code completion with clangd in your editor. Every time someone mentions microcontrollers I get mad at myself for wasting so many years with hobbyist-grade tools that never worked right. But I digress.)

An interesting exercise is comparing the nRF52 reference design (deep inside a ZIP file from here, https://www.nordicsemi.com/Software-and-Tools/Development-Ki...) with something like Adafruit's nRF52 dev board: https://cdn-learn.adafruit.com/assets/assets/000/052/793/ori... You will notice that Nordic's dev kit has a very complicated power stage, with solder bridges to turn off creature comforts for better power efficiency. Adafruit just has a random linear regulator on there.

On a slightly more practical note, if you want a rotary VoIP desk phone, there's a few models of Grandstream ATAs that understand pulse dialing. They can be registered to Asterisk as normal SIP clients. $20 for the ATA plus $20 for the phone, plus your time to do the software configuration.


In the 80's I had a Demon Dialer that stored a whole bunch of phone numbers and could automatically dial them with pulse or touch tone dialing, and switch between them for using alternative long distance services like Sprint, MCI, ITT, etc. Great for getting through to busy modems, radio contests, or just using as an electric phone book, etc.

Demon Dialer 176T (the high-end model that stores 176 numbers!)



>In the computer hacking scene of the 1980s, demon dialing was a technique by which a computer was used to repeatedly dial a number (usually to a crowded modem pool) in an attempt to gain access immediately after another user had hung up. The expansion of accessible internet service provider connectivity since that time has rendered the practice more or less obsolete.

>A similar technique was sometimes used to get the first call for prizes in radio "call-in" shows, thus leading to the adoption of random "fifth caller," "seventeenth caller" etc. by radio stations to circumvent this practice.

>The term "demon dialing" derives from the Demon Dialer product from Zoom Telephonics, Inc., a telephone device produced in the 1980s which repeatedly dialed busy telephone numbers under control of an extension phone.

>"Demon dialing" was popularized by the movie WarGames, which demonstrates the hacking technique (the technique eventually evolves into a technique called "phreaking"). After giving the program an area code and 3 digit prefix, the program then serially dialed every phone number in that prefix in order, recording which phone numbers were answered by a computer modem. Soon after, most likely because it was popularized by the movie, serial dialing was outlawed. Hackers got around this by simply randomizing the order the program dialed all the numbers. Later the terms "demon dialing" and "war dialing" became synonymous.


user manual for some DOS wardialing software:


> On a slightly more practical note

Why "more practical"? How is a desk phone more practical than a portable one?

You can’t drop it in the toilet.

In the days of wired handsets with really long cables, you'd be surprised at what could happen to the handsets, and where people took telephone calls. (-:

You can answer it for incoming calls like a normal office voip phone, without the need to worry about limited battery life or paying for a live SIM card just to have a toy to play around with. For people who already have their own SIP based IP PBX the cost of adding another phone is zero.

For outbound calls if you encounter an IVR menu (press 1 for english, etc), the ATA will translate pulse dialing into DTMF tones. Functionality I think this example is lacking.

They don't break - or at least they don't break under normal abuse. You can break them if you try hard enough, but your first try will probably fail.

...and when they do, they're very easily repaired --- at least the early, pre-electronics ones.

it becomes impossible (or at least rather challenging) to accidentally butt-dial somebody?

Very cool!

i concur with the whole functional simplicity thing.

But I think the dial is anachronistic, despite what's claimed; A bit like fitting your car with reigns for steering and a whip for acceleration.

The only 'functional' aspect of it is perhaps that it slows down your dialing enough (especially when your finger slips on a number, and you have to start over!), for you to think whether you really need to make that call.

I knew an early mobile phone adopter in the 80's who made a converter plug for his car phone, and plugged in an old black classical AT&T dial phone that he could make and receive calls on. Every time he drove over a bump with the car, the bell on the phone would ding.

The article makes no claim that it is not anachronistic. What it says, rather, is that that is not the point.


And I agree it's fun and a cool project, regardless of utility

It's fun, which is its own kind of "functional".

You could say it puts the fun in functional.

I'll see myself out.

That's a recursive portmanteau!

> A bit like fitting your car with reigns for steering and a whip for acceleration.

I just spent far too much time pondering the feasibility of a reign / whip controlled car. The trickiest part might be designing the filter to detect the crack of a whip reliably. Also wondering if there are any laws regarding the form and location of steering and acceleration mechanisms.

I want all of these features in a phone without the rotary part:

* Real, removable antenna with an SMA connector. Receptions is excellent, and if I really want to I could always attach a directional antenna.

* When I want a phone I don't have to navigate through menus to get to the phone "application". That's bullshit.

* If I want to call my husband, I can do so by pressing a single dedicated physical key which is dediated to him. No menus. The point isn't to use the rotary dial every single time I want to make a call, which would get tiresome for daily use. The people I call most often are stored, and if I have to dial a new number, or do something like set the volume, then I can use the fun and satisfying-to-use rotary dial.

* Nearlt instantaneous, high resolution of signal strength and battery level. No signal metering lag, and my LED bargraph gives 10 increments of resolution instead of just 4.

* The ePaper display is bistatic, meaning it doesn't take any energy to display a fixed message.

* When I want to change something about the phones behavior, I just do it.

* The power switch is an actual slide switch. No holding down a stupid button to make it turn off and not being sure it really is turning off or what.


I use an unmodified rotary phone as my main landline phone. In the UK it would seem that our exchanges still support pulse dialling. The only drawback are menu systems that expect to hear tones, but I get around this with a small app on my mobile phone that just plays DTMF tones. If i want to, i can pick up the receiver and play tones down it to dial a number rather than using the... dial.

Hand-held DTMF tone generators, for exactly this purpose, were once fairly widespread. A financial institution once sent one to me for navigating its dial-up automated teller system.

And before that less widespread, but in circulation [eg 0] for "free" long distance phone calls.

[0] https://en.wikipedia.org/wiki/Blue_box

You didn’t insinuate this, but to be clear, a blue box didn’t emit “touch tones.”

Yeah, i looked for something like that to keep next to the phone but couldn't find anything. I was sure they used to be a thing. What you _can_ get however are little kits you wire up inside the pulse dial phone which do an on the fly pulse to DTMF conversion.

I had one early century about the size of a car key. Googling for something similar now I find nothing. Not necessarily weird given the current state of telephony, but kind of a bummer.

You hold your mobile phone up to your rotary phone in order to play the tones? Why not just use the mobile phone to make the call?

Well, I don't. I rarely use the landline. I'm just saying that you can if you want to.

My mom refused to get touch-tone service, in the hopes of preventing me from becoming a phone phreak. But I had my touch-tone-enabled friends touch-tone me MCI codes and phone numbers I wanted to call over the phone, and recorded them on a cassette tape recorder, which I could then play back, with the cassette player's mic and speaker cable wired directly into the phone speaker and mic.

Finally there was one long distance service that used speech recognition to dial numbers! It would repeat groups of 3 or 4 digits you spoke, and ask you to verify they were correct with yes or no. If you said no, it would speak each digit back and ask you to verify it: Was the first number 7? ...

The most satisfying way I ever made a free phone call was at the expense of Bell Communications Research (who were up to their ears swimming in as much free phone service as they possibly could give away, so it didn't hurt anyone -- and it was actually with their explicitly spoken consent), and was due to in-band signaling of billing authorization:

Peter Langston (working at Bellcore) created and wrote a classic 1985 Usenix paper about "Eedie & Eddie", whose phone number still rings a bell (in my head at least, since I called it so often):




>(201) 644-2332 or Eedie & Eddie on the Wire: An Experiment in Music Generation. Peter S Langston. Bell communications Research, Morristown, New Jersey.

>ABSTRACT: At Bell Communications Research a set of programs running on loosely coupled Unix systems equipped with unusual peripherals forms a setting in which ideas about music may be "aired". This paper describes the hardware and software components of a short automated music concert that is available through the public switched telephone network. Three methods of algorithmic music generation are described.

Eedie & Eddie (And The Reggaebots) - Some Velvet Morning (Peter Langston)



>At the time I was also running a telephone demo that you could call up to hear samples of music composed for you while-u-wait by the computers in my lab. Most people called the mono telephone number that gave you a monophonic mix of the stereo channels, but you could also call either the left or right channel, or both, one phone on each ear, to get stereo -- this was how I worked from home. The elaborate phone connections were from the telephone switch in Brian Redman's lab.

>When the call would come in, the computers in my lab would start composing the 6 or 7 pieces that were played in the demo, while a voice synthesizer in Brian's lab (Eedie) would accept the call (we paid the long distance on a few collect calls from music researchers in other parts of the world) and introduce the demo, explaining what was going to happen, then the call would be switched over to my lab where another voice synthesizer (Eddie) would take over, introducing each piece as it was played.

>The idea for Eddie's voice came from the overly-happy shipboard computer in the Hitchhiker's Guide to the Galaxy radio serial. In the phone demo Eddie would say something like:

>Hi! This is Eddie! Your computer music phone, FRIEND!

>Oh BOY! You are rilly rilly going to like this music, I can TELL!


>First, I'll just make sure that Peter's computer is working, and then we can share-and-ENJOY!

>This is pretty much the kind of thing that Douglas Adams' computer said, and he was named Eddie, so we used Eddie as an homage. Then, when we wanted the voice-synth that answers the phone in Brian's lab to be female, we considered how, later in the Hitchhiker's Guide, Zaphod Beeblebrox chooses a different personality module for the computer and gets a worried old lady's voice -- "This will all end in tears!" but Adams never gives her a name, so I chose "Edie" as a feminized "Eddie". Unfortunately, the DECtalk voice synthesizer we were using pronounced E-d-i-e just like E-d-d-i-e with a short E vowel "eh" at the start, so we had to spell her name E-e-d-i-e in order to get the long E initial vowel sound.

>Anyway, the little chit-chat between Eedie and Eddie at the beginning of the recording of Some Velvet Morning is typical of the sillier moments in the phone demo introductions, and is included as an attempt to familiarize the listener with their strange accents.

When you called (201) 644-2332, the Dectalk speech synthesizer would answer, say "Hi! This is Eddie! Your computer music phone, FRIEND!", wait a few seconds, then cheerfully and sincerely announce, "Yes, operator. I will accept the charges."


LGR Oddware: $1,200 DECtalk PC Speech Synthesizer


Then Eddie would launch into his greeting and introduction to the phone tree of cool demos, like singing songs, telling jokes, generating improvisational music with L-systems or Stochastic Binary Subdivision, playing games, reading news summaries and weather reports, etc.

But then some "music researcher in other parts of the world" realized that it was perfect for making operator assisted third party collect phone calls!

Did you ever trick the phone company into cheerfully accepting the charges of a third party phone call? You will!


>A third number call or third party call is an operator assisted telephone call that can be billed to the party other than the calling and called party. The operator calls the third number for the party to accept the charges before the call can proceed.

So you could just dial 0, tell the human operator that you want to call any number you wanted, and charge it to (201) 644-2332. Then they would call that Morristown, New Jersey number for authorization, Eddie would answer, the human would ask "Will you accept the charges for a collect call from ... to ...?", Eddie would then cheerfully announce, "Yes, operator, I will accept the charges!", and the human would say "thank you" and hang up on the loosely coupled network of Unix systems with unusual peripherals at Bell Labs, then connect you through for free to whoever in the world you wanted to talk with, for however long you want to talk, charging the call to Bellcore!

"Share and enjoy!"


Why do you stick with the rotary phone?

Also, isn't it easier to have a second landline phone attached vs having to play DTMF tones from a phone app?

It's just a bit of fun. It's more ornamental than anything. It's just nice that it also works as a phone although I rarely ever actually use the landline.

If it works don't break it. Those old phones could stand nearly any abuse unlike modern phones that break if you look at them wrong...

This is great, I especially love the form factor. I did the same a few years ago but kept the original enclosure:


It was a very fun project and I learned a lot about electronics doing it!

I dunno, I like the iRotary form factor. Both are impractical, and yours leans into the impracticality.

in order to one up this very cool project one would have to: - create a cellphone in a Telephone magneto style from before there were rotaries - train a machine learning algorythm to emulate the behavior of a Switchboard operator so you can have the truely "authentic" old school behavior

I did something like this as a joke at a company I used to work for. Asterisk, a lot of wacky perl, and CMUSphinx. You'd pick up one of the rotary phones in the office[1] and say "operator", wait for a chirp, then you could either speak the number or say the persons name you wanted to call.

[1]: https://i.imgur.com/O7GLfsa.jpg

You could define a class called RingyDingyEnumerator!


Speaking of "old school", I love watching Joan Rivers shamelessly plug MCI, after which she audaciously asks Ernestine how old she is and grills her about her sexuality:



> In 1970, AT&T offered Tomlin $500,000 to play her character Ernestine in a commercial, but she declined, saying it would compromise her artistic integrity. In 1976, she appeared on Saturday Night Live as Ernestine in a Ma Bell advertisement parody in which she proclaimed, "We don't care, we don't have to...we're the phone company." The character later made a guest appearance at The Superhighway Summit at UCLA on January 11, 1994, interrupting a speech being given on the information superhighway by then-Vice President Al Gore. She appeared as three of her minor characters in a 1998 ad campaign for Fidelity Investments that did not include Ernestine or Edith Ann. In 2003, she made two commercials as an "updated" Ernestine for WebEx.

phone company sketch - we dont care - Tomlin


Ernestine's House Call - Saturday Night Live


Ernestine the telephone operator calls General Motors


Lily Tomlin as Ernestine the Telephone Operator



And this is a gem if there ever was one: “The point isn't to use the rotary dial every single time I want to make a call, which would get tiresome for daily use.“

when I started visiting this site almost 8 years ago this was the kind of content I wanted (and expected) to see more of. tbh it's been a little disappointing how silicon valley tech industry everything has been focused.

I don't understand how one would dial a 1 on that dial. There doesn't seem to be any room for it to travel.

I'm thinking of the same thing, looking at the pictures of old phones, there should be space reserved between the "1" hole and the metal stopper.

This is seriously cool! The device has an amazing set of features and functionality.

I am a bit surprised about the short battery time. I would expect that, without a display, the battery would last several weeks under normal usage.

I don't see it listing the battery capacity anywhere so it might not be very big. Also, proper power saving design is a bit more complicated than you might think, it takes a lot of careful attention to detail to take a mobile device from hours to days of battery life.

The headlined article states that it was taken from 2 hours to 1 day already.

Yeah, but that's from "first somewhat functional prototype" which probably didn't make any special effort to conserve power, to "rev A" which I'd guess does all the obvious things and is all you're going to get without significant effort. You don't get "several weeks" out of a phone without some wizardry or a humungous battery.

On my first Android smartphone, an LG Optimus One P500, I had no trouble to reach 4-5 days and the battery was far from humungous, only 1500 mAh. Most things were not running but I was using it to text quite often during the day. I don't believe there was any wizardry on there.

We see the battery on a picture, if it comes from Adafruit, that seems like the 2500mAh one they sell. There's no reasons that thing can't last at least a week, but I can understands it wasn't the goal at all to make it last, so almost no effort was invested in that.

> I don't believe there was any wizardry on there.

A lot of that wizardry is not really user-visible. Ensuring that chips spend more time in sleep mode, coalescing events so that wakeups happen together, ensuring that there are no components passively draining power, etc.

For me, wizardry would be to go beyond what any other manufacturer would do. No one is considering Android phones to be optimizing specifically toward battery life and that phone was the cheapest of the cheapest, I paid 50$ for it, without any contract in 2012.

The problem is that this wizardry is not usually user visible, and doesn't have much of a correlation with the device price or operating system.

If anything, it's even possible that the cheapest phone has the most engineering done on it to optimize the battery life, due to the fact that reducing the battery size (and resulting cheaper BOM) on a cheap device that has millions of units sold is worth the additional engineering time to reduce power consumption.

World's pretty big. A one-in-a-million engineer is pretty damn special compared to the rest of us, and there's 7000 of them out there. Cheap-per-unit hardware can have paradoxically high development budgets since volume is expected to be so high. We probably all have slightly unrealistic how hard it is to design even a "crap" piece of equipment.

It's like how non-software people can't fathom that it costs $50k to build a relatively simple phone app when MS Office only costs a couple of hundred bucks.

There's a paper display on the back. This is amazing.

The display takes next to no power when not changing.

The problem here is potentially a software bug. The cellphone radio probably is not getting put to sleep as long as it could be, and that's about 100 mWh.

Given the choice I'd much prefer a device which texts but doesn't have a "phone" than the reverse. The idea of a phone which doesn't text just sounds frustrating.

That is awesome. I remember dialing our beige rotary phone for my friend when he came over after school... Haha he was afraid to dial it for about a year.

(Scared of looking dumb, I guess) But aside from choking on our tears and those shameless bellyaches, we didn't give him any grief. This would have been probably 1994 - 2000... when my folks upgraded to touch tone and answering machine :)

This would be a great time to see this obituary of John Karlin, one of the early phone UX pioneers:


The link for the design files in the story is broken, but the files are up at this link instead:


Justine and I have a mutual friend. I sent him this link and he talked to her and she said she's been overwhelmed by the attention it's getting. I saw it on the Adafruit blog and I'm wondering if that's what started the avalanche.

It's cool project!

What I dont understand is why they just didnt use the trimline case as the case for the thing - or at least the blueprint for the case - this thing looks uncomfortable to hold for any length of time.

Maybe “uncomfortable to hold” is part of the point. This is basically an art project, after all.

Weird second time I've seen this this morning. https://news.ycombinator.com/item?id=22303956

I'd personally do it with the whole enclosure and handset intact. You could fit a pretty decent battery in there... to operate the bell.

She should put this up on a crowdfubd site like crowdsupply to mass manufacture. I for one would buy!

So how long until someone ports Doom to it?

Awesome, I will have to share that with my former colleagues from Martelsham.

24hrs battery?!?! For a low tech solution with no draining high res touchscreen and all the other battery hungry features of a modern smartphone this is shockingly poor. I thought it was going to be 2 weeks!

I would totally fund this on Kickstarter.

how about a phone where you just have a clicker to put in the number in binary?

Make it double as the hook, and you have just reinvented hookswitch dialling for the mobile telephone minded. (-:

or unary :D

The firebrigade and police not even a thousand clicks away.

i really like this, well done

Calling this open source is a bit of a stretch. At the heart of this phone there is a huge proprietary dependency (the Adafruit FONA module).

It's hard to get any open source software finished, what with all the goat farming.


I was about to post the same link

It goes opensource in my book maybe you will create an opensource alternative for the hardware. This sort of gatekeeping is maybe the reason why hobbyists avoid HN.

Huh? I didn't say that wasn't a cool project. On the contrary, I think this is a great project even if it was all closed source.

Maybe I should have clarified that I just meant to highlight the fact it's almost impossible to understand what many of the devices around us are doing because of the dependency on fundamental building blocks which are closed source and have no open source alternatives.

I'm sorry if my comment came out offensive in any way. That wasn't my intention.

I understand if that wasn’t your intention but i saw a cool project on HN and you gatekeeping. It seems the author open sourced everything they knew. Your comment didn’t mention any alternatives to open hardware or such. It felt condescending, something cool and hacked up.

I didn’t like the tone, hence the comment.

I think your comment is pretty fair and not an indictment of the project. It's 100% worth mentioning.

The problem with these open source phone projects is that all the modems use closed source drivers (I think).

Do you know a path forward to do this entirely with open source?

Unfortunately I don't know of any project in the area of open source basebands. There was OsmocomBB for GSM, which used the TI Calypso chipset, but I'm not aware of any such efforts for 3G/4G.

This module is basically just a 3G baseband chip on a board. By your definition, no circuit that contains a chip would be open source.

"just a 3G baseband chip" does hide an enormous amount of complexity. The point is just that this design contains both non-open-source hardware and non-open-source software (running on the 3G baseband chip). It's still mighty cool!

It's very cool indeed, even if it was all closed source :)

Except it's not "just" a 3G baseband chip. That chip[0] runs a closed source software that even contains a Lua interpreter. That's a proprietary binary blob that I don't know what it does, so this is not something that I would call "open source" even if the controller is open source.

[0] https://simcom.ee/modules/wcdma-hspa/sim5320/

Boy howdy, you are not going to like to hear what happens when those radio signals leave your cellular telephone.

Applications are open for YC Summer 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact