Warren Zevon appeared on The Late Show with David Letterman in October of 2002.

I drew my remembrance of the appearance below. (Apologies to Dave, who has been rendered here in a way that kind of makes him look like Richard Nixon. This was not intentional, but here we are. I’m still learning.)

Zevon had appeared on Letterman’s show many times through the years …

… but this time was different. Zevon knew he was going to die.

He was walking through the final days of a diagnosis with terminal cancer.

What could’ve been deeply unsettling instead became something deeply compelling. Something that has remained with me for years. The two talked about music and family, but also life and death. It was all somehow beautiful and human and … also funny and incredibly moving. All at once.

This part of their exchange has stuck with me probably every day since:

Dave: From your perpsective now, do you know something about life and death that maybe I don’t know now?
Warren: Not unless I know how much … how much you’re supposed to enjoy every sandwich.

Warren:
You put more value on every minute. You do. I mean I always thought I kind of did that. I really always enjoyed myself, but it’s more valuable now. You’re reminded to enjoy every sandwich and every minute of playing with the guys and being with the kids and everything.

I’d just finished re-reading (this time listening) to Jason Zinoman’s book on Letterman. My Youtube algorithm has been, wonderfully, feeding me classic Letterman clips. Last week, this clip snuck up on me from a different angle (and this time with an actual sandwich in my hand).

During a quick layover on my way home from a speaking engagement, I was sitting in a terminal. Traveling from Canada to Nashville. Happy to have had the work. Grateful for the experience. Anxious about upcoming work. Tired. Worried about all sorts of things. Laughing at a picture Kristi sent me from the kids. Hungry.

I bought a sandwich.

It was not a great sandwich.

Soggy and overpriced and maybe “food’“? (not sure) … it was what was available.

I thought of Warren’s words. It hit me that maybe he didn’t necessarily mean every sandwich is going to actually be great. Some sandwiches are bad. They just are. Some days will be a challenge, too. There will be disappointments. Hurts. Things that simply aren’t at all what we hoped they would be.

I don’t think the answer is to pretend everything is wonderful.

Maybe I’ve been hearing Zevon’s words a little differently lately. Against the backdrop of lots of good, but so many uncertainties and challenges, I find myself forgetting the point.

The goal might not be to exactly enjoy every sandwich. Some sandwiches are just straight up trash. Maybe the goal is to notice that you’re eating one.

That’s enough.

Life is strange. Twenty-four years ago a dying rock-and-roll star said something weird and funny on television. Last week I was in an aiprort eating a bad sandwich wondering if maybe I’ve finally understood what he meant.

Take a bite. Taste it. You’re here.

1. hire me to speak.
2. buy a book.
3. send this to a friend.

In my 2023 article on the evolution of the calculator, I griped about the way we teach the history of computers. I felt that we’re giving far too much credit to Charles Babbage. We have known how to make mechanical calculators at least since the 17th century; the concept of programmable machinery predated the English polymath’s writings by a wide margin, too. The remaining bottleneck was mostly technological: past a very modest scale, sprockets and cams made it too cumbersome to shuffle data back and forth inside a sufficiently complex computing machine.

The breakthrough was an electronically-controlled data register. These circuits allowed computation results to be locked in place and then effortlessly routed where needed, no matter how complex and sprawling the overall design. In contrast to the musings of Mr. Babbage, the history of this “working memory” has gotten less attention than it deserves. We conflate it with the evolution of bulk data storage and computer I/O and keep forgetting that RAM constraints were the most significant drag on personal computing until the final years of the 20th century.

The bistable cell

The first practical form of electronic memory was a relay. A relay is essentially a mechanical switch actuated by an electromagnet. Its switching action can be momentary — lasting only as long as the electromagnet is powered — or it can be sustained, latching in the “on” state until a separate “reset” coil is energized.

Latching relays can be used to store bits in a fairly obvious way; a momentary relay can also be wired to latch by having the relay supply its own power once the contacts are closed. A simple illustration of this principle is shown below:

In this circuit, pressing the “set” button will energize the relay, therefore also closing the parallel relay-actuated switch. The current will continue flowing through this path even after the “set” button is released. To unlatch the circuit, the normally-closed (NC) “reset” button must be pressed to interrupt the flow of electricity.

Relay-based main memory featured prominently in what was probably the first true electronic computer, Konrad Zuse’s Z3:

Electromechanical relays have switching speeds measured in tens of hertz; Mr. Zuse understood from the get go that the technology is a dead end. He took the expedient path in hopes of securing longer-term funding from the Nazi regime. Unfortunately for him — and fortunately for much of Europe — the officials were unimpressed and the funding never came through.

Zuse’s counterparts on the Allied side had more luck: early successes in codebreaking allowed them to pursue more costly and complicated vacuum tube designs. The blueprint for a tube-based memory cell came from William Eccles and Frank Wilfred Jordan — a largely-forgotten duo of British inventors who proposed the following bistable circuit as a replacement for an electromechanical relay:

The circuit latches when an input current flows through the transformer winding on the left. This momentarily makes the G1 grid voltage more positive, upsetting the current balance between the tubes in a self-reinforcing feedback loop.

In modern terms, both of these circuits would be described as set-reset (SR) latches. Their fundamental operation is shared with the cells that make up high-speed, low-power SRAM memories found on the dies of most microcontrollers and microprocessors today. A contemporary textbook example of an SR architecture could be this:

To analyze the circuit, let’s consider what happens if the “reset” line is high. In this scenario, one input of the AND gate is always zero, so the circuit outputs “0” no matter what’s happening on the OR side.

Conversely, if “reset” is low, the AND gate works as a pass-through for the OR. As for the OR itself, if “set” is high, it unconditionally outputs “1”. But if “set” is not asserted, the OR passes through the circuit’s previous output value via the feedback loop on top of the schematic. In effect, at S = R = 0, the device is latched and stores a single bit of data representing what previously appeared on the S and R input lines.

We could wrap up here, except for one tiny detail: most vacuum tube computers did not use vacuum tube memory for the bulk of their needs. The reason is clear if we consider that the memory unit in Zuse’s Z3 required about 2,000 relays, compared to 600 for the rest of the machine. The math for the relatively costly tubes was even worse, and that’s before we consider that each tube needed substantial, constant power for the heater filament. In short, until the advent of large-scale integration (LSI) chips in the mid-1970s, main memory usually had to be done some other way.

The merry-go-round

The solution to this problem were “dynamic” memories that relied on relatively short-lived physical phenomena — say, static electricity or acoustics — to store information in an inexpensive medium. Because the underlying physical process lasted only a brief while, the data had to be regularly retrieved and refreshed (or retransmitted) again.

One of the earliest examples of this was the Atanasoff-Berry “computer” (ABC) — in reality, a sophisticated but non-programmable calculator that used a spinning drum with an array of mechanically-switched capacitors:

The charge state of each capacitor conveyed digital information; the gradual self-discharge of the cells necessitated regular refresh. This was, in many respects, the precursor to modern-day DRAM. The DRAM chips in our desktop computers and smartphones still use capacitors; we just make them in bulk on a silicon die and have a way to address them electronically. Although this type of memory is slower and more power-hungry than SRAM, the enduring advantage of DRAM is that it requires far fewer components — and thus, less die space — per every bit stored.

But back to early computing: another fascinating incarnation of the idea were delay-line memories. These devices encoded data as sound waves pumped into a length of material and then sensed on the other side. Because sound takes time to propagate, some number of bits could be “stored” in flight in an electro-acoustic loop. The main advantage was that sound waves propagate pretty briskly, so the latency of this system was far better than that of a spinning capacitor drum.

The wackiest example of this genre is the mercury delay line. As the name implies, these devices relied on the toxic metal as the medium for sound waves:

Wikipedia will tell you that mercury was chosen because it uniquely matched the acoustic impedance of piezoelectric transducers, but I think that’s a misreading of the original patent (US2629827A). The problem with solid metal was presumably just that the longitudinal speed of sound was much higher than in liquids — so at a given clock speed and column length, you couldn’t store as many bits. Conversely, the issue with less dense and more compressible liquids might have been that they attenuated high frequencies, reducing the maximum transmit rate. All in all, I suspect that mercury simply offered the optimal storage capacity for the era — as long as you didn’t care about safety or weight.

Thankfully, this “mad scientist” period ended the moment someone thought of coiling the metal to accommodate more data in flight. Meet torsion memory:

The approach relied on acoustic waves traveling through spring wire, but this time, it used torsion (twisting) instead of longitudinal (push-pull) motion. This allowed the wire to be loosely coiled and held in place with rubber spacers; with longitudinal waves, much of the acoustic energy would be lost at every turn.

Yet another fascinating class of early dynamic memories included the Williams tube and the RCA Selectron tube. Both had no moving parts; the latter device is pictured below:

These memories operated somewhat like a CRT, selectively firing electrons at a phosphor substrate and storing several hundred bits as a pattern of illuminated dots. Although the stored data was visible to the naked eye, the readout relied on sensing changes to electrostatic fields instead. In essence, electrons were getting knocked around in the phosphor layer, creating a net charge which could be picked up for a while by a nearby electrode.

The magnetic era

Today, even among computer history buffs, magnetic media is typically associated with bulk, at-rest data storage: hard disks, floppies, and LTO tapes. But for about two decades — roughly between 1950 and 1970 — magnetic main memories were quite common too.

The earliest incarnation of this is the magnetic drum memory, a sort of an in-between stage between the capacitor-based memory of the Atanasoff-Berry computer and the modern hard drive:

In this design, read and write heads hovered above a fast-spinning ferromagnetic cylinder. The relatively small diameter of the cylinder, along with the high number of parallel heads that didn’t need to make contact with the surface, kept the read times reasonably short. But still not short enough: the computers of that era were already running at multi-megahertz speeds. We needed a better approach.

The best we came up with in pre-IC era was solid-state magnetic core memory. These devices consisted of an array of microscopic ferrite beads (“cores”) with a pattern of insulated copper wires woven through:

The key property of the cores is that they could be magnetized by an external field, but their magnetic polarity flipped only if the magnetizing field exceeded a certain threshold. Below that threshold, they retained their previous state.

To write to a specific cell, a sub-threshold magnetic field would be induced around a specific horizontal wire by supplying a calibrated current through its length. Another sub-threshold field would be created along a vertical wire. At the intersection of the two wires, the intensity of the combined fields would exceed the magnetization threshold and “flip” the selected core.

The readout process was a bit more involved: it relied on overwriting a cell and then sensing the induced current that’s expected if the polarity flips. If no such current pulse was detected, it meant that the written value must have been the same as what was stored in the cell before.

Magnetic memories were fast and quite tiny; the microscope photo above shows about 63 bits of a 32 kB module that measured 8×8” overall.

Interestingly, the legacy of magnetic RAM lives on: a company called Everspin Technologies sells magnetoresistive (MRAM) chips for embedded applications in sizes up to 16 MB. The chips behave just like SRAM, except they retain data with no power. At the moment, the tech doesn’t seem to be particularly competitive; I think it might have some uses in aerospace and defense applications because it is more resistant to radiation and EMP. That said, given the meandering path we’ve taken with memory technology, so knows what the future holds?

Other history-related articles you might like:

An illustrated history of the calculator

lcamtuf

·

August 26, 2023

Read full story

Random objects: Ball’s patent anti-garrotter

lcamtuf

·

December 10, 2022

Read full story

Random objects: the Petal camera

lcamtuf

·

January 20, 2023

Read full story

I write well-researched, original articles about geek culture, computing history, and more. If you like the content, please subscribe.

One problem that I’ve had in physics is remembering physical laws. I want to be able to use physical law in the same way that I can fry an egg, string together a sentence, or add up numbers. I think the world ‘fluency’, comes to mind. Imagine manipulating the physical world as fluently as one can communicate ideas – ‘oh my chicken gets overcooked, that’s fine, I know how to solve the heat equation so I just need to reduce the temperature by X and make my chicken a bit more spherical next time!’.

Sure, I could just look up Coulomb’s law or Maxwell’s equations, or ask an LLM, but I’m always amazed that the Greeks managed to learn so much about the world with much less tools. And I kinda want to be able to do the same. One time my friend Cyrus and I we were sitting on the beach in Brighton and then challenged each other to try to compute the radius of the Earth from first principles using the angle of the horizon. I think these types of challenges are both fun and revealing.

There is piece of advice in one of Shankar’s texts that stuck with me whilst I was chilling by the pool in Malaysia, from his textbook ‘On Fundamental Physics II’. Shankar writes that to get deep understanding of physics, one should think hard about how physical law could actually be tested and measured with real life instrumentation. Lest, we might get lost in ‘factors of 2pi’ just doing symbolic manipulations like the Cambridge Masters theoretical physics student I was.

I’ve always had trouble trying to remember something like Coulombs law, so I had a bit of a think about how I could feel it and test it with my own hands. After all, Coulomb’s law deals with point charges. And those are abstract ideas, it’s not straightforward to just magic up a piece of electric charge. One starting point might be to use something like a pen and rub it against your hair, but

Coulombs law is meant to tell us the force between two charged particles. If we have two charged particles q1 and q2, then observations like two charged balloons tells us that there should be a force between the two charged particles. Coulomb’s law quantifies this force.

It states that if the distance between the charged particles is x, then the force between them should be

Ok but let’s think critically about this equation to see if there are any new insights to become of it. How could one have come to a reasonable conclusion that this is indeed the law of force between charges? Presumably, the law is a function of distance between the charges, and the charge on both. So we have the force being a function of the following

To help us slim down what this function might be, we might ask what happens when one of the bodies are not charged. If one isn’t charged, then intuition says that there should be no forces in between the two things. This means that the charges should enter the equation as multiplicative factors. This would imply then that the force law looks like below, where a and b are some constants

But then by symmetry, we require a and b to be the same. So the force should look like

But it’s not at all obvious to me why the exponent a is 1 in nature. I think if I tried to write about a physical rabbit hole is out of scope but I will keep thinking about it.

The next question then is then figuring out the force varies as the distance between the two increases. A logical guess would be that the force law might obey something similar to the inverse square law observed in the intensity of light, or the gravitational constant. For example, Newton’s law of gravity reads that the force between two bodies is

And so then the question is – how would you check that Coulomb’s law indeed obeys inverse square law? It’s not entirely clear. And in fact, a few scientists in the 16th century actually proposed that the original Coulomb’s law was close to inverse square but not quite. To do this at home, you would probably need to figure out how to charge two electric balls with the same level of charge, hold one fixed, hold the other one by a string, and then measure the difference in angle deflection that the string makes with a vertical line to calculate how much force the ball experiences.

Coulomb did this with a sensitive piece of equipment called a torsion spring, but I’m trying to find easier ways.

Charging an metal ball cleanly and with home materials is proving kind of tricky. Supposedly, one way to make two equal charges is to charge one thing and then put it in contact with the other thing, so that the charge splits by symmetry. But then how would we check that indeed we have two equally charged things?

I thought of maybe trying to make a Faraday cage. But instead, today I tried to make an aluminium foil electroscope. To do this I took some insulated wire from by breadboarding kit, drilled a hole through the top of a jar, made the end of a wire a hook and then cut a thin piece of aluminium foil. I then folded this piece of foil and hung it on the wire. the other side of the wire lies on the outside of the jar where I scrunched up some more foil. And when I charge an object by rubbing it against my hair, putting it on the contact causes the tin foil in the jar to separate because of the repulsion of charge in both sides of the strip!

But right now this is a super messy and non quantitative way to do this. One of the main reasons is because of charge decay, usually because the air is humid and so takes away charge from the object. So I’m trying to think of a more robust way to create a repeatable charge at home to measure inverse square law more accurately as part of a killer experiment.