In the end, Gotham’s burqa ban just became unenforceable.

Chirp

“First tweeting – now chirping,” say the BBC on Google+, having noticed that two otherwise unrelated iPhone apps are, for unrelated reasons, named after birdsong.

Chirp is a new iPhone app that aims to solve the perennial problem faced by owners of smartphones whose manufacturers inexplicably crippled the Bluetooth support by design. This one uses sound: in the same way that dial-up modems and, before them, ZX Spectrums used to convert data to sound and send it down a phone line or store it on a tape, Chirp plays encoded data out of your phone speaker, and picks up incoming data with the mic. Brilliantly, they’ve taken the phone part out of a modem, and then put what’s left on a phone. It’s both elegant and faintly absurd, so obviously I love it, but I find myself wondering what the hell it’s for.

The fastest dial-up modem was rated 56kbps, and ran at 53.3kbps — and so took nearly three minutes to transmit a 1MB photo (not that anyone was taking 1MB photos back then). Chirp, clearly optimised for reliability rather than speed, runs at 25bps, and would take nearly four days to transmit the same photo. Instead, Chirp sends a 50-bit code that your phone sends to Chirp’s server, which returns the photo over 3G or whatever. So the process is entirely done with the phone’s data connection, except that the last part of the URL is ‘chirped’ through the air. Probably it could be sent via Chirp’s servers instead and nobody would notice for months — I expect some people still believe Bump actually works by bumping the phones together.

It seems to me that this makes Chirp slightly pointless for sharing files, links, and (they really do this) “140-character text messages” between friends. Partly that’s because I already have more ways to send short text messages to my friends’ iPhones than I could possibly ever need, but mostly it’s because I disagree with their chief executive’s claim that “it’s fairly novel to be able to transmit information to anyone who is in earshot”. In fact, humans have a built-in feature for exchanging short communiques with other humans within earshot. Building that into telephones, a device specifically invented to circumvent that limitation, therefore seems like the most monumentally pointless endeavour since the world’s most easily accessible encyclopædia was translated into Latin.

The idea of putting chirps out over tannoy systems or radio broadcasts is a more interesting. Then it’s essentially a QR code made of sound, and there are bound to be uses for that. QR codes, like Blippar and, well, more or less every attempt to make inanimate objects interactive, suffer because for every legitimate use for them there are a thousand stupid ones shoehorned in by PR nitwits. I’ve even seen an information stand with a QR code but no URL — it has machine-readable data but no human-readable data. It’s hard to shake the feeling that it was designed as a tool for the benefit of rebel robots.

The problem with chirps, QR codes, and the like is that there’s almost always a far more efficient way of doing the same thing. The obvious approach to broadcast links as sound would be to read out the URL. If that’s impractical then the correct solution is to come up with a better URL system — long strings of gibberish are rarely ideal. Even if you use sound, there doesn’t seem to be any need for the radio station to change their broadcast so much as to team up with Shazam so the app can directly recognise their station and connect users to whatever content is under discussion. Of course, most users will listen to one or two radio stations at most anyway, so 99% of the time they won’t even use that; they’ll just choose from their favourites list. And then you’re just tweeting links, which incidentally works fine.

But okay, let’s say you’re ‘chirping’ URLs. I can’t decode them: I almost never listen to audio on any device other than my iPhone, which will presumably mute said audio the moment Chirp starts listening for it. The one thing I’d like it to do — pull links out of Pod Delusion reports — is the one thing it can’t do. On the other hand, Downcast‘s ‘share via Twitter’ feature lets me send a direct message to @rtm with a link to the show-notes, which accomplishes the same thing without building a whole nother app or running yet another server full of short-coded URLs.

I’m sure there’s a use for Chirp, but I reckon you’d have to have it installed for a couple of years before it came in handy. The BBC rather worryingly suggest that the developers “see a future where you pay for a can of drink with a chirp” which seems like it would make credit-card-cloning scams so easy that there’d be an app for it on Cydia within the week.

As an alternative method of sending URLs over the phone or radio, I wondered a while back about creating a URL shortener that made links memorable rather than short. There’s a memory trick that encodes any six-digit number into a memorable image, by building them out of three pegs from a set of 100. If we could have more choices from smaller pools, we could build a site with long but easy to say, hear and remember URLs — a Bit.ly for the mind. You’d probably need a pictorial keyboard on the homepage to make it usable. I never bothered to make it principally because I don’t think it’d be very useful.

And neither, I fear, is this.

1bn Hiroshimas = 1 (Isle of Wight) x 20 (speeding bullets)

I think one of the problems with science journalism is that before a science story can be reported in the news media, someone has to convert everything from metric to journalist units. But some recent work may allow us to do science directly in journalist units, thereby making scientific papers immediately understandable to laypeople. According to a throwaway letter in the Guardian

1 billion Hiroshimas = 1 Isle of Wight × 20 speeding bullets

This is based on a G2 article about the asteroid that wiped out the dinosaurs that described the asteroid’s mass, energy and speed in those terms.

Unfortunately, the equation is wrong, as you can’t multiply speed by islands to get explosions. But it’s not far off. In fact, kinetic energy K = ½mv², where m is mass and v is speed — so actually

1 Chicxulub asteroid strike = 10⁹ Hiroshimas = ½ Isle of Wight × (20 × speeding bullet)²

I’ll forgive the correspondent the factor of 2, but he should have known that the speed needed squaring given that the only other sentence in his letter was “Who needs E=mc²?” Honestly, it’s as if some people have no grasp of dimensional analysis at all.

In fact, this is also wrong, because the Isle of Wight is more correctly a unit of area, not mass, so to use standard journalism units, we should really write

1 Chicxulub = 10⁹ Hiroshimas = ½ (Isle of Wight^³⁄₂ ρ_rock) × (20 × speeding bullet)²

Better still, that should be

1 Chicxulub = 10⁹ Hiroshimas = ½ (Isle of Wight^³⁄₂ ρ_rock) × (21 × speeding bullet)²

as the rock hit 20 times faster than a bullet, not 20 times as fast as one.

And we can test this hypothesis simply by typing “(1/2) * (density of rock * (isle of wight area)^(3/2)) * (21 * speed of bullet)^2″ into Wolfram|Alpha. It returns the figure 5.023×10²³J, and if you click on that figure, it rephrases it to “≈ 1.005 × estimated energy released by the Chicxulub meteor impact”.

Let’s just bask in the impressiveness of that for a moment.

Done basking? Then it’s time to admit there are a few problems with this. Alpha cites this as 8 billion Hiroshimas, not one billion. Alpha also takes ‘a bullet’ to be a rimfire .22LR usually deployed against small pests and tin cans, whereas Dr Collins appears to favour the somewhat meatier M16 assault rifle. Maybe that’s standard for a speeding bullet. Also I assumed the asteroid was a sphere that would cover an area of land equal to one Isle of Wight. In fact the Isle of Wight is long and thin so if we spun it around its major axis it would be a bit lighter than this; equally we could attempt to estimate the mass of the Isle of Wight and that could go either way.

The point is that you absolutely can do science in these units. They totally work. We use metric instead only because the numbers are easier — 1 Joule is 1 kilogram metre per second squared, avoiding having the annoying factor of 21 kicking around that the journalism units version above does. (I’m not going to quibble about the billions, though, as you only need to define the ‘gigashima’ to make that go away.)

To make life easier for anyone choosing to do science in journalism units, I have identified some relationships that may prove useful:

1 coal-fired power station ≈ 1 Hiroshima per day
1 thickness of human hair ≈ 1000 Olympic swimming pools per area the size of Wales
1 weight of a double-decker bus ≈ 1 Hiroshima per distance to the moon and back

All of these are approximate, but they’re all exactly true for at least one combination of reasonable guesses, so all we have to do is identify a mutually-convenient set of plausible values, then agree to use it forever. We can’t really fiddle with

1 Isle of Wight = 381km²
1 distance to the moon and back = 3.85×10⁸m
1 Wales = 20,779km²

and we know that defining

1 speeding bullet = 340ms⁻¹
density of rock = 2.65g/cm³
1 Chicxulub asteroid strike = 5.023×10²³J

gives us one neat relationship. Let’s add to that

1 Hiroshima = 1 Chicxulub ÷ 8 billion = 6.27875×10¹³J
1 coal-fired power station = 1 Hiroshima ÷ 24h = 726.7MW
1 double-decker bus = 1 Hiroshima ÷ 1 moon and back = 8.416 tons
1 Olympic-size swimming pool = 2,500m³
1 thickness of a human hair = 1000 Olympic pool ÷ 1 Wales = 120.3μm

Now all our relationships are spot on, and we can hopefully get on with doing some science with journalism units. At least, science that involves using coal to power buses to the moon.

Ie, the best science.

Might Qwerty be optimal on touchscreens?

It’s a common misconception that the Qwerty keyboard is designed to slow users down to prevent typewriters jamming. It fact, it’s designed to keep commonly consecutive letter pairs apart, so that two adjacent levers won’t collide.

(A more fun, but irrelevant, Qwerty story is that it is also designed such that the word ‘typewriter’ is all on the top row, to make demonstrating it easy. This story, if true, is itself fun but sucks all the fun out of the fact that the longest word that can be typed on the top row of a typewriter is ‘typewriter’. One of these is a fun fact, but I’ve no idea which.)

Nowadays, obviously, there are no swinging arms to collide, so we want the commonly-used keys to be reachable, and if possible to alternate hands as much as possible. Dvorak and Coleman have each had a stab at designing a better layout, but both aimed at the computer keyboard.

But increasingly, I type on my phone, using one very mobile thumb. I can get to any point on the screen, more-or-less right away — but sometimes I miss, and usually the phone figures out what I meant and autocorrects it. So maybe the most important thing about any given keyboard layout is how likely it is that a typo will result in a real word that the phone isn’t to know isn’t what I meant.

I wondered if suddenly Qwerty might be optimal again — separating pairs of letters that can be swapped to make another real word and that appear next to each other in English words aren’t totally different goals. So I thought I’d investigate.

So first I loaded the CSW12 Scrabble word list, and worked out a big table of how many places in the list you can replace each letter with each other letter to create a new word.

	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W	X	Y	Z
A		176	681	305	5253	182	216	252	4180	15	208	477	166	401	4717	297	5	483	953	453	2898	52	201	54	585	30
B	176		1240	1238	284	1157	1123	775	91	360	406	1003	1383	761	157	1579	14	1407	1127	1344	74	398	836	70	366	151
C	681	1240		1109	375	876	1218	843	162	261	1102	1004	1036	1379	265	1418	24	1117	1832	1783	120	418	816	153	274	209
D	305	1238	1109		715	776	1445	709	205	299	942	1600	1467	1838	227	1242	18	7549	10979	2348	107	549	726	135	616	307
E	5253	284	375	715		186	553	454	4712	22	459	938	1010	683	3434	470	4	956	2123	1734	1893	87	291	57	2162	57
F	182	1157	876	776	186		723	592	76	261	357	846	829	632	126	1040	11	782	1037	1121	57	387	620	30	203	89
G	216	1123	1218	1445	553	723		556	186	333	747	812	810	1013	191	1018	39	914	1194	1514	103	373	669	130	371	180
H	252	775	843	709	454	592	556		137	261	641	1186	979	635	260	1098	6	1079	1215	1453	81	239	805	34	356	137
I	4180	91	162	205	4712	76	186	137		35	129	519	130	380	2786	154	3	387	525	331	2671	42	200	23	1118	28
J	15	360	261	299	22	261	333	261	35		117	311	301	224	27	331	2	344	320	346	1	125	196	7	136	66
K	208	406	1102	942	459	357	747	641	129	117		947	779	892	145	911	50	816	929	1340	71	379	517	119	273	180
L	477	1003	1004	1600	938	846	812	1186	519	311	947		1420	1938	479	1363	15	3271	1876	2049	286	599	899	142	394	248
M	166	1383	1036	1467	1010	829	810	979	130	301	779	1420		1085	238	1898	14	1321	1225	2912	119	575	747	130	380	250
N	401	761	1379	1838	683	632	1013	635	380	224	892	1938	1085		286	1367	7	2439	1925	2091	301	529	711	259	440	250
O	4717	157	265	227	3434	126	191	260	2786	27	145	479	238	286		296	6	574	531	389	2291	58	314	36	596	27
P	297	1579	1418	1242	470	1040	1018	1098	154	331	911	1363	1898	1367	296		11	1255	1528	2061	166	580	1011	141	377	243
Q	5	14	24	18	4	11	39	6	3	2	50	15	14	7	6	11	0	12	30	16		5	10	2	2	2
R	483	1407	1117	7549	956	782	914	1079	387	344	816	3271	1321	2439	574	1255	12		4806	2173	447	591	885	205	613	250
S	953	1127	1832	10979	2123	1037	1194	1215	525	320	929	1876	1225	1925	531	1528	30	4806		3126	327	617	887	232	2621	6540
T	453	1344	1783	2348	1734	1121	1514	1453	331	346	1340	2049	2912	2091	389	2061	16	2173	3126		256	682	1187	215	602	429
U	2898	74	120	107	1893	57	103	81	2671	1	71	286	119	301	2291	166	0	447	327	256		43	416	15	239	12
V	52	398	418	549	87	387	373	239	42	125	379	599	575	529	58	580	5	591	617	682	43		353	96	142	154
W	201	836	816	726	291	620	669	805	200	196	517	899	747	711	314	1011	10	885	887	1187	416	353		108	400	136
X	54	70	153	135	57	30	130	34	23	7	119	142	130	259	36	141	2	205	232	215	15	96	108		74	58
Y	585	366	274	616	2162	203	371	356	1118	136	273	394	380	440	596	377	2	613	2621	602	239	142	400	74		108
Z	30	151	209	307	57	89	180	137	28	66	180	248	250	250	27	243	2	250	6540	429	12	154	136	58	108

As you can see, the letters involved in typos that are genuine words are also the most common letters — except C and P. (The frequency values are on an arbitrary scale to match the typo figures.)

Then I wrote a Python routine to generate a ‘badness’ score for each layout, which is the total number of words you can make by replacing a letter of another word with one of the six keys adjacent to it. Running it on 10,000 random layouts, the average badness is around 83,603, with a standard deviation of 14,024.

Here are some other layouts I tried:

Layout	Badness	STDs above mean
Qwerty	119,170	2.54
Dvorak	121,458	2.70
Colemak	112,354	2.05
Best random	46,414	−2.65
Worst random	151,438	4.84
Alphabetic	74,064	−0.68
Best I found	31,992	−3.68

(Predictable answer to question in title: “haha, no”.) Alphabetic uses the same key layout as Qwerty: 10 on the top row, 9 on the second and 7 on the bottom. The ‘best I found’ layout was derived from a random board on that Qwerty grid (since actually Dvorak and Coleman don’t really fit on a phone), by swapping letter pairs at random and keeping the change if it seemed to work. (This is called a ‘genetic algorithm’, albeit a crude one.) I think I did 5,000 steps, five or six times. Here’s the layout it found:

The most obvious thing it’s done is put S (the most typoable letter) in a corner and shoved Q up against it. Another potential improvement to the model is to account for second-nearest neighbours — since flagging an error but correcting it to the wrong thing isn’t much better than missing it.

Another thing it’s done is put all the rarest letters in the middle where they have lots of neighbours — almost precisely the opposite of what Dvorak and Coleman did. Which makes sense, both intuitively and because all the standard layouts are in the worst 5% of all layouts (assuming normal distribution).

Anyway, I think we can all agree this is plainly the best possible keyboard layout for smartphones, and we should name it Taylak and petition Apple and Google to include it as the default for everything ever. I certainly can’t imagine how using the same layout on phones and computers could possibly be more desirable than this.

Here, to end on, is the worst layout I could find, with 204,290 = μ + 8.61σ possible real-world typos:

Nobody use that layout.

Extrapolating into the past, we’ve missed three increasingly implausible opportunities to do this before.

On July 5 2010, Total Film fooled a lot of people into believing that that was the day Marty McFly and Doc Brown visited in Back To The Future Part II. In fact it was October 21 2015.

Yesterday, on June 27 2012, Simply Tap fooled a lot of people into believing that that was the day Marty McFly and Doc Brown visited in Back To The Future Part II. In fact it was still October 21 2015.

To punish people for falling for such nonsense, I propose more of these, increasing in frequency as we approach the actual date, so that when it actually happens, nobody believes it. But when?

The first of these errors was 1934 days premature. The second was 1211 days premature. (I’m not counting the copycat hoax on July 6 2010.) I think the reason these hoaxes are so seductive are that $\frac{1934}{1211}=1.60$, and that’s very close to the golden ratio, $\phi$. $\phi\approx1.62$, and has the lovely property that $\frac{1}{\phi}=\phi-1$ (or, $\phi^2=\phi+1$). It’s the only number of which that’s true, and it often appears in nature, art and architecture.

(Or, if you prefer, reports of $\phi$ appearing by accident are mostly coincidence and optimistic rounding, and so is this. I’ll leave that decision to you.)

Continuing the Golden Cascade of Back To The Future Hoaxes, we should have the next on September 23 2013. There will be two in 2014: on July 4 (when the alien mothercraft destroys the Hill Valley town hall) and December 28. The hoaxes will have to come thick and fast in 2015: April 18, June 27 (like this year), August 10, September 6 and 23, and October 4, 10, 14, 16, 18, 19 (twice), and then 25 separate hoaxes on the day before Future Day.

To be honest I suggest we use areweinthefutureyet.com for those ones.

On reflection, perhaps they shouldn’t go this far. It’s a bit sad, isn’t it?

I saw this post about the graphs of National Novel Writing Month on my brother’s blog ages ago, but an RSS malfunction showed it to me again today, and I got to thinking about his final graph: words written on any given day, plotted against how far behind he was that day (or, more precisely, words left to write / day left, normalised to the same value on day one). He correctly notices “a hint of a positive correlation”, and notes that that may not be causation, but could be an external factor, such as his determination to show his doubting wife what for.

I have my own theory, arguably more prosaic but also more interesting, so I created a simulation to test it. I assigned a random number to each of the 30 days of November, using the formula RAND()*RAND() to create a nice distribution. I assigned a number of words to each day by multiplying this random number by 50,000 and dividing by the total of all 30 random numbers. This created a month of simulated writing with random ups and downs, but a guaranteed total output of exactly 50,000. (For these purposes I allowed non-integer numbers of words.) I worked out the cumulative wordcount and behindness index for each day, and plotted them, with a trendline (in red) and R2 value, and ran simulation after simulation.

My theory was proven: every single one had a positive correlation.

Of course it did — any deviation from the 1666.7-word daily target will be reflected in your behindness score on every subsequent day, and if you write your 50,000th word on day 30, then it will be balanced by an equal and opposite deviation spread unevenly across those same days. Any novel of exactly 50,000 words will have this hint of positive correlation between behindness and words written. I suspect this would hold even if we allowed some days to have a net deletion of words. Novels of just over 50,000 (such as almost all of them) will get a slightly reduced version of the same effect. There are only two ways to avoid it. One is to write a wildly different number of words — say, 40,000 or 60,000. The other is to write exactly one thousand, six hundred and sixty six and two thirds of a word every single day, although that involves using a lot of three- and six- letter words.

I’m not sure that someone who writes 60,000 words really worked to the 50,000-word target at all, so if you want to know if ‘behindness’ affects performance, you’ll have to examine the stats from people who failed to complete the novel, between day one and whenever they eventually gave up. And to be honest, how useful a sample are they to a study of productivity?

Anyway, I suppose I hope this can be a nice example of how something plausible and supported-by-the-data-looking can turn out to just be randomness viewed from a funny angle.

Also we’ll put the verse about rebellious Scots to crush back in.

It’s rare for two people to say things that are obvious, true and contradictory, but that’s what’s happened here:

Scots will no longer be British if their country votes to leave the United Kingdom, Labour leader Ed Miliband has warned in a keynote speech on national identity.

Miliband insisted that leaving the union would mean that Scottish people would lose their British identity – challenging the argument put forward by the Scottish Nationalists, who have insisted that Scottish people would continue to be British in a geographical sense.

Sigh.

Look, the geography gets a bit complicated, so let’s break it down:

“Great Britain” is the name of the big island that England, Scotland and Wales are on.
The country that occupies this island and a few nearby bits (most notably Northern Ireland) is called “The United Kingdom Of Great Britain And Northern Ireland” and is long enough already without adding “Except Scotland”.
The “British Islands”, apparently, is the UK plus some other, distinct states that the Queen is also head of.
The “British Isles” is every damn rock between Iceland and France.

So yes, Scotland will remain British, because it’s part of Britain. But it won’t be part of Britain any more so it won’t be British, really.

I would dearly love to know what Ed Miliband thinks will happen if we vote to “leave Europe”. I’m picturing an army of tugboats, or drills to perforate the tectonic plate along the Channel.

If you refuse, you lose your benefits and/or head.

It has come to my attention that some people do not think about how things will look. I mean, what would we think of this behaviour from, say, the host nation of Eurovision?

A group of long-term unemployed jobseekers were bussed into London to work as unpaid stewards during the diamond jubilee celebrations and told to sleep under London Bridge before working on the river pageant.

I think it’s a bit despotic. An absurdly rich woman who is head of state simply because her dad was has been thrown a £3 billion party at the citizens’ expense, spent the afternoon heading a £12 million flotilla, while on the banks, the pageant was helped along by poor citizens working in shitty conditions for no money and against their will.

I realise it’s not quite as bad as I make it sound, but it’s very nearly that bad. Like when Amazon deleted 1984 from everyone’s Kindle. Seems like it’d be quite easy to forsee how this shit might look when it inevitably gets out, and maybe think, you know what, just this once maybe we’ll pretend to be mature adults. You know — purely in the interests of avoiding bad press.

Tomorrow’s Front Page

Crappy interfaces in real life

I spend a lot of time getting cross at crappy interfaces on software, but the fact is that real life objects are just as bad. I’m typing this at a laptop, for example, with a trackpad. And while Apple have multitouch, click-sensitive trackpads that make sense, this one scrolls using the right hand half a centimetre of pad, which is visually and tactilely indistinguishable from the rest of it. It’s the little things, like Jeff Atwood’s cat feeders, but it’s also the really big things. Our old DVD recorder which asked, when you put a DVD in, if you would like to “access the disc contents” — and to get to the DVD menu, you had to say ‘no’. Toasters which measure time in completely arbitrary units, at least as far as I’m aware. Washing machines that have a key to their own interface on the front. TVs that have a hundred buttons to do basically one thing and you still have to tap ‘circle with an arrow’ four times to watch a DVD.

It’s like people don’t learn. I remember when I could operate a microwave oven by typing in a time and pressing ‘cook’. Ryan North has one where you can just bark numbers at it. And yet this is the thing we have at work:

Obviously that’s absurdly overcomplicated for a device which, much as manufacturers kid themselves otherwise, has only two modes: on and off. (Really, who wants to ‘slow defrost’ anything?) But my main objection is that I have had a bagel spinning around in there for ages, with the microwave all humming to itself and lit up, and without the bagel getting so much as lukewarm. Apparently it only heats food if you set the timer. Where exactly the clue to this is supposed to be is not clear. I assumed when a microwave lit up and span your food round it was cooking. Apparently this one also has a ‘shop demonstration’ mode.

Because the fact is that outward appearances matter. I forgive the absurd obsession of washing machine manufactures with putting any and all options on dials because that at least makes their products obviously washing machines. Humans in Western society have loads of cues and associations built in — say, we pull doors with handles — and it’s daft not to take advantage of them and insane to actively work against them. Putting a handle on a push-only door will confuse and annoy, and this… This is the sink in a bar in Manchester:

Perhaps they considered it designery and artsy, but in fact what they have done is add taps to a urinal and call it a sink. The thing is that I took it for a urinal when I first saw it, and while I noticed the taps fairly quickly, I don’t trust the entire inebriated male population of Manchester to all have done so on any given night. That is a sink which I confidently predict has been pissed in and that is enough to put me off washing in it.

What you need, when cleaning yourself, is a clear interface. So don’t get this brand of shower:

I put it to you that it’s less than totally clear which way is ‘hot’ on this dial. The arrow points right, but it’s on the left. If you’re standing in a room full of steam and you wear glasses, that’s fairly dangerous. And it’s obvious what the dial does, so all they had to do was convey which way was which. That is literally the smallest amount of information it is possible to encode. And still it’s ambiguous.

Even stranger is the dimmer switch in our new conference room.

This has two buttons, one at the top of the switch, and one at the bottom. Here is how I assumed it would work:

Pressing the top switch would make the lights brighter.
Pressing the bottom switch would make the lights dimmer.

Here is how it works:

A short tap of the bottom switch turns the lights on or off.
A long hold of the bottom switch makes the lights brighter.
A second long hold of the bottom switch makes the lights dimmer. This alternates between brighter and dimmer.
The top switch exists only to make the object resemble a lightswitch.

The upshot of this is that we all look stupid when presenting the work of the Unit to outsiders, because we are supposed to be carefully controlling illumination in clinical trials and can’t operate our own office lights.

And I suppose I just thought that stuff like toasters and lights would have been around for long enough by now for us to have basically mastered the art of making them usable. We don’t have Steve Jobs around any more to fix this stuff up for us, so perhaps it would be a fitting tribute to him if we all stopped making watches that require one long and fourteen short button presses, each with its own high-pitched beep, just to correct for daylight savings, hmm?