Introduction

Everybody makes things.

I don’t mean that everybody is an artist, that’s the kind of sentence that makes people close the tab and discard my whole point as baloney. I mean it literally. You have, at some point in your life, put a playlist in an order that mattered to you. Written a message and then rewritten it before sending. Built something in Minecraft, or a spreadsheet you were quietly proud of, or a joke you turned over in your head a few times before saying it out loud. The urge to make a thing and then show it to somebody is about as human as it gets, and it doesn’t switch off just because you’ve decided at some point that you aren’t creative.

Why do we make things? Why do we want to make things?
Nobody is forcing you to, there’s no external reason for most of it. And yet.

Here's a question for you: how much should it bother you that most of what you make isn't any good?
Keep that in your head, i’ll come back to it.

On America’s Labor Day, in 1953, science fiction author Theodore Sturgeon made a talk at the World Science Fiction Convention in Philadelphia. Theodore Sturgeon had spent the previous couple of years defending science fiction to the masses, to the kind of people that had never read the genre and never planned to. At this convention specifically, he has been recounted in saying:

“When people talk about the mystery novel […] they mention The Maltese Falcon and The Big Sleep.
When they talk about the western, they say there's The Way West and Shane.
But when they talk about science fiction, they call it 'that Buck Rogers stuff,'
and they say ‘ninety percent of science fiction is crud.’ ”

Sturgeon continued:

“Well, they're right. Ninety percent of science fiction is crud. But then ninety percent of everything is crud, and it's the ten percent that isn't crud that is important. and the ten percent of science fiction that isn't crud is as good as or better than anything being written anywhere.”

Part One.

If 90% of everything is crap (I’ll be saying the word ‘crap’ instead of ‘crud’ from now on), and you make things, then 90% of what you make is crap. This isn’t an insult and it isn’t a prediction about you specifically, it’s just the rate. It means that the note-taking app that got as far as a database schema, or the song that got as far as four sets of chords, or the half-finished discord bot, or the half edited youtube video, or whatever example applies to you specifically -- all of that was always going to happen. None of that is evidence that one is ‘bad’ at making things, it’s just the way human creation works. Most things are bad, and most things don’t work out, even if and when they are finished!

I feel like most people understand this in the abstract, but the main problem is that knowing it doesn’t help much because as humans we don’t experience our failures as a statistic -- we experience them one at a time.

Tom Scott makes this point in a video about getting started online, and it's stuck with me since. I don't think I've fully embraced it even as I write this, but it turns up as an objective truth every time it's applicable: that the chances of a project succeeding don’t really have much to do with how much time, effort or money has been spent on it.

People, myself included, complain that the specific project they’ve poured all of their heart and soul into doesn’t ever reach anywhere. But the one little quick and dirty thing that took no effort whatsoever? Others found that more interesting and funnier, and enjoyed it much more than the large project.

You can’t simply buy your way out of this phenomenon, if you spend £0 and an hour on a project it is likely to fail, yet if you spend £1,000,000 and a year on a project it is still likely to fail. Money doesn’t move the rate here at all! It only moves the cost per attempt. And starting out, with nothing behind you -- no followers, no subscribers, no users, no recognition beyond the one niche community you find yourself in -- that can be a trap that demotivates you and causes you to make nothing in the end.

If you instead took that time, that year and that money that you would’ve spent in that year, and made five, or ten, or twenty or thirty or a hundred projects all throwing them at the wall and putting them on the internet for all to see; you’d have a much larger chance of hitting that 10% that is genuinely a good idea beyond your own biases, a larger chance of hitting a good number of projects that other people not only want to see, but have seen. Have watched, have played with, have signed up to. If the rate is fixed, which Sturgeons Law tells us that it is, the number of attempts are your only lever to success. Programming’s cost per attempt is near zero! A stupidly good deal, and one that every aspiring developer should take for their playground of learning new concepts and shipping fun ideas or web toys.

Part Two.

Of course, all of this assumes that you count ‘success’ as being popular. The argument for volume only works if you want to maximise the chances of yourself or your project being popular on the internet. Say you spend a Saturday morning developing something, put it online, and nobody comes. What actually failed there? The code works, you’re better at something than you were on Friday. The only thing that failed is a number of views or users...?

Well -- not nothing. Let's not pretend. If you wanted people to see the thing, and people didn't see the thing, that's a real disappointment and I'm not going to talk you out of it. But really remember what that number measures, at the end of the day it’s whether someone bigger posted something similar that week. It’s what time you hit publish. It’s whether the ‘algorithm’ we catch ourselves in nowadays was in the mood. It’s a number that is measured by whether or not the people who would’ve loved it happened to be scrolling at that time. It is a measurement of other people’s behaviour, and you have almost no influence over any of it.

That’s fine as a thing to want, yet it’s disastrous as a thing to grade yourself on. That could very possibly be the grade that keeps you going or not. How many people do you know, maybe even yourself, who stopped streaming on twitch or making youtube videos because, deep down, nobody was seeing it and that made them lose passion? Using this measurement as a baseline for carrying on makes the decision to continue doing whatever you love based on something you could have never truly controlled.

Here’s what to measure instead!

What do you know now that you didn’t know on Friday? Not vaguely. Specifically. Do you understand how OAuth actually works, rather than how to install a library that does OAuth for you? Did you write a parser for the first time and find out it wasn’t the dark art you assumed it was? Did you learn that you hate writing frontend, and that you’d rather never do it again? That last one counts too, by the way. Ruling things out is data.

Now, I know how this sounds. It sounds like the thing that people say to you when a thing has flopped and theyre being kind about it. “Never mind! You learnt a lot!”. I’m not doing that, because I don’t think it’s the nice metric, I think it’s the harder one.

Views tell you about last Tuesday. They tell you what the algorithm was doing, who happened to be awake or what else happened to be posted at that time. That’s about it. It’s a report on conditions that no longer exist. What you learnt, by contrast, tells you about the next thing you make -- the only part that you have any say over.

Sturgeons law is a ratio, it isn’t a quality bar. It says 90% of what you make is crap -- it says absolutely nothing about how good your crap is. Your hundredth bad project is a wildly different object to your first bad project; and while the rate never moves, the floor definitely does. Every time you make something, including and especially all of the times that nobody looked, the floor has moved. Nobody is watching that number increase. It increases regardless.

The other thing about this measurement is that nobody can take it off of you! Platforms die! As I’ve written about before, twitter’s API went away and took a decade of engagement numbers with it. Vine is gone. Sites get delisted, accounts get banned, view counts get reset by a redesign. None of that, no matter how demoralising, ever touches what you actually learnt by building the thing.

You only get this number, however, if you go and look for it. Failing without noticing what you’ve learnt is the one way to actually waste the attempt. So write it down. One line in a text file when you abandon something -- what it was, what broke it and why. It takes thirty seconds, if that!, and is the only record you’ll ever have of getting better.

Part Three.

Go and look at the Github profile of somebody you think is better than you.

If you don’t develop, bear with me. This can be relayed to Instagram pages, or youtube channels, or anything social that you could compare yourself to in an unhealthy way.

You’ve probably done this before, and you’ve probably felt a bit ill afterwards. Six pinned repositories, all of them finished, all of them with a readme that has a logo in it and little badges along the top. Meanwhile you’ve got a folder on your desktop called projects with eleven things in it that don’t run.

It’s the Netflix shelf again.

That’s genuinely all it is. A pinned repository is a curated shelf in exactly the same way the science fiction aisle is also a curated shelf, and you are doing the same thing the audience does at Waterstones; quietly concluding that the ninety percent stopped being made. It didn’t, it’s just not on the shelf, because why on earth would anybody put it there? Nobody pins the graveyard.

Every developer whose profile makes you feel inadequate has, somewhere on a drive, a directory called test, and next to it test2, and next to that one test2-actual. They may have branches they never merged; they may have repositories set to private because they were embarrassed by them. You are comparing your own unedited everything against somebody else’s top ten percent.

Github makes this worse than most places, mind. The contribution graph! A year of little green squares, and not one of them says what it was for. A year of green from a project with fifty thousand users looks precisely identical to a year of green from a thing that got abandoned and never spoken of again. You cant tell the difference from the outside, and neither can anybody looking at yours. Which is the bit that took me a while to get, actually. You are somebody else’s intimidating profile. Somebody has probably looked at your pinned repositories -- the two or three you were willing to admit to -- and had this exact feeling about you, with absolutely no idea about the four things you binned that month, or the one you deleted entirely because it made you cringe to look at.

The shelf is not the shop.

Part Four.

Nobody made you do this.

It’s worth sitting with that for a second because it’s a really easy thing to forget. There’s no deadline with any of it. Nobody who set the brief, nobody is marking it, and if you stopped tomorrow the only person who would really notice is yourself. Whatever it is youre building at the minute, you started it because you wanted the thing to exist, or because you wanted to find out whether you could actually build it. That was the reason and it was a perfectly, perfectly good one.

So where did the scoreboard come from?

Because you didn’t write it, somebody else did. The follower count on your profile, the star count on your repository, the little view count under the video -- all of them is a design decision somebody else made in a meeting. They are there because showing you a number keeps you coming back to look at the number. It’s not a conspiracy, im not calling out these platforms for it. It’s just, the way things work. It’s just product design, and it works so well that most of us have quietly accepted a definition of success that arrived preinstalled, and that we never actually agreed to at any point.

It’s the trap I mentioned earlier in part one. When you’re starting with nothing behind you whatsoever, that borrowed scoreboard doesn’t read as “this got low numbers”, it reads as a verdict. It reads as the answer to whether you should keep going at all, and it is a genuinely terrible answer to that question.

So take it back off them, and pick your own criteria instead.

I don’t mean that in a vague inspirational-poster sort of way. I mean literally decide, before you start, what would make the thing worth having done. And the good ones tend to be smaller and stranger than you’d expect. Getting to use a language you’ve never touched before. Working with somebody whose stuff you’ve admired for years. Making one specific person laugh. Finally understanding a thing you’ve been avoiding because it looked to hard. Having something that exists, that you can point at, that you made.

The best example for me would be this very website that you’re viewing this content on. I had no idea how to use discord’s authentication before the guestbook on the home page, I had no idea how to get my website on webrings or the kagi small web. I didn’t even really know how to write blog posts in a nice way for me, without needing to contend with html every time I want to share my thoughts. But it all means that my experience from making this site not only serves a purpose for me, it gives me a dumping ground to write up my feelings, but it also meant I learnt a lot whilst doing it. Skills that will help me on whatever the next thing is. And it doesn’t have a counter next to it, there is no number on my github profile for “I finally understand this concept”. But It is important. It is a measurable thing for my success. And it would be wrong to ignore it.

It’s a score that you keep to yourself. That sounds like a consolation prize at first, I get that, but only until you truly notice what that number actually means: that every other number you have ever graded yourself against was a decision somebody made, in a meeting, for reasons that had absolutely nothing to do with you.

Yours is a decision you get to make.

So build that web toy. Make that youtube video. Write that song or start that twitch stream, and don’t let it bother you that most of what you make won’t be any good. You should never expect your first tries to be. You should never expect your 500th try to be.

After all, its human to want to create things. Not to be perfect at it every time.

Thank you for reading.

- Dia

This post is also available as a PDF.

Pre-ramble

I am, unfortunately yet unapologetically, British. America receives a lot of flak for ‘not yet switching to metric’ -- despite the fact that they actually have ‘switched’ to metric in the legal sense -- and in my opinion it’s us who get away with the same, if not worse, behaviour.

In America, it’s very obvious what you should be measuring your stuff in; the measurements may be hard to follow or not make obvious sense to one, but one should always be able to identify which measurement is used. Britain has none of this. In the UK, every single thing you can think of has its own unit¹, there is no sweeping ‘all liquids are gallons’ rule or anything of the like. For example: soft drinks are sold by the litre, milk is sold by the pint. Draught beer and cider can be sold by the pint, but all other alcohol must be sold by a specific subset of amounts millilitres.² Bottled/canned beer and cider can never be sold by the pint, but rather ‘568ml’. All of these rules for alcohol are outlined in the Weights and Measures act.

‘Weights and Measures’ is mostly known in British culture as ‘those rozzers³ who come in to pubs checking that a shot is actually a shot’.

All of this is to say that my relationship with units of measurement has been, on paper, rough. In practice none of this matters, and it comes to British people the same way that grammatical gender comes to those native to languages that have it. This section of the post is purely contextual for the landscape I’ve grown up in (and ultimately gotten used to) regarding the units of measurements of stuff.

Introduction

The metric system is obviously better than the imperial system, and I at no point will be arguing otherwise. It succeeds in its goals of being an international standard, it is easy for everybody to understand, and it allows the base of everything ever to be either water or the size of the earth itself⁴. I have even written at length in the past about metrics use cases on the road, in an old script that I may write up as a blog post sometime in the future.

I also believe that the ‘Imperial Units of Measurement’ get a very bad wrap, and in some cases undeserved animosity. Whilst it as a single system isn’t ideal for every day use, that doesn’t mean that every single time a unit of measurement from it is used, that unit of measurement is automatically bad -- an opinion held by many people the world over. Okay, mainly just Europe-over.

First things first, a couple of definitions and disclosures:

• Whenever I mention the ‘Imperial System’, I’m talking about the United Kingdom’s imperial system of measurement. The real Imperial (as in Empire) system of measurement. This is the one referenced in UK law, including but not limited to the prior mentioned Weights and Measures act.
• Unless stated otherwise, I am not talking about the US Customary Units. This is the ‘imperial’ system used in the United States, and has become what non-brits think is the ‘imperial’ system.
• The ‘Imperial System’ is not one set of units. As put by jan Misali -- the writer of an excellent source I found whilst writing this piece -- it’s a family of slightly different systems with a common origin.

If it's a family rather than a system, then “is the imperial system any good?” isn't a question with an answer. The only question you can actually ask is whether a given unit is any good, one at a time. So let’s do that, starting with the one Britain defends most fiercely.

Culture, and The Pint.

Let me tell you a tale of the pint in British law.

The pint is the only unit of measurement that Britain has gone out of it’s way to protect in law. The pint is not simply a legacy unit that nobody got round to killing, this is a unit that the country actively chose and has a huge stake in the culture around here. In 2024 the UK got a new government. After around 14 years of populist right wing nut jobs, the extremely boring socially-left-of-centre, fiscally-right-of-centre Keir Starmer took hold with his Labour party. At the time of the election, satirical journalist Johnathan Pie called the new prime minister ‘50 shades of beige, captain white bread.’ on a piece with the New York Times.

It felt fake, therefore, when this government tried to get involved with the ‘culture’ of the UK. The opposition saw this as an opportunity to get a dig at them.

In early 2025, the pint was recognised in law in a seemingly controversial way. The vehicle was the Product Regulation and Metrology Bill, a piece of legislation so dry that its own scrutiny committee complained it was a skeleton. Buried in it was a power for a minister to decide, by statutory instrument, which units of measurement may be used and what sizes goods may be sold in -- along with a power to amend the Weights and Measures act to match. The Conservatives -- those ‘right wing nut jobs’ that had just been beaten in their worst defeat ever, and gagging for some positive press -- spotted this and claimed the government intended to ban pubs from selling pints. It did not, and everybody involved knew it did not, but a Bill that could theoretically touch the pint is close enough to a Bill that bans the pint if you want a headline.

What followed was an almost satirical chain of events that need not concern us in this post, but essentially lead to the three biggest parties in the UK arguing on the best way to preserve the pint. An afternoon of the House of Lords competing over who loved beer the most. They all agreed that ‘the pint’ was a direct extension of British culture, and should be protected by law, it was the how that took so long to resolve.

Eventually, the result is that the pint is now protected by statute, ministers may not restrict its use for draught beer, draught cider, or milk in returnable containers.

Here’s the kicker: In order to protect our most imperial of imperial units, the enduring symbol of a nation that refuses to be told what to do by continental Europe, Parliament had to write down what a pint actually is. They defined the pint as 0.56826125 cubic decimetres.

But who cares! ‘The Pint’ was saved, even if it didn’t need saving.

But this isn’t a political post, at least it’s not meant to be. So let's ask the actual question using this cultural knowledge as part of the evidence. Ignore the heritage and the Lords and the flag-waving. Is the pint a good unit to use? I think it is, and for the same reason the mile is: because nobody ever ‘measures’ with it. It’s just a handy word that people understand the meaning of.

The Pint isn’t and shouldn’t be treated as a volume, it’s a serving of beer or cider. When you order one, the number ‘568’ does not enter your head, and neither does any other number. “A pint of Guinness” is a complete instruction -- it tells the bartender the drink, the vessel, the price bracket, and the amount in three words. There’s no arithmetic anywhere in the transaction, and everybody is happy.

And when you do need less than a full one, the pint divides in exactly the ways a drinker actually wants. A half. You could have a third or two thirds, as mentioned earlier, but I’d genuinely really argue that nobody needs that. Even if they did, every one of these legal measures is a phrase rather than a sum -- nobody has ever ordered "189 millilitres of bitter". A litre divides beautifully into tenths, which is brilliant! But it’s a property of no use whatsoever to a man deciding whether he's got time for another before the train.

There's something else the pint does that no other British unit manages: it's physically embodied in the object you're handed. The glass itself is the measuring instrument, stamped and verified, and you can see whether it's full. You don't have to trust a scale or a pump or a label. Every other unit in this post is an abstraction printed on something, where the pint is a thing you can hold up to the light and check.

Meanwhile, underneath, the whole thing runs on metric anyway. Alcoholic strength is a percentage by volume. The units of alcohol printed on the side of a can (the ones the NHS wants you to count) are calculated from millilitres of ethanol. A pint of four percent beer is around about 2.3 units, and that number was produced by doing metric arithmetic on 568 millilitres. Nobody in the pub does that sum, and nobody has to, because not only has the sum has already been done and printed, every beer is different! Having ordinary people do maths to calculate their units would be silly.

Legal and health stuff, metric. Regular people? Imperial. Not only is it culture, it also just makes sense. Some things just are because they are, and asking for ‘a pint’ is one of these things.

This is the same pattern as the road! The imperial unit is the interface, the metric is the implementation. This ‘imperial interface’ survives purely because it was never really doing the job people assume units do.

It also survives because I’ll be old and grey before brits give up the word ‘pint’.

Miles and Yards.

There are many, many reasons why metric units of measurement are better than imperial ones on the road. As I mentioned earlier, I have already researched and written about why the UK should switch to metric on our roads -- but these arguments are really quite obvious and you’ve heard them a million times from a million people. Let’s talk about the usability and strength of the mile on the road.

The United Kingdom uses Miles and Yards on their road signs; whilst the American’s may be infatuated with feet, over here it’s the yard that won out. It’s entirely crucial to mention though that miles and yards, and feet for that matter, never mix. This can be difficult to fathom for those who don’t use the system. I feel as if many people who have only ever used metric sort of assume that people use yards or feet as a smaller division of the mile. This just isn’t true, and would be really silly. You wouldn’t ever say

“The distance between Manchester and Sheffield is 38 miles and 1,410 yards”

If you needed that level of precision, you would say

“The distance between Manchester and Sheffield is 38.8 miles.”

So, why use miles over kilometres? I’ve genuinely got a really good reason and you’re gonna have to hear me out on it. It’s by total accident, and it’s all to do with how long an hour happens to be and how fast 60mph happens to be.

The national speed limit on a single carriageway road in Britain is 60mph. Sixty miles per hour. Sixty minutes in an hour. Which means that out on an A road, one mile takes one minute, exactly, and you don't have to work that out because it isn't a conversion -- it's the same number twice! A sign saying ‘Buxton 24’ is also a sign saying ‘Buxton, twenty-four minutes’. Nobody taught me that. Nobody teaches anyone that. You just absorb it after enough journeys, and from then on every distance sign in the country is quietly telling you the arrival time as well.

It holds up at the other limits too, roughly. Thirty in a town is two minutes to the mile. Whilst it’s true that you can go up to seventy on a motorway, that’s still fifty-one seconds if you do the maths, which any sane person rounds to a minute. So across every road in Britain, a mile is somewhere between one and two minutes, and mostly it's one.

Now try that with kilometres! A hundred kilometres an hour, which is the standard limit on a French or Spanish main road, gives you thirty-six seconds per kilometre. Thirty-six. It's not wrong, it's just kinda useless and crappy for this kind of job. Objectively. Nobody is estimating a journey in thirty-six second chunks, so you end up doing actual division, which is exactly the thing metric is supposed to save you from.

The mile is not defensible on its own. They are useless for many reasons, many reasons that Metrication UK will point out and many reasons that I agree with. Miles are only good because they come with miles per hour.

The unit is fine and the unit of speed is fine, but what's actually good is the pair because together they hand you distance, speed and time in a single number with no arithmetic in between. Break the pair and you lose it. If we switched to km/h and kept the mile, or the other way round, the whole thing collapses immediately. If you switch both you end up like France and Spain.

Which is why, incidentally, the argument doesn't extend to the rest of the road. Fuel is in litres and always has been, but efficiency is inexplicably in miles per gallon, a unit describing a gallon that no petrol station in the country has sold since before I was born. Bridge heights are in feet and inches, and also in metres, on the same sign, because getting that one wrong writes off a lorry. British roads are not coherent and I'm not going to pretend they are! I'm making a much smaller claim. Of everything on the road, the mile is the one imperial unit that earns its place, and it earns it because of the company it keeps.

So the mile isn't good because it's a mile, far from it. A ‘mile’ is actually quite hard to learn if you didn’t grow up with it. It's good because of the accident of arithmetic it fell into, and because that accident happens to hand you the one number a driver actually wants. It’s the same shape as the pint, arrived at from a completely different direction; The pint works because it's a serving rather than a volume, the mile works because it's a minute rather than a distance. Neither of them is doing the job we're told units are for, and both of them survive because of it.

Stone, and their British dominance.

I’ll be honest, this section will be hard for me to write. I don’t like the stone as a unit of measurement, I don’t like the pound as a unit of measurement. Pounds are better but, still not ideal. I’ve been using kilogrammes⁷ my entire life and don’t plan on changing now.

But Stone does have it’s place in society and it’s only to do with talking about change of weight. One stone is roughly 6.35 kilogrammes, or precisely 14 pounds. Stones are only used for the weights of people, never the weights of ingredients or produce, or even the weight of live farm animals. Only humans. And whilst I can’t prove this is why it stuck around, I think the stone survived because it's a coarse unit of measure.

Daily weight fluctuation is natural and normal, it can be attributed to water weight caused by water retention by sodium and carbohydrates. According to healthline.com this can be around 2.2 to 2.7 kg per day. If someone was to weigh 60kg, I’d imagine it would be pretty hard to tell when their weight changes to 62kg, or 58kg. However, the consensus that I have heard a lot from people who use stone day-to-day, is that the minute somebody who weighs 60kg gains to 66 or 67 kg, it’s much easier to notice.

Daily fluctuation is the noise floor, a kilogramme sits below that floor -- so a kilogramme of change tells you nothing! It could be salt, could be water, could be what you had for dinner. A stone by contrast sits well above it, so a change measured in whole stones cannot be noise, it is necessarily real.

This isn’t speculation about perception, it's only maths against that 2.2-2.7kg figure! “I’ve lost a stone” is a claim that survives the measurement error, whereas “I’ve lost two kilos” isn't. This is why ‘losing a stone’ is a big deal, it’s why one is an announcement and the other is a shrug, and it’s why I believe many many people in the UK use stone as a unit of measure in everyday speech. Just like the pint, it doesn’t really matter what a stone actually equals. Just the fact that you have ‘lost two stone’ is notable and understandable by the person who you’re talking to.

Now it is true that if somebody is talking about their full weight, they’re likely to say “eleven stone four”, so the unit people report their weight in is really the pound, which is finer than the kilogramme, and my whole coarseness argument should collapse... Except it doesn't, because those are two different jobs. Stones and pounds are how you say where you are. Whole stones are how you say what's changed. Nobody has ever announced that they've lost nineteen pounds; they say they've lost a stone and a bit, because the stone is the bit that means something and the pounds are the bit that moves on their own overnight.

Again, just like the pint and lesser so the mile, the ‘stone’ doesn’t need to mean anything. Nobody knows that a stone is 6.35 kilogrammes, at least I didn’t until I researched for this. But I do know that gaining or losing a stone is a noticeable change, and people around here understand that.

Inches, and their global dominance.

The inch is genuinely cool. So far I’ve been talking about how Britain is a tad quirky, but it all works out in the end. Inches?? Inches won globally. I’m so serious even if you’re reading this from a nice metric country where you always use millimetres or centimetres, it’s likely that many things in daily life are based on whole amounts of inches. The inch won here, it just might be reported in a decimal conversion.

As a “computer scientist” -- a far too fancy term that doesn’t actually describe what I do, I’m a software student -- the example I go to is servers. Server racks are 19 inches wide! Not only in Britain or America, everywhere. Every data centre from Hong Kong to Honduras, from Moscow to Miami, from Aberdeen to Abu Dhabi⁸ is built around a 19 inch opening, and the vertical space in them is measured in rack units of precisely 1.75 inches each (which is why a server is described as “1U” or “2U”, and why those numbers have no relationship to anything metric.) The entire physical infrastructure of the internet is dimensioned in a unit that most of the engineers maintaining it have never used for anything else in their lives.

Or, okay, look at a circuit board. The standard header pin spacing is one tenth of an inch. The whole industry writes this as 2.54mm, which is exactly one tenth of an inch, because (in my opinion) 2.54mm sounds like a design decision and 0.1 inch sounds like an admission. Every breadboard, every Arduino, every ribbon cable connector on earth is built on that pitch, with 0.1inches between every pin.

Or, elsewhere, look at hard drives! They’re usually either 3.5 or 2.5 inch. Screens are sold by the diagonal inch in every country in the world regardless of what that country uses for anything else -- you can’t ever buy a ‘140cm television’. Camera sensors are generally measured in inches.⁹ Car rims and bike wheels are measured in inches. And planes? Aviation flies on feet as an international standard. Today only China, Mongolia, North Korea, and parts of Russia and Tajikistan are working in metres.¹⁰ Why? Because in aviation getting it wrong -- for lack of a better term -- kills people, and the thing that stops you getting it wrong is everyone using the same number. It doesn't matter in the slightest whether the foot is a good unit. It matters enormously that the aircraft, the altimeter, the chart, the controller and the aircraft three thousand feet above you all agree.

That's my argument; much different to the pint or the mile. Those survived because of how they feel to use. The inch survived because of lock-in, an admittedly much less romantic reason yet a much more powerful one. Once a million racks are 19 inches, the 19 inches has stopped being a unit and become a shape. You can't argue a shape into being 500mm. You'd have to replace the racks.

Any software engineer reading this might even recognise this pattern immediately, because it's the same reason we're all still dealing with QWERTY, and CSV files, and time zones. The standard that wins is never the best one. It's the one that got there first and got copied enough times that changing it costs more than living with it. I think I have an entire blog post worth of thoughts about QWERTY.

Feet are, actually, a really good scale.

In the olden times, nearly every culture that needed to measure things arrived at a set of units based on the body. Because the body is the one measuring instrument everybody owns, and nobody can leave at home or pretend that they didn’t have the correct equipment to measure their things at market, “anthropometric” systems of measuring became entirely normal and incredibly useful.

None of these were accurate, but that mattered far less than you'd think. If you're pacing out a field boundary or a ship's rigging, being within a few percent is fine, and being able to do it at all with no equipment is worth much more than being exactly right. The precision came later, when trade and then engineering demanded it, and by then everyone had spent a thousand years or so thinking in units the human body happened to produce. This means that yes, while the ‘foot’ is a silly name for things, and Americans often receive a lot of flack for naming length in ‘feet’ (Britian receives much less flack as we often use centimetres or yards for length), it leaves the foot as a really nice measurement to use.

Today, whilst it’s obviously larger than an average humans foot; at around 30 centimetres the metric system doesn’t have anything close to it. A metre is about three times bigger and a decimetre is about three times smaller. Whilst these are, fine, the foot was designed with measuring human-sized things in mind, which can be incredibly useful. In metric, due to the power of ‘ten’ prefixal naming conventions, you jump from a decimetre at 10cm to a metre at 100cm. Most human sized things fit in that gap.

The benefit to this? It's not that a foot is ‘natural’ or that it feels nice, that's just a story. The benefit is objective arithmetic-free reading! Everything at human scale lands between one foot and about thirty. A doorway is 7 foot. A ceiling is 8. A room is 12 by 14. A sofa is 6. A person is 5 or 6. That's the whole of domestic life expressed in small whole numbers, in one unit, with no decimal point anywhere. Now say those in metric. The ceiling is 2.4 metres, or 240 centimetres -- pick one, they're both slightly awkward, and whichever you pick you'll be switching to the other one within about a minute because a doorway is 2.1 and a person is 180. You end up sliding between two units and a decimal point to describe the inside of a house.

And the second benefit follows from the first: because the unit came off a body, you can measure with your body. If you want to know whether the sofa fits, you can pace the room heel to toe and get an answer within a few percent, right now, with nothing in your hands. Try pacing a room in decimetres.

Metric units are only ideal when they can be measured with tools and written down. Just talking to a person, saying an integer number of feet is much easier to understand than even an integer number of centimetres. Knowing how long one foot is, give or take, it’s not difficult for somebody to imagine three feet or four feet, and allows for imprecision because this is human communication.

The moment you’re building something or working out clearance space, British people switch to millimetres or centimetres. It’s only ever metric that gets written down for serious stuff, the way it should be.

‘Miles per gallon’, and the case against me.

I've spent this whole post picking the units that work, so it's only fair I pick the one that doesn't.

Fuel efficiency in the UK is measured in miles per gallon. Petrol is sold in litres and has been for decades, the gallon hasn't been a legal unit for selling fuel in this country since the mid 90s, but there’s evidence of garages switching far before that.

It was as early as 1981 that petrol stations could display metric if they wanted, and many did as the price of petrol had risen to more than £1.99 per gallon. This was important, as many analogue meters back then physically couldn’t display a higher price. Switching to the litre, which was smaller and therefore cheaper per unit, allowed a smaller number to be displayed without needing to replace every meter. Smart!

There are also accounts from the time that garages were more than happy to change to the litre, as it made their prices look cheaper. It was at the time that garages -- or more accurately the shops within them -- absolutely hated the idea of switching to kilogrammes over pounds as this would make the price per unit increase, and therefore look more expensive.

Anyway, both of these together saw a gradual transition to litres throughout the 80s. The Energy Institute statistical series says that their records changed from “new pence per gallon” to “new pence per litre” at the start of 1989, and selling petrol by the gallon become illegal entirely by New Years Day of 1995.

So what are we left with? A measurement of fuel consumption expressed as a ratio to a unit that has been illegal to sell fuel in for thirty years. To find out what a journey actually costs, you have to convert mpg into something involving litres, because the number itself doesn't connect to anything you can buy.

It's also, objectively, the wrong way round. Miles per gallon gets bigger as consumption gets smaller, which flattens exactly the differences that matter. Going from 20mpg to 25 saves you more fuel over a year than going from 50 to 60 does. The physical numbers say the opposite as a 10pt difference should save you more than a 5pt difference, but that’s just not how the maths work out. That actually means that you can go three times less per gallon.

Over 10,000 miles: 20 to 25mpg saves 100 gallons, 50 to 60mpg saves 33 gallons.

That is, bullshit. Every metric country uses litres per hundred kilometres, which is the same measurement inverted, and inverted is correct. And before you write this off as a leftover we're stuck with, we did the same thing again this century. Electric cars in Britain are quoted in miles per kilowatt hour. An imperial unit of distance over an SI unit of energy, invented from scratch, in the 2020s, by people who had every opportunity to do literally anything else. It's even the same shape of mistake. A bigger number means less energy used, with the same flattening at the top.

There's no defence available here, every other unit in this post passed a test that this one fails: it did something useful for the person actually using it, something the metric version wouldn't have done as well. Miles per gallon does nothing for anybody. It's a ratio between a unit we use and a unit we abolished, and it isn't easier to say, easier to picture, or easier to act on. It's just what we happened to be using when we stopped paying attention.

Conclusion

So, should we switch to metric?

Yes, obviously we should. I said at the top that the metric system is better and I meant every word of it. If Britain switched everything tomorrow, the roads and the pubs and the scales and the lot, we'd be better off within a generation, and the transition would cost less than the arguing about it already has.

But I don't think that’s the interesting question, and I don't think it's the question people are actually asking when they laugh at us either. Every unit in this post that survived, survived for a reason, and not one of those reasons was heritage. The pint is a serving. The mile is a minute. The stone is coarser than the noise. The inch is a shape. The foot works for humans because its literally based on the human body!

What none of them survived on is arithmetic. Not one of these units is any good at being converted, and it turns out that being good at being converted was never the job. That's the job of the numbers underneath, and underneath, every single one of these is already metric. The pint is 0.56826125 cubic decimetres because Parliament said so, and the number was sitting in the Weights and Measures Act long before anyone thought it needed ‘saving’. The inch is 25.4 millimetres exactly, and has been since 1959. Even the mile is defined in metres!

Britain went metric decades ago. We just kept the old words for the bits where a human being is standing at the end of the transaction.

Which is why I think we get a worse press than we deserve. It looks like a country that couldn't finish the job. It's closer to a country that finished the job underneath and kept the interface it already knew how to use. That isn't a defence of the imperial system, because the imperial system as a system is indefensible and I haven't tried to defend it once. It's a defence of about six words that happen to still be the best way of saying certain things out loud.

Metric is what gets written down. Imperial is what gets said.
I'd switch tomorrow. I'd still ask for a pint.

Thank you for reading.

Further reading

This blog post has mainly been about culture and lightly linguistical. For a more mathematical defence of the imperial system, consider watching https://youtu.be/iJymKowx8cY by jan Misali.

On miles per gallon, autocar has a brilliant post about the struggles of fuel efficiency in the UK. Consider reading https://www.autocar.co.uk/opinion/anything-goes/buying-fuel-litres-measuring-efficiency-gallons-makes-no-sense by Kris Culmer.

Footnotes

I don't have a reason for writing this. Usually there's a hook; a launch, a review, something that actually happened. Yet this time there isn't one. The Made by Google event being as bland and lacklustre as it was probably put me in the mood, but I can't honestly claim that's the reason, because Apple are bland and lacklustre plenty of the time too. I just want to get this off my chest, so here it is.

And before anyone decides what kind of post this is: it isn't a post about Apple being brilliant. I have always said (and I will keep saying!) that Android phones have the capacity to be far better than whatever iPhone Apple has just shipped. I'm not being sarcastic, I do believe it.

The people who defend Android to the death will tell you the camera is better, the battery is better, they can sideload whatever they want, they can pirate Spotify easily, they get real customisation, they get form factors Apple won't touch. I mean, when I say form factors I mainly only mean foldables. But every single one of those reasons is a good reason, and I agree with all of them.

The problem is that nobody gives a shit.

Any iPhone you buy is a good phone

If you buy an iPhone, you have a good phone. That's the whole story, that's where it ends.

Apple have never made a bad iPhone. You can bring up the 5c, you can bring up the 6, you can bring up the XR, you can go all the way back to the 3GS if you want - and I'll go through them, because my point survives all of it.

The 3GS wasn't bad. The 6 might have bent but it wasn't a bad phone, it was actually a genuinely great phone for the time. The 5c is interesting, because the 5c didn't fail on quality. The 5c failed because it had no reason to exist. DankPods on YouTube put this brilliantly in his full length retrospective of the device, but to put it briefly: It was an iPhone 5 in plastic for roughly the same money, without the slow-mo camera or Touch ID or any of the things that made the 5s feel like a step forward - and you could still pick up a real iPhone 5 cheap. The 5 was so good that buying it instead was the obvious move, and most people did exactly that: the 5c wasn't a bad phone, just a really badly positioned one.

And the XR? The XR was the best selling iPhone Apple ever made when it launched, that phone changed their entire strategy. The very next generation, the 11, was the follow-up to the XR and the 11 Pro sat above it. The whole "Pro" split we've lived with ever since exists because of how the XR sold. You cannot look at that and call it a bad phone.

So the point isn't "iPhones are the best phones". The point is that the floor is high. There is no version of walking into a shop, buying an iPhone, and coming out with absolute rubbish.

Samsung will absolutely sell you rubbish

Samsung will make a bad phone and sell it to you without a flicker of shame. Other manufacturers obviously do too, but I'm talking about the big dogs here, anybody can make anything but 'anybody' isnt Samsung and Apple.

Everyone talks about the S26 Ultra, the Fold 7, whatever the current flagship is. Great phones. Genuinely. But are you buying one? Because the phones people actually walk around with are the £200 A-series ones, and I couldn't even tell you the model numbers, which tells you everything about how relevant they are. They're slow, they're bloated, they get a couple of updates and then they're done.

And if you go second-hand to get a good one instead, you're buying an old flagship. Which means it's out of support, or nearly, or it could be fragmented and not in any developer's modern ecosystem.

So yes, there will always be an Android phone better than the current iPhone: But I don't care, and most other people don't either. The world isn't built around the best phone: the world is built around the iPhone.

Social factors to iPhone

Nobody has ever said "ugh, you've got an iPhone". People say the other thing constantly.

iMessage. Game Pigeon. The 0.5 lens, which has been on every iPhone since the 11 and is now just how a certain kind of photo is taken. That's not just a spec, that's become an entire culture in my generation. I went to Iceland with friends in 2024 and some people there had android phones without 0.5 lenses, and I felt bad for them! It sometimes felt like they were missing out, atleast to me. And anyway, even just talking about the regular camera: Snapchat looks better on an iPhone, Instagram looks better on an iPhone. Film a TikTok on Android and it looks worse, and it looks worse even when the phone's camera is better, because these third-party apps can't get at it properly.

Then there's all the boring plumbing that turns out to matter more than any spec sheet. AirDrop. Apple Pay. Sending someone a shared album instead of forty images through WhatsApp. Adding someone to a note instead of screenshotting it. Android has versions of most of this - Google has, Samsung has - but "a version of it exists" and "everyone around you already uses it" are completely different things. For my generation, everyone around me is on an iPhone.

I know the reply to this, because I get it every time: yes, Google Photos also has shared albums. It does. But the majority of people you talk to don't have Google Photos. Why limit yourself to only Android users when there are fewer of them than there are iPhone users? It's a real chicken and egg problem. If nobody has an Android, nobody will buy one, which makes the experience even worse for the people who do have them, which makes the next person even less likely to buy one. It's a constant negative cycle, and no amount of "but the feature exists" gets you out of it.

That's the bit that frustrates me most, honestly. None of this is a technical failure. The phones are fine. The software is fine. It's better than fine in places. It just loses to the fact that everyone else already picked the other thing.

And as a developer

I'd much rather build for iOS. Swift and SwiftUI are really lovely to work in. If I'm asked to make an Android app I don't even reach for a native one. I'm not opening Kotlin or Java, I'm most likely going React Native even with my distain for webapps and getting on with my life. That's partly me, but it's also why third-party Android apps so often feel like the port rather than the product. Because they usually are.

Conclusion

So whenever I see someone defending Android to the absolute death, some part of my brain goes: you don't have a social life.

I know that's wrong. I know it's a bad thing to think. And I want to be clear that I don't mean people who own an Android - loads of people own one for perfectly good reasons and never think about it again. I mean the people who defend it, at length, forever. Because to argue that none of this matters, you have to be in a position where you never need iMessage games, or a shared note, or a shared album, or to AirDrop something to the person sat next to you.

And in a world full of iPhone users, in a country where the iPhone is the default, to not care about these things is... weird. You have to be a little bit of a hermit to not care about the fact that your friends are all on iPhones and you can't play Game Pigeon with them, or you have to be a little bit of a hermit to not care about the fact that your friends are all on iPhones and you can't share a photo album with them.

Again, to be as clear as I can be, because this whole post has been negative and I don't want it read as the wrong thing: Android phones can be better than the iPhone. They often are. The best phone released this year probably won't be an iPhone. But the phones people actually have are frequently bad phones, and there is no such thing as a bad iPhone. They've objectively never made one.

That's my position and, unfortunately, I'm comfortable with it.

- Dia

Introduction

Today as I write this, on Wednesday 12th August 2026, the United Kingdom was set to view the closest eclipse to totality since 1999. This is an awesome event no doubt, but not one that I really planned for. I had heard about this eclipse probably days later than I should have -- i've actually been much more excited about the meteor showers scheduled for tonight at around 2am! But, with a nice park with good views out west not that far from where I live, I thought i'd go and give it a watch anyway.

This eclipse shouldve been brilliant, why wasn't it?

With my naive and unplanned self, I took nothing with me that would have been any use. I brough some sunglasses, my phone and tripod, a bottle of water, and a book to pass the time waiting. Crucially, no eclipse glasses, until today I thought they were just tacky sunglasses that were cheap enough to be given out for free.

After all, why would I need any more than this!! RMG cited that Manchester and Liverpool would have around 90% obscuration of the sun. In my mind it should've gotten really dark, the sun only visible partly (10% partly), and the entire thing would've been a spectacle.

Oh how i was wrong.

What actually showed up to the naked eye was... nothing. I looked up with my sunglasses on and saw, well, the sun. If nobody had told me there was an eclipse today I wouldn't have known. There was no blockage at all! Let alone 90% of the damn thing!

When I got my camera out to try and capture the thing -- after all, this should've been the 'peak time' for the eclipse -- i saw another whole load of nothing. Frantically lowering the focus brightness did next to nothing, it took going into video mode (for some reason) with these focus tricks before anything showed up.

About, a third of the sun was covered? Maybe a bit less? I was entirely perplexed and felt lied to, I was specifically told - we were all specifically told that 90% of the sun was going to be out of sight!

(90% obscuration, properly, means 90% of the sun's area covered. Not some vague diameter measurement, not a rough approximation. A proper thin sliver left over, not the fat bite i was looking at.)

Look, look at this photo from 19:10:03. Literally bang on the dot of the predicted maximum. It's clearly not a 90% crescent.

Look at this way better photo, from 19:18. It still isn't a 90% crescent! It doesn't illustrate my point for this post, I just think it's a better photo.

So whats going on here??

Nerdy camera nerd talk for nerds

Well. In one of my photos i think i've found the answer, and it's caused me to realise that eclipse glasses aren't just cheap sunglasses that can be given away for free. I dont think i've ever actually worn one.

The image below shows a big blown-out overexposed sun, no sign of an eclipse here. However there is a small, separate, green crescent elsewhere in frame. This is what I have learnt is called a 'lens ghost', a secondary internal reflection phone lenses throw off of something this bright.

So then, what's that? A phone lens isnt one piece of glass, its a stack of five or more elements and every single glass-to-air surface in that stack reflects a small slice of whatever light hits it, even with anti-reflective coatings on. Normally that lost light goes nowhere and you never notice; point the thing at something as violently bright as the sun though, and those tiny reflections stop being ignorable. Every so often light bounces off two of those surfaces in just the right sequence to refocus and land back on the sensor as a second, much fainter copy of the sun, usually shifted off to one side and usually tinted because the coatings are tuned to kill reflections at specific wavelengths and whatever slips through comes out green or purple. Thats the ghost in the photo above!

Why does that ghost show something truer than the real image right next to it?? Because each of those bounces throws away most of the light. By the time it lands on the sensor as a ghost its maybe a hundredth, maybe a thousandth as bright as the actual sun sat two inches away in the same frame. Dim enough to actually fit inside what the sensor can record properly, instead of maxing it straight out.

Which is also the answer to why the main image only showed about a third covered, and I dont think 'blooming' on its own quite covers it. It's not that the crescent grew fatter, its that the same stray light bouncing around inside the lens doesnt just form one clean ghost off to the side, it also fogs the area right around the real sun -- a real, named thing apparently, veiling glare. That fog lands right on top of where the moon should have been sat as a hard black edge. The sensor should read that bit as properly black, moon blocking 100% of the light there, but a little bit of scattered light landed on it too, and that was enough to drag it back over the 'this pixel counts as lit' threshold. So the edge of the 'sun' in my main photo was never actually the edge of the moon, it was just the point where the scattered light happened to fade out.

Sorry for ruining the magic if you too got an eclipse photo today.

So: the eclipse itself was exactly as advertised. The phone camera just can't render something that bright without the stray light inside its own lens fogging the shot, and that fog eats into the crescent shape and makes it look far less eclipsed than it is...

Here's another example of this behaviour. At the start of it you can see the green flare flying around as i hastily pan to the sun whilst trying to avoid videoing people. That's (apparently) more true to what the media meant than the bite taken out of the sun visible afterwards.

So was it 90%??

Yeah.. It was there the whole time. If i'd bothered getting actual eclipse glasses like a normal person, i probably would've stood in that park and watched the sun genuinely vanish down to a sliver, same as everyone was promised. Instead i spent forty minutes trying to photograph nothing, on a phone that flatly could not cope with how bright the sun is, and only worked out what actually happened afterwards digging through a lens artifact i didn't even know i'd photographed.

Lesson learned: dont trust an unfiltered phone photo of an eclipse. Ever. For the next one of these that happen i'm buying the glasses first, or atleast finding a watchparty-place that gives them away for free.

Atleast I had an excuse to go to the park.

Thank you for reading.

- Dia

That's it, by the way. The blog post is done.
All that follows are some utterly brilliant eclipse photos taken from all around the North West. Enjoy!

one last thing. the meteor showers i mentioned at the start of the post? unreal. by about 10pm they filled up the night sky, i've never seen so many in my life. apologies if you missed them! they got very overshadowed by the eclipse, pun intended.

goodnight readers!

Pre-Ramble.

Campbeltown has it's roots as a fishing village, however in the early-to-mid 1800s Whisky arrived to the town. It's real claim to fame. At the town's peak, there was over 100 distilleries! Now, granted, only 28 of these were legal, which is where many knowledge sources such as Wikipedia get their number from -- but locals say that there were profoundly more than that.

From Wikipedia:

Campbeltown is one of five areas in Scotland categorised as a distinct malt whisky producing region, and is home to the Campbeltown single malts. At one point it had over 30 distilleries and proclaimed itself "the whisky capital of the world".

Springbank Distillery, Whisky, and a mighty tradition still in operation.

The tour actually had us started in the shop. The email had required that we arrive '15 to 30 minutes early', so we did.. I dont think we needed to. Despite obviously Whisky, which I am sure we will get to later on, they were also selling lots of wood products made out of old Whisky barrels!

Bottle openers, coasters, fridge magnets, keychains, bookmarks, they even had a cheese slicing set. Anything that they could feasibly make out of old aged wood, it seemed they were ready to sell. To the right as you walk in was four sample stations of different whiskys. For free, anybody could try and sample some of their expressions, which i think is insanely kind and something you dont see increadibly often today.

Tricking barley into thinking it's spring.

The tour guide first took us outside to the floor where they malt the barley. It's difficult to get a sense of scale over a picture, but this entire warehouse floor was covered with tiny little barley beads. It was only around 4 to 6 inches tall, as this depth allowed for all of the barley to remain at the same temperature.

Before any of it reaches the floor, the barley spends a couple of days in the steep soaked in cold water, drained, left to breathe, then soaked again. This repeats until it's gone from bone dry to about 45% water. This water allows tiny rootlets to start pushing out, meaning the barly as started to chit.

In the same room they had a manual turner machine that puts the barley at the bottom at the top, and vise versa. Purely for temperature management.

Cool!

What peat is actually for.

After that came the peat. I had absolutely no idea was peat did in Whisky, not a clue. Sometimes labels talk about the 'peat content'?? Yeah. No idea. Well:

Peat, the soiley but hard amalgomation of decaying matter, is what gives smokey whiskeys their smokey taste! Most places use local peat, however due to the location of Campbeltown their peat contained bits of sand which would turn to glass in the whisky making process -- not really ideal. Today, they predominantly get their dry peat from Peterborough and their wet peat from Inverness.

Setting fire to a bog.

The peat and the barley meet in the kiln.

Underneath the drying floor is a furnace, and the furnace is where all of that peat ends up. The floor above it is perforated -- thousands of little holes -- so that the smoke and the hot air rise straight up through the grain sitting on top of it. This is me looking up at that floor from underneath.

The kiln is doing two jobs at once. The obvious one is flavour -- smoke goes into grain, grain eventually becomes whisky, whisky tastes smoky. The less obvious one is that drying the barley out stops it growing. Remember that we spent two days tricking it into thinking it was spring; if you let it carry on believing that, the little plant keeps growing and eats all of the sugar we went to the effort of creating. So the kiln kills it at exactly the right moment, having got what we wanted out of it.

Slightly grim when you put it like that.

This is the bit where the three expressions they do here differentiate! Same barley. Same water. Same building, same people, same everything right up until this point. Longrow goes into the kiln with a long, heavy peat burn and comes out properly smoky. Springbank gets a short one and comes out lightly smoky. Hazelburn gets no peat at all, just hot air.

Three completely different whiskies, separated by nothing more than how long somebody keeps a fire lit.

Ten bins and a bit of slate.

Once the barley has malted, it's brought on old-school conveyor belts to get in malt hoppers. Each one of these 10 malt bins can hold 20 tonnes of barley. It takes 2 months to fill them, followed by a 2 month resting period.

And i want to show you how they keep track of which bin is which:

That's it. That's the inventory system! Two hundred tonnes of malted barley managed with a bit of slate and a stick of chalk. There really is something lovely about a place that could obviously afford a spreadsheet and yet has decided it doesn't want one.

The mill that killed its own manufacturer.

Next the malt gets crushed, which turns it into something called grist. That's all grist means -- crushed malt. You crack the grain open so that water can get at the starch inside.

The machine that does the crushing is a Porteous, and it dates back to the 1940s. This is where our guide told us my favourite fact of the entire tour:

Porteous mills were built so well that they never wore out. Nobody ever needed to replace one. Nobody ever needed to buy a second. So Porteous, having made a perfect product, had absolutely no repeat business, and went bust. Every distillery in Scotland is still happily running the mill it bought eighty years ago from a company that no longer exists.

Look at the little black plate bolted on beside the big brass one. It names the last UK trained and qualified Porteous engineer -- a Mr Alexander Wilson Featherston, of A. F. S. Hull Ltd -- and gives his office number and his mobile.

A working distillery has one man's mobile number screwed to the side of its mill because he is the only person left who knows how to fix it. I love it here.

The mill has one job, an average batch is 3.4 tonnes, and the grist coming out of it should be 20% husks, 70% middles, 10% flour.

Which is meaningless until you know what happens next, so:

• husks are the shells
• middles are the coarse chunky bits from the middle of the grain
• flour is the fine powder

In the next stage the crushed malt sits in a big tank and hot water gets drained down through it. The husks form a natural filter bed that lets the liquid pass through. Too much flour and the whole lot turns to wallpaper paste and clogs solid, though too much husk and the water rushes straight through without picking anything up. 20/70/10, every time, from a machine older than most of the people looking at it.

Sugar water.

The tank in question is the mash tun, and Springbank's is a big open-topped cast iron thing painted post-box red. Almost everywhere else has moved to sealed stainless steel with a lid on it. This one you can walk up to and look into:

Those rotating arms with the teeth on them are the rakes, and they stir the mash around. The perforated plates on the floor are what the liquid drains out through.

The sweet liquid you drain off is called wort, pronounced "wert". It's only sugar water, just loads and loads and loads of it.

They put four waters through it, and the numbers are on a sign on the side:

The whole cycle takes about ten hours.

The temperatures going up like that isn't arbitrary. The first and coolest water is the one doing the actual work, because the enzymes turning starch into sugar are alive at 63.5°C and would be killed stone dead by anything much hotter. The later, hotter waters are just rinsing out whatever sugar is left. And because the third and fourth come out weak, they get kept and used as the first water of the next mash instead of being thrown away.

The soggy grain left at the bottom is called draff, and it goes off to local farms as cattle feed. Genuinely nothing is wasted in this building.

Beer, basically.

Now we add yeast, and once you add yeast to sugar water you are making beer. That is not a joke or a simplification -- what comes out of this stage is a flat, unhopped, slightly sour beer at around 6%, and if you drank a pint of it you'd know exactly what it was.

It ferments in washbacks, which is the industry's word for "big tub". Springbank has six of them, and they're made of wood:

Twenty one thousand litres of wort and 75kg of yeast go into each one, and it sits there for a minimum of 72 hours and up to 110.

Both of those details are deliberately awkward choices. Wood is porous and impossible to properly sterilise, so bacteria live permanently in the timber. Dont think of it as a hygiene failure, it's a flavour source. And 110 hours is an enormously long fermentation; most distilleries run about 48. Once the yeast has finished its work, those resident bacteria keep going and produce all sorts of fruity, funky, complicated flavours.

Steel tubs and short ferments would be cleaner, faster, cheaper, and would make a duller whisky.

The still house.

Distillation sounds complicated and isn't. Alcohol boils at a lower temperature than water. So you heat up your beer, catch the vapour coming off it, cool it back down into a liquid, and what you've caught is stronger than what you started with. Do it again, it gets stronger again.

Springbank's wash still is direct fired, meaning there is an actual oil flame burning underneath the copper. Practically nobody does this any more because it's fiddly and it scorches the wash slightly -- and the scorching is precisely the reason they keep doing it.

There are some words to get through here, but they're all straightforward:

• low wines -- the weak spirit off the first distillation, about 20-25%
• foreshots -- the harsh, volatile stuff at the very start of a run
• feints -- the weak, oily stuff at the very end of a run
• middle cut -- the good bit in between, and the only part that becomes whisky

The foreshots and the feints don't get binned, they get tipped back into the next run. And the decision about exactly where to stop taking the foreshots and start taking the middle cut is made live, by hand, by the person standing there.

Then there's the diagrams on the wall, which explain Springbank's whole strange setup better than i can:

Longrow is distilled twice, which is normal. Hazelburn is distilled three times, which is actually the traditional irish way to do it! That's why irish whiskeys tend to be more floral or fruity.

Springbank is neither traditionally scottish nor irish, as it's distilled two and a half times, which is nonsense, and which nobody else in Scotland does. It goes through the wash still, then the first low wines still, and then only part of it goes through the second one -- the diagram breaks it down as 80% feints and 20% low wines. That partial third pass is the half.

The locked box.

This is the spirit safe, and it might be my favourite object in the whole distillery:

Brass and glass, padlocked shut. Historically, the only key was held by the Customs and Excise officer.

The reason is tax. Everything flowing through that box is untaxed alcohol, so the man actually making the whisky was not permitted to touch it. He could look at it, and he could test it with hydrometers through the sealed case, and that was it. Hence the glass -- the entire object exists so that a distiller can do his job while being physically prevented from getting at the product.

There's a desk right next to it with the handwritten record books on it, why use a computer when a book does the job.

There's also a sign nearby with the year's numbers on. 264,000 litres a year, of which 80% is Springbank, 10% Longrow and 10% Hazelburn. And it refers to what's coming off the still as plain British spirit, because legally that is what it is. It cannot be called whisky until it has sat in an oak cask in Scotland for three years.

Into the barrel.

Every cask gets filled at 63.5% alcohol, and they fill about sixty a week.

J & A Mitchell & Co Ltd -- the family who have owned the place since 1837 -- 2026, Springbank Single Malt, Campbeltown, cask number 1444. That means its the 1444th barrell made this year! The 'A' means that this is the first time the barrel has ever been used, as they can only be used twice. The wood of 'B' barrels still do the job of flavouring well enough, but a 'C' barrel (having now been used twice) is useless for whisky production. They're predominantly used as storage barrels or flowerbeds.

Where it sleeps.

These are dunnage warehouses, which is the old way of doing it. Earth floors, thick stone walls, and casks stacked only three high on wooden rails. Cold, damp, and very stable in temperature. The modern version is a racked warehouse -- steel, twelve casks high, dry. Dunnage is worse in every single commercial respect and better for the whisky, which is a sentence you could apply to this entire distillery.

About 2% of every cask evaporates each year. That's the angel's share, and it's why a warehouse smells the way it does.

Our guide described the relationship between the spirit and the wood as a marriage, a honeymoon, and eventually a divorce. The marriage is the day it goes in the cask. The honeymoon is the years where the wood is giving it colour and vanilla and spice and the whisky is getting better every single year. The divorce is when the wood has given everything it had and starts taking over instead -- the whisky goes dry and tannic and over-oaked.

The entire skill of the job is bottling it before the divorce.

The bit i can't show you

We did get into the bottling hall. I just can't show you any of it -- phones away, no photos, because unreleased bottlings sit in there and they'd rather one didn't turn up online before they've announced it. Entirely fair.

You'll have to take my word for it, then, which is annoying, because bottling is the part most people don't realise is unusual. Nearly every distillery in Scotland tankers its whisky off to an enormous central bottling plant somewhere in the Lowlands. Springbank does it here. It was actually a brand new hall, only opened a couple of months back.

That is what "the only distillery in Scotland that does everything" actually means. The barley is malted here, distilled here, matured here and bottled here, and none of it leaves the walls until it's in a bottle in a box.

Next door.

On the walk back we passed the sign for Glengyle, which is owned by the same family and makes a whisky called Kilkerran. The distillery and the whisky have different names because the Glengyle name was already spoken for elsewhere.

It was revived in 2004, and it's a decent part of the reason Campbeltown still counts as a whisky region at all. You need more than a couple of working distilleries to justify being a region, and for a while there it was touch and go.

And then: the whisky.

The tour finishes in the Washback Bar with a tasting, where you can choose between Springbank or Longrow. They give you a single for free to try, and a double in a bottle to go with it. They also have single bottles as driver's drams should you not be able to drink a shot of whisky at midday.

Having spent two hours watching cold water, a chalk slate, a bankrupt company's mill, a locked brass box and a damp stone shed all conspire to produce it, i will admit the Longrow tasted quite different to how it would have done that morning.

Didn't get a photo of the glasses full, like. Woe is me.

On the way out we spotted Cadenhead's, just down the road:

Cadenhead's is the oldest independent bottler in Scotland, founded in 1842, and - of course - owned by the same Mitchell family.

What got me about the whole thing wasn't any single piece of it. It's that at every point where there was an easier, faster, cheaper way of doing something, Springbank has looked at it and decided not to. Malt it by machine somewhere else. Ferment it in steel. Use a modern condenser. Tanker it to the Lowlands to be bottled. Every one of those is available, and every one of them would make a duller whisky, and they've said no to all of them.

Everything you drink out of that place is the result of somebody deliberately choosing the harder option, over and over again, for nearly two hundred years.

If you're anywhere near Kintyre, i'd reccomend visiting. It's only £15 per person and you leave with more than £15s worth of whiskey and a glass!

Thank you for reading.

- Dia

Why web?

is a question that im sure, even subconsciously, you've thought of whilst browsing this site.
By 'web' i mean the personal web -- the "Indie Web" as it has come to be known. The idea that everybody can own and host a space for themselves and their thoughts. The idea that anybody can choose how that space looks and feels, and customise the ways in which their content can be viewed. Maybe you want to post photographs for your friends to see, maybe you want to keep track of all of your thoughts and feelings, maybe you want to archive your life even if it's only yourself who is going to see the archive. The web lets you.

Issues with centralisation.

Today, and quite rightly, most people would solve this by signing up to one of the thousands of companies happy to take your time and your data on the promise that you can do these things outlined. It could be an instagram account, a private twitter. It could be an online drive service offered by companies like google and microsoft. The modern way of using the net is centralised, the idea that everything you do is put in one place, operated by one company, on the promises of ease, discoverability, and zero hassle.

In reality, nobody likes these companies. Nobody likes twitter, nobody liked twitter even before elon got involved. Nobody particularly enjoys instagram or discord, or even sites like twitch. We use these services because they are easy, they do make us discoverable, and they are relatively zero-hassle. It helps that many of these sites are free to use! But we still dont like these platforms. It doesnt feel like our accounts are ours, it feels like our accounts are a depiction of ourselves that we've set up to play the system. Anybody who has mucked around with their discord profile to get everything just visually right - bio length, colour choice, etc - has done this without realising.

Furthermore, all of these platforms contain algorithms at the heart of how they work. All of them show you content not in the hopes that it's content from your friends or family, or content that you'd even particularly enjoy. It's solely content that keeps you on the app. Content that makes you addicted.

Our brains, believe it or not, need silence. They need moments of doing nothing. They need periods of wait.
In a modern algorithmic world, we are controlled by these companies to never have this time. The minute we feel bored, we reach for tiktok or twitter or instagram. Our brain is never given that time to wait.

Even sites that dont offer this addictive content stream, such as twitch and longform youtube, still feel icky to many people -- me included. I dont feel like i'm watching this stream on Twitch because it's on Twitch, i feel like i'm watching it on Twitch because the streamer had no other choice. Because Twitch is where everybody is, it's where everybody continues to be: no matter the individual's opinion of Amazon.

It's all well and good complaining and educating about these things, but the fact of the matter is that it's obvious, and common knowledge. Everybody knows that humans need to go outside more and spend less time on screens, everybody knows that constant content consumption is bad for you -- but nobody does anything about it. At the end of the day, we still open tiktok, or instagram, or youtube, or twitter, or one of the other thousands of addictive sites, simply because we feel as if we have no other choice.

If you're into reading, i'd recommend The Dose Effect by TJ Power. If you're not, know that he explains how slow dopamine is extremely important to human growth, and today we have almost none of it. All of the dopamine we get is quick release: eating a slice of cake, placing an order online, betting on live sports, checking a phone for notifications. Today, we dont have slow dopamine releases almost at all, and when we do they are often mitigated and downplayed by all of the quick-release dopamine surrounding it.

So, with all of this said, and it has been a lot, i return to the question asked in the title of this post. Why web? Why bother styling a website in utterly horrible to deal with html and css. Why bother sludging through a website builder or learning how to host a site. Why bother paying the money for a domain name?

Why the effort is worth it.

I wasn't alive in the days of the early internet, so bear with me as i reminisce about a past that i never experienced, but in the world where everybody owned their own space and could do with it what they please -- people were happier online.

There were issues with those days, there are issues with this model all around! Accessibility, exclusivity, discoverability are all worse and regularly completely missed when it's down to one person to make and maintain a system. And it wasnt some peaceful utopia either - people have been having furious arguments with strangers online since before the web existed, and plenty of people lost entire years to IRC and text adventures.

But none of that was designed. Nobody built a flame war on purpose, and nobody was measuring how long they could keep you in a chatroom. You logged on when you wanted to catch up with friends the same way you'd choose to walk into a pub or a park and meet up with people in real life. If you didn't want to, you stayed offline -- after all, you wouldn't want to be clogging the phone line up for no reason!

Under this system, it's not just that you couldn't experience the negatives to centralised social media -- those negatives couldnt even be fathomed. There were no damaging comparisons to be made. There was no constant stream of information being battered into your brain 24 hours per day. You got the information you wanted to see, and no more. You chose everything you saw. Under this system, you would find new content through recommendation from friends. You would find new content through email lists, email forums, or chatrooms. You would find new content from people. And in the world of Artificial Intelligence, doesn't that just sound lovely.

Somebody has to start somewhere. This utopia only works if everybody has a place to put their stuff -- if everybody has their own place that can be bookmarked and checked regularly by others. It doesnt matter how many people see it, there are no likes, there are no retweets. It is remarkably hard to appreciate how much those numbers shape you until you spend a while somewhere they dont exist. You write the post you wanted to write, and that's the end of it. Comments, were they not included on the web page, would be sent through email. They weren't public, it was one person talking to another person about something that they had made.

That sounds amazing.

To me, yes, all of this is more effort. But when has the better outcome ever been the easiest? There is no movement without, well, a movement. If you can take one thing away from this blog post, please have it be a raised awareness of how rotted our generation has become. Actually think about the word 'brainrot' for one moment, think about what those two words together imply.

Why is your own website the best way to do it?

There's a fair objection to be made here, which is that you can already do most of this somewhere else. Substack, Medium, Tumblr, Neocities, free static sites on somebody's hosting -- these already exist and some of them are genuinely lovely. But most of them are still somebody else's house, and the terms of your stay are still written by somebody who isnt you. The difference between those and your own site is really small... until the day it suddenly isnt.

Nobody else gets to decide what it looks like. Remember the discord profile with the bio trimmed to fit and the one colour picked as a header (provided you dont pay, of course!!). That's what expression looks like inside somebody else's box -- on your own site there is no box. If you want your writing in comic sans on a background of your cat, that is a decision available to you, and it will be more you than any profile page has ever managed.

It survives, too. Vine is gone. Google Reader is gone. Tumblr deleted half its culture overnight in 2018 to please an app store. Twitter's API went from free to unaffordable in a week and took a decade of tools with it. None of these were announced in advance, and in none of these cases did the people who'd spent years building something there get a say. If your host turns nasty, you copy your files somewhere else and point the domain at it. The address stays yours and nothing you've made goes with them.

The effort is a feature, not a bug. It's genuinely a faff to write a post, style it, and put it live, which means you only do it when you've actually got something to say. You cannot post to your own website out of boredom. That friction is the thing every platform has spent twenty years and billions of pounds engineering out and it's the thing that keeps what you make worth making.

It's also a place, rather than a feed. A post on twitter is a thing that happens to you and then is gone. A website is somewhere people go. It can be bookmarked, it can be put in an RSS reader, it can be linked from somebody else's page by someone who thought you were worth linking to. It sits still long enough to be found on purpose.

And -- my personal favourite point -- you'll have built something. This is the slow dopamine i was banging on about earlier. Learning enough HTML and CSS to make a page that works is a reward that takes weeks rather than seconds and the site itself ends up being a record of you getting better at it. You dont get that from choosing a theme.

None of this makes it easier. It makes it yours, which is a different thing entirely and, i'd argue, the more valuable one. So make one. It doesn't have to be good.

Thank you for reading.

I welcome your comment, i truly do. The indie web is about connecting humans with humans; something we might be lacking in today's age. Send me an email, DM me over discord, or leave a comment right under this post if that feature is available. Let me know what you think of this blog post and any blog post to come. If you found me linked on somebody else's web page, let me know that too! Sticking together might be all we have in these horrible times.

- Dia

Hello World

If you're reading this, the blog works. That's the whole post.

Secondary title!

The web used to be fun. Let's keep it that way.

Second secondary title... ?

Tertiary

Quaternary

pent-er-ery?

hexerenery?

normal text _subscript

normal text ^superscript

^{_{Image credit: Nina Amigo Canino}}

- Dia