If you ever needed an example of the increasing importance of personal brands in the AI era, I’ve got a whopper. 

Over the past few months, I’ve noticed something that construction professionals and jewelry makers on Etsy have in common: they’re always looking out for better ways to sort and store their small parts. This seems to be a universal struggle for anyone who makes or repairs things — whether it’s their job or hobby.

You would think this is a solved problem by now. It’s just a plastic box with plastic boxes inside of it. You can go into a big box store and find small parts organizers for as low as $5 on sale. 

If you want one that’s made from materials that won’t eat away at your parts over time, you may pay around $20. 

If you want one that will survive years of working under harsh conditions at outdoor job sites, you’ll pay $50 to $70.

But if you want the Adam Savage signature model box, it’ll cost you $120 plus shipping — if it’s in stock.

Often, it’s out of stock and eBay sellers list it for around $200. It sells out fast at both prices. This is why you are limited to 10 boxes when ordering at Adam’s website. He’s trying to discourage resellers.

Adam has spent decades cultivating an audience. These days he makes YouTube videos in his personal workshop, where he makes stuff — anything he feels like at the moment. His videos are not flashy. He speaks openly and honestly about life, career, hobbies, failures, successes, and parenting.

This has built so much trust with his audience over the years, that he can command up to $200 for a plastic box.

This is extraordinary when you consider the needs of 90% of his audience could likely be met by a $5-$20 box from a big corporate brand, without limitations on order size and without added shipping charges.

This is why personal brands are not only a real thing, they are the realest thing right now.

People pay for a narrative. People pay based on trust and relationships. 

The biggest companies on Earth are trying desperately to automate this kind of trust into existence right now, under faceless brands. How do you think that’s going to turn out?

This is really dumb.

So if you use regular paper clips to stick pages together, you’re probably used to them being slightly bent out of shape after using them a few times.

Then I try to make them last longer or using them while they’re not really that useful anymore.

Why am I doing this?

They are cheap as fuck. And there is still like 2/3 of them in the box of 100 I bought six months ago

If you live in the United States, you might already be familiar with the quirks of territorial sovereignty (though actually most Americans don’t know this 😬) After all, millions of Americans live in places like Puerto Rico, Guam, the U.S. Virgin Islands, American Samoa, and the Northern Mariana Islands.

Subscribe now

Now, while these territories fall under U.S. sovereignty, the day-to-day reality of their residents is distinct from those living in the fifty states. For example, residents of Puerto Rico are U.S. citizens, but they don’t pay federal income tax on locally sourced income (though they do pay Social Security), they also lack a voting representative in Congress, and they can’t vote for the President in the general election. It’s a nuanced relationship, bound together by a flag and matters of defense, but otherwise separated by layers of constitutional law. It can get complicated and it often doesn’t work out for the territories all that well.

But that’s the United State’s territories. Which, while complicated, if you look across the Atlantic, you’re gonna find a system of geographic and political relationships that makes the U.S. territorial system look downright simple. I am, of course, talking about the vast network of continued territories and Crown Dependencies that exist under the umbrella of the United Kindom… sort of.

The British have had centuries longer to complicate things and oh boy have they taken advantage of that. So to understand the modern British footprint, we have to untangle three distinct categories: The United Kingdom, the Crown Dependencies, and the UK Overseas Territories.

Here is a breakdown of what these regions are, how they function, and, most importantly, why these distinctions matter to the people who call them home.

The United Kingdom: the core state

Let’s start with the anchor. The United Kingdom of Great Britain and Northern Ireland (to use its full, very long name) is a sovereign state in the world. In this way, the United Kingdom is no different from the United States, or Canada, or Germany, or China, or Kenya, and so on. They have their borders. Within those borders they make laws. Outside those borders they deal in treaties and, sometimes, even war.

And, in a similar fashion as the United States and Australia have states and Canada has provinces, the United Kingdom has its, well, kingdoms. Specifically, the kingdoms of England, Scotland, Wales, and Northern Ireland. Now, I’m not here to say that these are the same as the state of California in the U.S. But, for simplicity and comparison’s sake, they at least act in a similar manner under a much larger umbrella of a single government.

That government is, of course, the UK Parliament in Westminster (London). While Scotland, Wales, and Northern Ireland have their own devolved governments that handle local issues like health and education, Westminster retains ultimate authority over the entire country.

If you live here, you’re likely a full British citizen. You pay UK taxes, you’re subject to UK laws, and you vote for a Member of Parliament (MP) to represent you in the sovereign government. There are, of course, people who have different visas, but you get the picture.

This is relatively simple stuff as far as national sovereignty explanations go so, let’s start getting weird.

The Crown Dependencies: close to home, but a world away

Just off the coast of Great Britain lie three island territories: the Bailiwick of Jersey, the Bailiwick of Guernsey (both in the English Channel), and the Isle of Man (in the Irish Sea).

Here’s the twist though, they are not part of the UK. These are considered “Crown Dependencies.” This means they’re self-governing possessions of the British Crown (the Monarch), but not the UK government.

They have their own entirely independent parliaments (like the Tynwald in the Isle of Man, which claims to be the oldest continuous parliament in the world). They pass their own laws, set their own taxes, and manage their own economies.

The UK is only responsible for their defense and international representation (like signing treaties), and even then, the UK is supposed to consult them first. I can’t really speak for how often they do, but they should be I guess.

Because of this unique arrangement though, they are famously low-tax jurisdictions. There is no capital gains tax or inheritance tax, making them highly attractive hubs for global finance and retreats for the wealthy. Residents also live under local laws that can differ significantly from the UK. Even things like speed limits are different from the UK proper.

But perhaps most importantly, islanders fiercely protect their independence. While they carry British passports (with slight variations on the cover), they don’t vote in UK elections, and UK laws do not automatically apply to them. Not unless they independently pass their own legislation that mirror those from the mainland.

This would be as if Long Island in New York State was something called a Presidential Dependency. It had the President of the United States as its figure head, but Congress had no actual legislative jurisdiction over it. Weird right?

But there’s one more level to go down in this rabbit hole.

UK Overseas Territories: The British Empire lives?

Scattered across the globe are 14 UK Overseas Territories (OTs). These are the last remnants of the British Empire. They range from the wealthy financial hub of Bermuda and the contested Rock of Gibraltar to the remote Falkland Islands and the tiny, volcanic Pitcairn Islands in the Pacific.

There’s that famous saying from colonial times: the sun never sets on the British Empire. I actually think this might still be true. I haven’t fully mapped it out with all times of the year, but given the geographic dispersion of OTs, it certainly feels like.

Now, like the Crown Dependencies, these territories are not part of the UK proper. They are internally self-governing, with their own elected assemblies, constitutions, and local laws. And like with the Crown Dependencies, the UK retains responsibility for defense, foreign affairs, and internal security.

But unlike the Crown Dependencies, the UK government appoints a Governor to each territory, who acts as the de facto head of state on behalf of the Monarch. Crucially, the UK Parliament theoretically retains the power to legislate for the OTs, though it rarely exercises it without consent.

Since 2002, most residents hold full British citizenship, allowing them to move to and work in the UK freely. However, this is a one-way road more often than not. For example, those who are born and live in the Falklands can move to the UK to work and live with no issue. But those who are born and live in the UK can’t move to the Falklands and work and live. They need a visa just like anyone else.

And speaking of territories like the Falkland Islands (or even Gibraltar historically) the UK’s commitment to their defense is existential due to territorial claims by neighboring countries (Argentina and Spain, respectively).

Despite being subject to ultimate UK authority on foreign policy and defense, residents of OTs have no representation in the UK Parliament. When Brexit happened, Gibraltar and the Falklands were deeply affected, yet their residents had no vote in the referendum (except for Gibraltar, which was granted a special inclusion).

Why the distinctions matter

For the average person in London, the difference between a Crown Dependency and an Overseas Territory might just seem like pub trivia. But for the people living under these agreements, these constitutional boundaries dictate their daily lives.

1. Economic Survival: For places like the Cayman Islands (an OT) or Jersey (a Crown Dependency), their ability to set their own zero-or-low-tax policies is the lifeblood of their economies. If they were fully integrated into the UK, their economies would likely collapse and need to be subsidized heavily by the UK.

2. Democratic Voice: Residents of OTs and Crown Dependencies live in a democratic gray area. They govern themselves locally, but the ultimate guarantor of their security and international standing (Westminster) is a body in which they have zero elected representation.

3. National Identity: These agreements allow people to be proudly Bermudian, Manx, or Falkland Islander first, and British second. It provides the security of a major world power without the total loss of local culture and autonomy.

The British system is a patchwork quilt, stitched together over centuries of conquest, trade, and diplomacy. It’s messy, contradictory, and largely held together by tradition and gentleman’s agreements. But for the millions who live on the fringes of this post-colonial web, it’s the delicate balance that defines their place in the world.

And it certainly makes U.S. or Canadian territorial arrangements simple by proxy.

Share

This is the second part (I, II) of our somewhat silly look about the plausibility of warfare in Frank Herbert’s Dune. Last week, we looked at the system of warfare that is dominant in the setting when the first book opens: warfare among the Great Houses. While I noted some worldbuilding issues I see – some of the physics doesn’t quite work out, I don’t think lasers are satisfactorily dealt with and the implied social system doesn’t seem even remotely stable– we’re going to accept for this part that the system works more or less as Herbert describes it.

The various Houses (Major and Minor) maintain relatively small militaries of trained close-combat fighters who fight using shields. Because shields reduce the effectiveness of ranged combat nearly to zero, this system of warfare dominates among the Great Houses and because untrained, unshielded fighters are so profoundly vulnerable to trained, shielded ones, outside military challenges to this system are generally unsuccessful, enabling the small, closed and mostly hereditary elite with their retinue-armies of shielded fighters to maintain a stranglehold on political and military power. They use that power to run relatively inefficient patrimonial ‘household’ governments over entire planets, siphoning off what little economic production they can – because their administration is so limited – to fund their small armies.

What keeps the armies small is both that the resources of the Great Houses are limited – again, small administrations – but also that the core components of industrial military power in this setting (trained fighters, shields, ornithopters, frigates) are clearly very expensive, both to build and to maintain. And as an aside, because it will be relevant below, it is clear even in the books that wear and tear on shields is a major cost: “The Harkonnens certainly used plenty of shields here, “Hawat said. “They had repair depots in every garrison village, and their accounts show heavy expenditures for shield replacement parts.” (Dune, 88, emphasis mine). In short, these elements of military power represent ongoing expenditures, requiring maintenance and logistics which is going to matter a bit below.

This week we’re going to look at how the Fremen disrupt this system and ask if the Fremen success in doing so seems plausible. We’ll do so generally accepting Herbert’s clear description of the Fremen as superlative warriors, even though long-time readers will know that I find the idea of the Fremen being such superior warriors broadly unlikely. But as we’ll see, even if the Fremen are remarkably skilled warriors, they are unlikely to succeed in their jihad against the society of the Known Universe.

But first, as we’re going to cover below, equipping a fighting force with Dune’s version of modern military power – shields, ornithopters and frigates – is expensive. If you want to help me equip a Great House of trained fighters to challenge the Imperium, you can support this project over at Patreon. If you want updates whenever a new post appears or want to hear my more bite-sized musings on history, security affairs and current events, you can follow me on Bluesky (@bretdevereaux.bsky.social). I am also active on Threads (bretdevereaux) and maintain a de minimis presence on Twitter (@bretdevereaux).

Wars of the Fremen

We should start just by outlining exactly what the Fremen do, both what we see in Dune and what we are told about in Dune Messiah.

The Fremen are, at the time Dune begins, the native population of Arrakis and we are told there are about 15 million of them. They maintain some small levels of industry – mostly things which can be rapidly moved – back lack large industrial systems and notably lack the ability to produce any of the elements of industrial military power (shields, aircraft, frigates) essential to the warfare of the Great Houses, though they do at time capture and use this equipment.¹ The Fremen are already highly capable warriors, but because they lack these elements of industrial military power – especially shields – it is easy for the militaries of the Great House to oppress them. In particular, the Fremen have no defense against laser weaponry, which is devastating against unshielded opponents.

When Paul arrives, he organizes the Fremen for what is initially a classic protracted war campaign against the Harkonnen occupation, which eventually sufficiently disrupts spice production to bring the emperor himself to Arrakis. The result is something of a science-fiction rerun of Dien Bien Phu: the foreign occupier, convinced that his industrial military renders him unbeatable in a conventional engagement intentionally and arrogantly extends his force into enemy territory only to be cut off and defeated.

A few things make this Fremen success work. First, the Fremen operate from a terrestrial base that their enemies cannot attack effectively (the deep desert). The Fremen also operate with tremendous local knowledge: because they are the indigenous population, it is easy for their agents to infiltrate into the settled zone the Harkonnen control, meaning that the Fremen have good visibility into Harkonnen operations even before their leader becomes a prescient demigod. Perhaps most importantly conditions on Arrakis negate most of the advantages of industrial military power. As Hawat notes, ornithopters suffer substantial wear-and-tear on Arrakis, making it expensive (but not impossible) to maintain large fleets of them; shields too apparently are hard to maintain. The large sandstorms that rage basically anywhere except in the small area protected by the ‘Shield Wall’ mountain range (which is where all of the cities are) can disable shields at almost any scale. But most of all, shields attract and drive mad the large local sandworms, making their use on the ground in the open desert essentially suicide.

Consequently the Fremen able to win in part because they occupy the one place in the whole universe where the military ‘package’ of the Great Houses does not work.

And to be honest, I do not find the way the Fremen win on Arrakis to be wholly implausible. Given their mastery of the local terrain and infiltration of the local population, it makes sense that the Fremen would be very hard to uproot and might steadily bleed an occupying force quite badly over time. At the same time, the idea that Shaddam IV and House Corrino might – somewhat arrogantly – assume they that could safely extend themselves down to the surface is the sort of military error regular armies make all the time. Finally, it also makes sense that the Harkonnen and Corrino armies coming to Arrakis might fail to adapt to Fremen warfare – fail to adapt to warfare without shields, for instance – because they do not perceive their primary security threat to be the Fremen (the Harkonnen, we’re told, consistently underestimate how many Fremen there are). So while they should respond to the Fremen with guns and artillery, it makes sense that initially they respond with the sort of armies that work for all of their other problems: trained melee fighters with shields.

And if – again, we’re accepting this for the sake of argument – if the Fremen are the superior close-combat fighters, the result of that effort might well go this way. Especially with a prescient leader pushing them forward to victory. Crucially, the victory at Arrakeen fundamentally depends on these local factors: Fremen knowledge of terrain enables Paul to mass his forces undetected and observe the Corrino disposition safely and to thus to stage a coordinated surprise attack against his opponents. Sandworms enable him to deliver an attack force rapidly through a sandstorm and the storm itself disables the defender’s shields, enabling him to disable their frigates and also neutralizing much of their airpower. Fremen victory is almost entirely reliant on factors unique to Arrakis.

So that is more or less fine. The problem I have is really with everything that happens next.

What Happens Next…

I think we should be clear what Dune and especially Dune Messiah lead us to understand comes next to avoid unnecessary wrangling in the comments. While we do not see it, the Fremen wage an absolutely massive, known-universe spanning war in which they conquer thousands of worlds and kill sixty-one billion people (the statistic given in Dune Messiah).

Equally, we are supposed to understand that this result was inevitable. Indeed, this is one of the central themes of Dune, that by the time Paul’s prescience has developed sufficiently for him to understand the road to his Jihad, it is already too late to stop it. As we are told of Paul’s thoughts, “He had thought to ppose the jihad within himself, but the jihad would be. His legions would rage out from Arrakis even without him. They needed only the legend he already had become.” Just after, right before his duel with Feyd, he thinks, “from here, the future will open, the clouds part onto a kind of glory. And if I die here, they’ll say I sacrificed myself that my spirit might lead them. And if I live, they’ll say nothing can oppose Muad’Dib” (Dune 482, emphasis original). The point is the Jihad happens either way.

I want to stress that: even without Paul Atreides’ prescience, the Jihad happens and at the very least burns across the known universe doing massive destruction; in fact, even without Paul the Fremen win.

That position – that the destruction of the Fremen Jihad is not merely possible but inevitable to the point that Paul cannot stop it – puts a very, very high bar on its military plausibility. In particular it rules out any defense that Fremen victory is simply because Paul, as a prescient military leader, can simply pull an endless series of ‘inside straights.’ Remember: the Fremen explicitly still win even in Paul Muad’Dib Atredies is dead at the hands of Feyd Rautha Harkonnen. It is not enough for it to be possible for the Fremen to win, it must be impossible for them to lose.

Now in the thematic world of Dune, that is because military victory is fundamentally a product of the Fremen Mirage: societies have an inherent vitality to them and the Fremen are vital, hardened by the harshness of Arrakis, in a way that the Great Houses are not. In Herbert’s mind, that is enough: the ‘hard men’ created by the ‘hard times’ of Arrakis will inevitably triumph once an event – the emergence of Paul as a heroic figure – spurs them into action. Paul is thus die Weltseele zu Pferde, “the world-spirit on horseback,” the archetypal ‘great man of history’ who embodies supposed historical forces which are larger than him, which act through him and which would act without him.

Except of course the problem is that both the Fremen Mirage and the Great Man Theory of history are, to put it bluntly, rubbish– grand historical narratives which simply do not fit the contours of how history actually works. ‘Hard men’ from ‘hard places’ and ‘hard times’ lose all the time. Societies only seem ‘vital’ or ‘decadent’ when viewed in retrospective through the prism of success or failure that was contingent, not inevitable. History is full of movements and moments which cannot be explained through the agency of ‘great men.’ There is, in fact, no ‘world spirit’ guiding history like an invisible hand, but rather a tremendous number of contingent decisions made by billions of people with agency acting with free will.

So rather than simply assume that because the Fremen are moving with the ‘universe spirit’ of history as it were, that because they are a vital people, because they are ‘hardened’ by Arrakis, that they win by default, we’re going to ask are the Fremen actually likely to win in their Jihad? Remember: the books present this not merely as likely but inevitable. Is it likely?

Oh my, no.

The War With the Great Houses

I think we actually want to think through this conflict in two rough phrases. Initially, the Fremen leaving Arrakis are going to be confronted by the traditional militaries of the Great Houses. We’re never told how many Great Houses there are, but it is clearly quite a lot – the institution still very much exists in God Emperor of Dune despite the fact that we’re told 31 Houses Major (the upper-rank of the Great Houses) had collapsed. The implication is that 31 Houses Major do not represent even a majority. Likewise, the entire political system of the Corrino Imperium only works if the Houses of the Landsraad collectively had more military power than the Corrino Sardaukar, such that the emperor had to keep them divided at all times (and such that, acting collectively, groups of them might force concessions from the emperor). Given that Baron Harkonnen thinks just two legions of Sardaukar could easily overwhelm his entire offensive force of ten legions, the implication has to be that there are quite a few Houses Major with military forces on the scale of House Harkonnen.

In short the Fremen are likely to be faced by many dozens of ‘House armies’ ranging from the high tens of thousands to the low hundreds of thousands, probably collectively representing several million trained fighters with shields (I’d guess a few tens of millions, once Houses Minor are accounted for), ‘thopters,’ frigates and all of the other components of ‘modern’ (for the setting) warfare.

The main advantage the Fremen have – and it is a very significant advantage – is that their control over the Spacing Guild (via control over the spice on Arrakis) means that they can face these forces one-by-one, rather than having to face a large coalition of the Landsraad all collected in a single location. The secondary advantage the Fremen have is that the Great Houses are likely to try to meet them with the same rigid, formulaic armies they have long prepared for use against each other: trained fighters using shields engaging in melee combat. They will probably not be, in the first instance, rapid military innovators – they aren’t set up for that.

But the disadvantages the Fremen face are enormous. First and foremost – and this is going to be central – Fremen manpower is fundamentally brittle. On the one hand, the Fremen do not have a civilian class – all of their people are trained fighters, so basically their entire adult population is available for combat. The problem is that means that there is no underlying ‘peasantry’ as it were to refill the ranks of their losses and the harsh conditions of Arrakis – essential to the entire Fremen thing – are not conducive to a ‘baby boom’ either. Fremen losses will thus be functionally permanent: every Fremen Fedaykin lost is lost forever – a long-term reduction in the total Fremen population and thus available Atreides military force. Meanwhile, Hawat estimates the total Fremen population at roughly 10 million. That represents a fundamentally finite resource which cannot really be replenished: it must provide for offensive forces, for casualties, for garrison forces to hold conquered worlds and with enough left over to maintain both the logistics of the Jihad and the basic rhythms of life in the sietches of Arrakis.

The other major problem the Fremen face is that most of their key advantages evaporate once they are off of Arrakis. Indeed, some invert. The Fremen knowledge of local terrain was crucial to their victory on Arrakis but if anything the Fremen are remarkably badly equipped to understand and fight in other terrains. These are men who cannot conceive of a thing called a ‘sea,’ for instance and one supposes they would not fair well in snow or forest either. Urban terrain is also, crucially, mostly foreign to them. Their mastery of stillsuits, of walking with irregular strides in the desert, of concealment in sand, of the use of sandworms all matter exactly not at all off of Arrakis and in most cases will be active hindrances. At best they will have to face the armies of the Imperium in ‘stand up’ fights, at worst they will be repeatedly ambushed.

What is even worse, the Fremen are stepping into a kind of warfare they are unfamiliar with, for which their society was not designed. Remember: Fremen victory on Arrakis depended on most of the technology of industrial warfare not working there. Sandstorms grounded ornithopters and shields were broadly unusable outside of the towns and villages (and disabled by a sandstorm for the final battle). None of that is true the moment the Fremen step off world.

Worse yet the Fremen supply of industrial ‘firepower’ is fundamentally limited. The Fremen themselves are incapable of manufacturing any of this. One of the sleights of hand here is that while the Fremen disable all of the Harkonnen and Corrino frigates at the opening of their battle at Arrakeen – blasting the noses off – these very ships are handwaved back into functionality for the off-screen Jihad. One wonders how the Fremen – who have never seen this technology before, technology which is built nowhere on Arrakis (we’re told the Harkonnen’s equipment is all off-world import, nothing is manufactured locally) – were able to swiftly repair dozens of high-tech spaceships. Equally, the Fremen lack both the ability to manufacture shields or ornithopters, but also lack the knowledge to maintain shields or ornithopters.

They have exactly what they captured from the Harkonnen and Corrino troops and nothing else, with almost no means to repair anything that breaks – this is where my earlier point that shields evidently require a lot of maintenance and replacement matters. While the idea of running an army entirely off of captured weapons is a thing often thought of, functionally no one ever actually makes it work: open the hood on armies claiming to run primarily off of captured equipment and you almost invariably find foreign sponsors providing the bulk of their weapons. The Fremen have no such foreign sponsors – or at least, won’t have them the moment it becomes clear they intend to burn down most of the known universe – so their access to military material is going to be limited.

As a result, the Fremen are going to be a remarkably two-tier force: a small body of troops equipped with looted shields and supported by what aircraft can be maintained, with a larger body of Fremen fighting ‘light’ as they did on Arrakis, but without storms or worms or mastery of local terrain.

On the one hand, the Fremen would presumably be able to outnumber the first individual Great Houses they targeted. Great House armies are small, as we’ve noted, so while the Fremen would have an overall numerical disadvantage (the Imperium has more trained fighters than there are Fremen) locally they would have the advantage, created by their control of the Spacing Guild. It would be less overwhelming than you might first think though, for a fairly simple reason: though the spacing guild is compliant, the Fremen only have the space transports they can capture. Note that the Spacing Guild supplies heighliners, not frigates and the Fremen do not know how to build frigates. So their ground-to-orbit and orbit-to-ground capacity is going to be limited. High – the Harkonnen and Corrino fleets captured on the ground at Arrakeen were large – but limited. Still probably enough to give the Fremen local numerical superiority everywhere they went.

The problem would be attrition: Fremen manpower is brittle. This is made worse by the fact that achieving numerical superiority on multiple fronts – and we’re told this fight encompasses a great many worlds (and planets are big things – most of them do not have all of their major settlements packed in one small area like Arrakis does), so they fight on multiple fronts – would require deploying large numbers of those ‘second tier’ Fremen forces. Those Fremen are going to be lethal in close combat, but extremely vulnerable to the industrialized firepower of the setting: one thing we’re told very clearly is that lasguns are evidently extremely powerful against unshielded enemies.

Meanwhile, as capable as the Fremen are, we also know they are not trained how to fight in shields (it is an entire plot-point in Paul’s duel with Jamis that they do not understand Paul’s slower movements), so once forced by military conditions outside of Arrakis to fight shield-against-shield, some part of the Fremen qualitative edge will be lost even for the ‘first tier’ troops.

And simply put, a few million Fremen is probably not enough to actually sustain that campaign, though I will admit it could end up being borderline, depending on the size of Great House armies and the loss-ratios the Fremen are able to put up. Once you have siphoned off the tens if not hundreds of thousands of soldiers required to garrison worlds that have been taken and accounted for losses fighting technologically superior opponents in unfamiliar terrain, I would guess that Fremen manpower would end up badly overstretched.

Very roughly, we can start with 15 million total Fremen. While Fremen women are trained to fight and Chani is on the front lines, we do not see any other women do so: the Fremen do not employ their womenfolk offensively as fighters, as a rule.² So accounting for women and children – in a society that we may assume has almost no elderly – that 15 million total Fremen might give us 5 million military aged males available for offensive deployment. Some portion of those will still be needed on Arrakis for spice production, administration and so on, but perhaps it is a small portion.

So perhaps 3 million Fremen available for offensive action off world, of which perhaps only a few hundred thousand can be moved at a time given the limited supply of frigates, charging out into a universe with perhaps something on the order of 15 to 30 million trained fighters. That offensive force will be depleted not only by casualties, but also by the demands of holding and administering captured territory and also that army needs to still exist when the fighting is done, both to deter what Great Houses remain and also to enable the continued existence of the Fremen as a people. If Paul conquers the universe but gets a majority of all military-aged Fremen men (over a decade, so more than one full generational cohort) killed, Atreides rule isn’t going to last very long.

Worse yet (it gets worse) the manpower pool the Great Houses operate from is absolutely vast – there are evidently tens if not hundreds of billions of people in the Faufreluches – so any Great House not entirely wiped out is going to be able to reconstitute fairly rapidly. If you do wipe out a Great House but leave the planet, there are no shortage of richece willing to take their place and then reconstitute a Great House army fairly rapidly. The Fremen are going to be playing whack-a-mole quite a bit, because their opponents have enormous demographic reserves to draw on, while by contrast the Fremen’s own are very limited. Of course the Fremen could start recruiting people out of the faufreluches, but that seems both unlikely (the Fremen do not bother to conceal their contempt for the people of the villages of Arrakis, whose conditions are already much harsher than the average worker in the faufeluches) and would also dull the all-important qualitative edge the Fremen need. So while the perhaps 5 million or so total Fremen military-aged-males is a exhaustible, set resource the 15-30 million Great House fighters is a resource which can be almost endlessly replenished.

It is easy to see the ways this could go wrong. First, the Fremen lack of industrial military power could cause the casualty ratio to turn the wrong way once they are off world. Sure, they have the superior close-combat fighters – we’ve stipulated that – but if you lose half of every attack group to lasguns, hunter-drones or other ranged weapons on the way in (because you haven’t enough shields), the Fremen are simply going to run out of Fremen before they subdue the Great Houses. The other path is one where the campaign sputters: the Fremen win initial (costly) victories due to numbers and mobility advantage but are then forced to dissipate much of their force in garrisons and administration. That in turn enforces something that happens to many great conquering peoples: they become like the regimes they replaced. Fremen leaders with their small military retinues settle down to control and exploit the worlds they garrison while being vassals of the Atreides – in short, they become Great Houses, likely losing whatever distinctiveness kept them militarily superior in the process. In either cause, because the numbers are so lopsided, the loss of momentum for the Fremen probably spells collapse as the balance tips back the other way and the Great Houses, with superior manpower and economic resources, begin whittling down what is left.

In short, Fremen victory against the Great Houses strikes me as possible but implausible, it is an unlikely outcome – one that probably would require a prescient warlord directing everything to perfection in order to win. Which as we’ve noted already, is a failure point for the narrative of the books, which require this war to be a thing that succeeds regardless of if Paul lives or dies.

Of course this assumes broadly that the ‘military resources’ – trained fighters, shields, supplies, frigates and so on – in the ‘system’ remains fairly static: that the Great Houses mostly fight as they have always done, with the weapons they’ve always had. One result of that is that the Fremen never get access to the quantity of weapons to fully modernize their own forces – the Great House armies are, ironically, too small to furnish them enough systems to capture.

Of course those limits might not hold. War is, after all, the land of in extremis. The Fremen assault might be enough to really break the static nature of the faufreluches and unlock a lot more economic potential, which might increase the military resources the Fremen could unlock from captured worlds.

That scenario, it turns out, is both likely and much worse.

Fremen: Total War

First, let us start with the part that this seems likely.

So far we’ve been discussing this as a war between the Fremen and the Great Houses, with the much larger mass of the population left out of it. We’ve done that because I think it is the only version of this war the Fremen could win. But it is also clearly, explicitly not the version of the war that happens.

Again, we’re told in Dune Messiah that the Jihad ends up killing 61 billion people, wipes out forty religions, and sterilized ninety planets.

In short, under Muad’Dib’s leadership the Fremen are not merely waging a war against the noble families of the Great Houses, but rather a war against the people of the Imperium. There is something of an irony that Frank Herbert seems to be clearly thinking in terms of something like the rapid expansion of the Rashidun Caliphate (632-661) here, but the Rashidun caliphs quite deliberately avoided this sort of thing, often offering religious protections to the underlying peoples beneath the empires (Roman and Sassanid) they were attacking to avoid a situation where they faced broad popular resistance. That said, this aspect of Islamic conquest was often not emphasized in the 1960s popular understanding, so Frank Herbert may not have been aware of the degree to which local religions and communities were largely and intentionally left in place during early Islamic expansion.

Either way, it seems almost certain that Paul’s Fremen attempting to extirpate entire religious traditions and sterilize entire worlds, are going to start facing broad popular resistance.

Now obviously the first problem here is that it makes their manpower problem much worse. When the Fremen were just facing the Great Houses, they were outnumbered perhaps 5-to-1, which is quite bad but in the fiction of the setting superior skills can overcome those disadvantages at least some of the time.³ But against a, say, Earth-like planet – of which there must be very many, given that killing 61 billion people did not even cause much of a social collapse in the Imperium – the Fremen might face mass-mobilized armies on just that planet in the high tens of millions. The USSR mobilized an astounding 34.5 million troops during WWII out of a population (pre-war) of about 200 million. Naturally it would be hard to mobilize a whole planet on that basis, but doing so on a modern-Earth-like world would net you around one billion soldiers.

So the idea that the Fremen might find themselves landing forces of, say, 300,000 Fremen warriors (representing basically the maximum carrying capacity of the Corrino and partial-Harkonnen fleets they captured) on a planet only to find themselves facing an opposing force five million or fifty million or five hundred million foes is not out of the question. One of the few ways to force that kind of mobilization from modern societies is to attempt to genocide a population or extirpate their long and sincerely held religion and the Fremen are trying to do both.

Now the Great Houses can control these populations because they maintain local legitimacy, because shield-based fighting gives them a huge advantage against populations that cannot afford shields and because they have demilitarized the lower classes. But the Fremen will have removed all of these factors. The Fremen do not have long-standing local legitimacy – they are a barbarian foreign force trying to take away your religion. They also do not have a shield-based fighting system and lack enough shields to fully equip their force in any case and so take to the field without a technological edge over a mass-mobilizated populace. And worse yet, the very threat they pose is going to push the lower classes to militarize.

Now in pre-industrial societies, this effect was somewhat limited because pre-industrial societies were not capable of fully militarizing their lower classes. But the societies of Dune are post-industrial societies. It may be impossible to provide the high tech instruments of warfare to an entire mass army – not enough shields, ‘thopters and frigates – but it would be trivially easy for these societies to equip the great masses of their population with spears, swords and simple guns.

Ironically, the Fremen would now find themselves immediately caught in the same trap as the Great Houses: trained in a fighting style that emphasizes close combat, they would try to have close-combat mass-battles with huge, unshielded armies of melee combatants, rather than being set up to use their shields to maximum advantage by conducting the fighting at long range.

Facing even relatively modest mass armies would require the Fremen to deploy a lot of their available manpower simply to be able to hold ground on the kind of scale these wars would be fought on, which would make the two-tier structure of their army even more of a liability because it would force them to field those second-tier troops in quantity. And while a Great House might be dumb enough to fight those second-tier unshielded troops in close combat – that being their habit – one imagines a mass army of resistance might approach it differently. After all a mass army is going to look for cheap ways to arm hundreds of thousands or millions of fighters and guns and artillery are relatively cheap compared to shields and ‘thopters. And we know that the basic technology of artillery is not lost, because Vladimir Harkonnen uses it as a surprise tactic against the Atreides.

Heaven help the Fremen if some planet somewhere stumbles on the same idea and expands it out to a fifty-million-soldier army against a largely unshielded, close-combat-based infantry Fremen force. Ask the survivors of the Battle of Omdurman (1898) what happens when the most skilled, motivated, desert-hardened and determined ‘hard men’ attempt to charge machine guns with contact weapons. While the ‘first tier’ Fremen troops with captured shields might still be effective, after their ‘second tier’ supporting units were obliterated they would be horribly outnumbered, easy enough to simply mob down with bayonets.

Even if the Fremen qualitative edge remained intact – perhaps because their opponents continued to operate in the contact-warfare frame rather than rediscovering projectile weapons – the attritional structure of the conflict would become unsustainable pretty quickly. Paul could easily lose half of his entire offensive force fighting a single partially mobilized world of this sort with a 15:1 casualty ratio in his favor.

But there’s an even worse outcome here for the Fremen, especially given the length of the conflict: total economic mobilization. So far we’ve considered worlds with perhaps days or weeks of warning doing panic mobilization while under attack, churning out as many rifles and swords as they can to put together mass armies, relying on the fact that planets are very big and so any conquest would take months if not years.

Paul’s Jihad lasts twelve years, canonically. For a sense of what twelve years is in ‘mobilization time,’ the United States went from producing almost no tanks in 1939, to just 400 in 1940, to 4,052 in 1941 to 24,997 in 1942, to 29,497 in 1943. In 1939, the United States built 5,856 aircraft; by 1944, it was building more than 8,000 aircraft a month.[efn_notes]Statistics via Overy, Why the Allies Won (1995), 331-2.[/efn_note] Again, as we’ve already noted, the only way the small armies of the Imperium make sense with its attested population (which must be more than the 61 billion Paul kills) is if this society is mostly demilitarized. We see plenty of industrial capability – aircraft, space-ships and so on – it is just that these noble houses with their limited administration cannot mobilize that capacity for war.⁴ The technology and population exists, what is lacking is the administrative capacity and political will to employ it. And while we might imagine that Dune‘s frigates and ornithopters are more complex machines than WWII-era aircraft, tanks and warships, it is equally the case that we’re thinking about the economies of entire planets rather than individual countries.

But for a planet that found itself not immediately under attack but very obviously in the path of Paul’s Jihad – perhaps with a well-entrenched local religion – that calculus is different. Information might spread slowly in the Imperium, but not infinitely slow – at least the elite do seem to have some sense of affairs in distant places. Those richece, perhaps with their nobles or without them, might well opt to do what those noble houses with their tiny, underdeveloped administrations could not: mass mobilize not just people but industry, unlocking the productive capacity of several billion people and turning much of the civilian economy over to a war-footing in a way that the Great Houses, with their small administrations and very limited legitimacy never could. Show the people film-strips of Paul Muad’Dib’s army murdering billions and sterilizing worlds and say, “that is coming here unless you line up to work in the factory churning out ninety thousand ornithopters a year.” Big posters that say, “to keep the Fremen Fedaykin murderers away from Our Holy Sites, we need YOU to hit our target of launching two thousand heavy weapons frigates this year!” Industrial societies engaged in something approach total economic mobilization can produce enormous amounts of destruction very rapidly.

The Fremen Jihad lasts more than long enough for the more populous worlds of the Imperium to adopt this kind of war economy in preparation and the tremendous violence that the Fremen inflict – again, sixty-one billion casualties – are more than enough to motivate a lot of these worlds to do exactly that.

Paul will, in that event, at least be lucky that the Spacing Guild might let him isolate such worlds, although if you are the Spacing Guild (or an anti-Fremen group of smugglers) you might just be willing to roll the dice to see how Paul’s base of power on Arrakis handles the arrival of thousands of frigates with tens of thousands of ‘thopters carrying millions of heavily equipped troops showing up in the skies above Arrakeen.

The Failed Jihad

Now of course the natural response to all of this is that Paul Muad’Dib Atreides can avoid all of these outcomes because he is the Kwisatz Haderach, able to see the present and the future and thus able to anticipate and avoid all of these outcomes, threading the needle of probability perfectly to guide the Jihad to its victorious conclusion.⁵ And of course we’ve already noted the flaw in this: Dune is explicit that by the time Paul fully grasps his prescience, it is too late to stop the Jihad, which would happen and succeed even if he was dead. Paul is merely the catalyst for what Herbert imagines as historical – nearly ecological – merely the manifestation of the ‘world-spirit’ of the age moving through history. The Jihad would happen without him. Only the catalyst is required; the rest is inevitable.

And it just clearly isn’t. There are, in fact, quite a lot of ways the Jihad could swiftly fail.

And fundamentally that goes to how Frank Herbert’s vision of military power – one shared by quite a lot of people – differs from how military power is actually generated. In Frank Herbert’s vision, military power is a product of the individual capabilities of fighters, which in turn is produced ecologically based on the harshness of the environment they come from. He imagines huge gulfs in capability, where two legions of Sardaukar can easily overpower ten legions of Harkonnen and Fremen in the desert can inflict even more lopsided casualties on Sardaukar.⁶ There is a direct correlation then between the harshness of a place and the military power it can produce.

And equally, there is a strongly gendered component of this view in Frank Herbert’s writing: militarily effective societies in Dune are masculine in key ways.⁷ Harsh conditions, for Herbert, produce intensely masculine societies (whereas the decadence of the Imperium is signaled in equally gendered terms: the gay sexual deviant Baron, the genetic eunuch Fenring, the emperor with his household of daughters and his failure of “father-head”-ship), which in turn produce militarily effective ones.

It is not hard to see how intense and pervasive a view of military power that is, how frequently in popular culture ‘manliness’ is presented as the primary source from which military effective flows. This isn’t the place to get into the modern manifestations of this sort of ideological framework, but it is not particularly hard to find recruiting and propaganda videos that attempt to communicate military effectiveness almost purely through gendered visual language of masculine fitness prowess, as if victory belongs to the army that can do the most push-ups. Herbert’s vision is somewhat more sophisticated than this, but only somewhat. It is water drawn from the same well.

And that simply isn’t how military power is actually generated in the real world. Training certainly matters and there are some kinds of fighting – like horseback archery – that almost have to be deeply socially rooted to be effectively trained. Cohesion also certainly matters, but it can be generated quite a few ways and strong cohesion is certainly possible to produce ‘synthetically’ through training and drill. But the strongest armies do not generally come from the harshest places – indeed, the opposite: for most of human history the military advantage has gone to resource-rich places with dense populations. This is obscured somewhat in popular culture because the exceptions to this rule are so striking but they’re striking because they are exceptions.

But especially after the industrial revolution – and Dune is a post-industrial (very post-industrial) universe – military power is largely generated by economies, a brute-force product of the ability of societies to deploy the most men (supported by their agriculture), the most metal, the most explosives and these days the most electronics. Weaker powers can still win by protracting conflicts and focusing on degrading the will of an enemy, but they do this because they are weaker powers who understand that they do not have much of any chance of winning in a direct confrontation. Indeed, the armies that have put the most emphasis on the ‘fighting spirit’ or individual physical superiority of their soldiers have tended to lose modern wars to armies of conscripted farm-boys and shop-keepers backed up by tremendous amounts of modern industrial firepower.

Of course, as Clausewitz reminds us (drink!) war is the realm of the “play of probability and chance” – a contest in which the stronger does not always win. Military strength may be, in modern times, almost entirely the product of industries, economies and demographics (and the first two more than the last one in most cases), but such raw strength is not the only thing that determines the outcome of wars, which equally depend on the strategic importance of the objective, the political will of the two parties and the vagaries of chance that are omnipresent in war (drink again if you got the reference).

None of this makes Dune a bad book or Frank Herbert a bad author – it is a fascinating book that raises these kinds of ideas and questions. But equally precisely because the book’s understanding of where military power comes from derives not from historical facts but from fictional events, it is worth thinking hard about how it imagines this works and if that actually corresponds to historical trends.

In this case, Frank Herbert imagines that ‘historical forces’ have created an effective inevitability that once roused the Fremen, on account of their harsher society, would storm the universe basically regardless of the balance of logistics, military equipment or numbers because the vague ‘hardness’ of their society makes them unbeatable. It makes for a fascinating narrative, but this is not how history works and indeed the wastelands of history are littered with the half-remembered names of a great many peoples who were ‘hard’ and ‘tough’ and ‘aggressive’ and utterly slaughtered or overrun because the ‘wealthy’ ‘decadent’ and ‘unmanly’ societies they fought also had greater numbers and superior weapons.

So to answer the original question: no, one way or another, the Fremen would fail, though they might fail in the most interesting way – failing not by replacing the faufreluches, but by galvanizing them into producing (or reproducing) a different kind of self-governing society that was far better able to mobilize itself and its resources – and capable of far more destructive, horrifying forms of war.

One wonders what the Dune universe’s version – after the collapse of both the faufreluches and the Fremen – of the First World War, a horror-show of industrial warfare on unprecedented scale – would look like.

Welcome to WordPress. This is your first post. Edit or delete it, then start writing!

I just completed a major update to my website. Two things at once. I’ve moved my site from blog.andrewshell.org to andrewshell.org (dropping the subdomain) and migrated it from 11ty to WordPress.

I put a lot of effort into setting up redirects, so everything should keep working (I hate broken links). If you find I missed something, please contact me.

I have a few reasons for doing this:

First, I’m very excited about adding ActivityPub support (you can follow @andrew@andrewshell.org on Mastodon), so anything I post will hopefully show up in AP easily.

Second, I manage several WordPress sites for friends and realized I was getting rusty on my day-to-day WordPress skills.

The theme I created is still a “classic theme,” which isn’t what I initially planned, but good enough for now. I do a lot of microformats/indieweb stuff, and I’m not sure how much control I have over this with the block editor.

So for now, everything looks like it’s in a good place while I figure out what I want to keep or change. I’m happy to be back in the WordPress ecosystem.

I don’t need to but it is convenient.

My pens can be put into three categories.

• The Kaweco Sports: very useful to put in pen holders on notebooks.
• Pens with broader nibs. Great for drawing likes and scribbling out stuff.
• Pens with finer nibs. My current preference for general writing

Having this many pens in rotation makes it easy to just rotating out pens that are dry and in new ones from my desk and being able to clean and re-ink pens as I have time.

One of the reasons I migrated my site to WordPress was all of Dave Winer's evangelism. One of the things Dave has done is an external editor for WordPress called WordLand. I had tested it out briefly with a test site on WordPress.com, but now that my main site is on WordPress, I'll see how it works. This post, was written with WordLand. I wasn't sure if it worked with self-hosted apps, but I guess because my site is linked up with WordPress.com via Jetpack, it just works.

This evening I added an idea I have had for a while to Artemis: a "dense" layout.

By default, the Artemis interface appears in a single-column layout. The dense layout creates several columns. Each column from left-to-right shows posts for a given day. You can scroll down to the next row of columns to see posts from previous days, too.

You can experiment with a new "dense" layout, designed to be skimmable, by setting your "Choose a Layout" preference to "Dense" in your account settings.

Here is what the dense layout looks like:

ALT

The Artemis dense layout showing three columns of posts for February 22nd, February 21st, and February 20th.

I made this layout with a question on my mind: how much information could I fit on the page while preserving the readability of its contents? A grid layout lets me pack in many more posts. The UI is busier than I prefer, but I like how skimmable a dense grid can be.

On reflection, this layout also lets me see clusters of posts that I haven't read (marked in white). (Although I have read more posts than the ones above via someone's site directly but that doesn't show up above.)

Addendum

This layout fits in my idea of what a "Lab" set of features would look like for Artemis (a la Ghost Labs). Such a Lab would contain features that are experimental. I don't have list of features in a Lab that you can toggle on/off, although I would love to create a Lab section at some point so people can play with experimental features that have an explicit label indicating they are experimental.

Addendum #2 (Added Feb. 23rd 2026)

I originally experimented with the idea of a "dense" layout in May last year and came up with a design that packed even more information onto the screen:

This design was the original inspiration behind what is now the "dense layout" feature, although the current "dense layout" has a little bit more whitespace.

If you use Artemis and would prefer to have this even-more-dense layout, please let me know.

The IndieWeb community chat has a feature that lets you create a wiki page from a chat interaction. You can say “what is {term}?” in the chat and, if there is a definition on the community wiki, the definition will be returned by a bot. Otherwise, the bot, Loqi, will say that you can send a message in the form “{term} is…” to create a stub wiki page with your definition.

This allows community discussions to start in one medium – a community chat, which in this case is bridged between Slack, IRC, and Discord – and end in another: the wiki. In this way, a definition “graduates” from a message in a chat which may not be easy to refer back to into a definition with a permalink to which the community can easily refer. Importantly, when information is on the wiki, others can contribute and develop the page further.

I see this pattern of what I am calling “graduation” in other places too. The group chat feature in Instagram lets you start a call with friends in the chat. You can start with text messages as normal and, at any time, move into an audio call. You can do this without having to switch contexts (i.e. go to a different application and figuring out how to invite everyone to the chat).

You can “graduate” from a text chat to a videoconference in professional contexts. I have see this be effective in many contexts, especially when a text chat has been going on for a long time and talking through a problem via a call would be useful.

Now I think about it more, the W3C has an IRC bot, Zakim, that is used for meetings. At the end of a meeting, chairs can turn the text discussion from the meeting into a permalink that serves as minutes (example).

You can “graduate” between contexts for both creative (i.e. to create an artefact that lasts) and communication purposes (to move to a medium that has better affordances for the type of discussion you want to have, like a video call with screen sharing to work through a professional task).

“Graduate” may not be the perfect term in that it may imply that something should begin in one form (i.e. text) and end in another (i.e. a video call). I don’t think that’s true; a call could graduate into a document, for example. Another term may be needed. Indeed, mediums work together: text chat has its own affordances, as does video chat, wikis, websites, and so on. “Graduation” feels like it applies to the moment when it feels like another medium would help advance a discussion at a point in time.

It has been helpful for me to think about how a place expressed through technology – for example, a chat, or a software application – may be a place to start, nurture, and develop a discussion or creative work, and also a springboard to other mediums that have other affordances (permanence of wikis for archiving, the ability to share screens and talk through a problem on a videoconference, etc.).

I wonder what other instances there are of this concept.

Artemis lets you subscribe to ActivityPub feeds (i.e. accounts on Mastodon). To do this, you can type in an ActivityPub handle like @jamesg.blog@jamesg.blog [1] on the “Add a website” page.

When you subscribe to a feed using an ActivityPub handle, Artemis can use the information in the feed to determine if a post is in reply to a previous post by the same author. This is sometimes called a “thread”; longer threads of posts by the same author used to be called “Tweetstorms”.

Rather than present all posts by an author individually, Artemis groups threaded posts. Here is what this feature looks like:

ALT

An example of a thread in Artemis where the first post in the thread is at the top in a white colour, then replies by the same author are nested below with an indent. Replies are in a muted colour to distinguish them from the first post.

The post at the top of the thread is at the top. Subsequent posts by the author in the same thread are nested below the top post, with an arrow icon to indicate the visually nested post is a reply. Only the first two posts in a thread are shown this way so that long threads don’t take up a lot of space int he reader (although I will likely add a [and {n} more] label soon).

In addition, ActivityPub posts are presented slightly differently in Artemis. Posts with images will have [photo] to indicate there is an image in the post; posts with content warnings will have a [content warning] label; hyperlinks are replaced with [link]. Using these textual cues, you can get a sense for the post with a quick skim. Artemis acts as a preview: to see the full post, you can click through to the author’s original post.

It is also worth noting Artemis only saves posts published by an author and replies to their own posts, rather than all posts by an author (which would include their replies to others' posts, re-posts, etc.). I made this decision because I generally like to follow what someone has written as my first priority. I may extend this feature in the future to allow people to choose what they want in their reader if this is requested by users.

Conclusion

The concept of nesting posts is related to the “Grouping link posts in a web reader” post I wrote last year, except that post was specifically addressing a scenario where one website you follow responded to another. The thread grouping feature discussed in this post is specifically for threads of content by the same author published via ActivityPub.

I wanted to document this feature as an example of how to present previews of threads. If use Artemis and have any ideas on how this could be better, or have seen other implementations of this idea that you like, please feel free to email me!

Footnotes

[1] @jamesg.blog@jamesg.blog publishes a post with the title, first few words, and URL of new posts published on my website. I use Bridgy Fed for this. I set this up in case people wanted to get notifications for when I publish something new on my website. I haven't announced this anywhere yet, so consider this footnote the announcement.

Recently, a feed a few users – including myself – were following with Artemis published a “bookmark”-like post. The markup in the corresponding feed was a bit different than expected, so Artemis ended up linking directly to the bookmarked post rather than to the author’s post itself.

I fixed the underlying bug in Artemis, but it left me thinking about an idea I have been considering for a while: should Artemis let users know when a post has a permalink that corresponds with a site other than the author’s?

I recently subscribed to a feed of bookmarks that someone published and had an idea: I could add “via” in the author’s name. So when they publish a post in this feed, dedicated to bookmarks, Artemis would show it as being published “via {their name}”. I liked this pattern but it meant that I had to proactively update bookmark feeds I follow to include “via” in the author name. And I would have to remember to do this when I followed other such feeds in the future.

This led me to develop a new feature: Artemis “via”. This feature shows “via” when Artemis thinks the permalink for a post is not the same as the site you are following. Here is what it looks like if a post in your reader links to a site other than the author’s:

Above there are three posts. The first post is published on the author’s site, so their author name appears as normal. The next two posts link to sites other than the author’s website that I am following. “(via)” appears before the author name when the entry links to another website. “(via)” is in italics which provides a subtle visual indication that Artemis has added this. Users can’t add italics to author names.

With the above "(via)" labels, I can see without clicking or hovering over a link that it will take me to another website.

The following things must be true for “via” to show up next to an author’s name for an entry:

• When your feed is calculated, the hostnames of all entries are calculated. The most popular hostname is found. If the number of posts with the most common entry name is not equal to the number of posts published in the author’s feed (as defined by how many days of posts you opt to show in your reader; ideally Artemis would look further back, but I still need to scope out this work) a given post in the author’s feed will be considered for a “via” label.
• The entry URL hostname must be different than the feed’s domain for the entry to be considered for a “via” label.

If both of these conditions are met, “(via)” shows up before the author’s name for an entry.

For example, if I have a feed jamesg.blog/bookmarks and the post permalinks in the feed point to example.com and coffee.com, my author name would appear as “(via) James’ Bookmarks” in Artemis.

This implementation accounts for an important edge case: some feeds may publish the same hostname for every entry but the feed URL itself has a different hostname. I saw this come up in the case of feeds for a blogging platform that supports custom domains; sometimes, the underlying feeds linked to the blogging platform itself rather than the custom domain.

Suppose I publish jamesg.blog and all posts use jamesg.example.com as the hostname. This would mean the first condition in the above bullet point list (a hostname for an entry must match the most popular hostname) would fail, so “via” would not show up. This means that if a site has a different hostname for all entries, “via” will not show up.

Without this condition, I saw sites that I knew to be linking to the right URL showing up as “via”. This is suboptimal: seeing “via” in cases where it is not applicable would be confusing.

The “via” logic is still very much under development: it may be the case that I need to come up with other heuristics. With that said, I think it is a good start. The overarching motivation is to make it more clear when an entry permalink points to an external site, as is common in the case of bookmark feeds. By adding “via”, Artemis can help users understand that what they are seeing is not going to be on the site they followed, but rather go to another site that the author they followed found interesting enough to put in their feed.

When I announced Artemis in 2024, I titled the announcement “Artemis, a calm web reader, is available (in beta)”. So central to the philosophy of how I build the software is the principle “calm” that, when I write about Artemis, I still use the phrase “a calm web reader” to describe what the software is.

While exchanging blog post titles with Britt, she asked me a terrific question:

How you maintain calmness in Artemis while still adding new features. Is there an upper limit on the amount of things a calm reader could do while still being calm?

I ended up writing a post on another suggested topic “A piece of art I would love to see in person”, but the question about Artemis stuck in my mind. How do I preserve calm in Artemis over time?

I have been using Artemis for over a year and a half now. In that time the software has gone from a static site that updates once a day to a service that others can use. I have added a lot of features since Artemis was a static site. With that said, I don’t think new features have reduced the sense of calm I feel when I use Artemis. I think “calm” is an attribute of a feature and its design: a feature can afford calm, or make the experience of using Artemis less calm.

Adding features while preserving calm

My own experience using the software is that it feels like Artemis exists in the background. I go to see what my friends have written lately and then I go to their website. It doesn’t feel like it demands my attention; the update cadence is much slower than social media and other readers.

The new features I have added, and continue to add, are mostly around the topic of user preferences and control. How can I let a user customise their reader more? How can I make sure the user controls what they see? That latter question, for example, relates to the work I put in on keyword filters (which can now be applied both to all feeds and to individual feeds!). I have a few keyword filters set up which allow me to not have to think about topics I don’t want to see in my reader. In this way, the feature affords calm.

Preserving the calm I associate with the software is made up of intentional choices I make whenever I choose to add, or not add, a new feature. I ask myself questions like: How can I make this as unobtrusive as possible? Does this feature give a user more control over what they see in their reader? Will this feature affect the main reading experience and, if so, how will I make sure it is not disruptive? (For the last question, the answer is often to make something opt in by default.)

In addition, if someone thought a feature was intrusive or confusing, I would consider the feedback thoughtfully before writing a response and see what I can do better. Indeed, I have the experience of Artemis as the person who builds it; others’ feedback would help me improve the software beyond what I can see from my perspective.

Roadmaps, ideas, and red lines

There have been periods of weeks where I haven’t worked on Artemis. This is important to me. I don’t have a roadmap for Artemis with dates on when I plan to deliver features. Rather, I build what feels right when the idea comes. “via” came after a user highlighted a bug that I also noticed. “roll-ups” came after I realised people may want to follow feeds that publish many posts at once, but may not want to see all their post sin their main reader.

There may come a time when I plan a short roadmap for Artemis, but at present all I can think of would be simplifying the code such that things are more stable. I would like to work on removing repetitive code in Artemis at sone point. It would take a lot of time and I would need to plan it out, but this work would make the software easier for me to understand while reading the code and therefore easier to maintain.

There are a few “red lines” for Artemis that I don’t want to cross. These are all essential philosophies of why the software exists. First, Artemis does and will not strive to refresh posts in real time. Second, I don’t want Artemis to become an inbox, so there are no unread counts. Third, I want Artemis to be a stepping stone, rather than a destination. Artemis should point you to sites you enjoy rather than trying to be the place where you go to read the web.

Regarding the question “Is there an upper limit on the amount of things a calm reader could do while still being calm?”, I think the answer is “yes”; the more features something has, the more complex it is. I am cognisant that having lots of user preferences will, over time, feel more overwhelming. To counter this I try to set as reasonable defaults as possible (and, related to the topic of software maintenance, I need to revisit the defaults to make sure they are as good as they can be; I need to do a full run-through of using Artemis from sign-up to adding feeds to keep building my understanding of how the software works for new users).

Conclusion

The original question that motivated this post is prescient. I want to make software that is useful and makes people feel empowered. I think “calm” as a design philosophy is one part of this.

To keep following through on the reason Artemis exists – to provide a calmer way to follow websites you enjoy – means I need to continually consider that reason as I build the software. The obligation to be considerate in adding features is especially great because other people use the software. As a software author making something for other people, I want to make sure people have the experience they signed up for, and for that experience to persist over time.

Search is one of my favourite disciplines in computing. In 2024 I spent a lot of time working on a NoSQL engine that I called JameSQL. This tool now powers the search engine on my website.

Designing search engine ranking systems is tricky to say the least. When I use my blog search engine, I sometimes notice that the article for which I am looking does not show up at the top of the search results. Google set a high standard for search; when I type something in Google in a site search, I can often find what I am looking for.

I am not yet ready to delve back into the world of search, but I did want to take a note of an idea I had today: I want my next search project to have tooling for ranking introspection. By this I mean I want to have tools that let me know why a particular article ranks above another.

At present, JameSQL only returns a single attribute, _score, which is computed using either TF-IDF or BM25, with any additional boosts you have specified (i.e. give h1s 3x more weight). I imagine having a value like _score_answer that would tell me how much weight each attribute used in ranking had, for example:

[
  {"bm25_on_post": 100.01},
  {"score_after_h1_boost": 101.01},
  {"score_after_inlinks_added": 109.01},
  ...
]

This would be an ordered list that specifies what calculation has been made, followed by the score at that point in time. This could then be used to calculate how many points each ranking factor added onto the final score. This can be done by calculating the difference between scores after each weight is applied.

This information would help me answer the question “why is this post ranking in this place for this query?” much more effectively than right now, by letting me see exactly how each calculation and ranking factor affects the final search engine ranking.

I started building a tool that lets me interactively experiment with different algorithms (see image of the dashboard) which was useful. I think I would like to revisit that dashboard to make it more useful if/when I work on a search project in the future.

Outside of the scope of this particular, developer-focused context, I generally want to use software that gives me a clear idea as to why I am seeing what I am seeing. As a user, I shouldn’t be left thinking “why did this show up?” With many opaque recommendation systems used on the web today, I am often left feeling exactly like that: “why did this show up?” This makes it a lot harder for me to understand, and therefore trust, a system.

Last year I was introduced to the idea of “Dopplr colours” in the IndieWeb community. This refers to an accent colour assigned to cities on the now-defunct travel website Dopplr. You can see examples by clicking through different Dopplr city pages in the Internet Archive and paying attention to the borders of the map.

While I haven’t been able to find an authoritative description of the algorithm, to the extent I understand the Dopplr colours were assigned using an MD5-based algorithm. Aaron implemented a demo of the Dopplr colour system and described the algorithm in PHP as:

substr(md5($_REQUEST['city']), 0, 6)

Here is an equivalent Python implementation:

import hashlib
​
colour = hashlib.md5("jamesg.blog".encode()).hexdigest()[:6]
​
print("#" + colour)

These code snippets calculate the MD5 hash for a string, then take the first six characters. This creates a hexadecimal value that can then be used as a colour. The Dopplr colour for my domain name is #e228f3. It’s pink! Of note, you can calculate a Dopplr colour for any string, not just city names.

The IndieWeb community uses Dopplr to assign colours to cities in its timeline of in-person IndieWebCamp events:

ALT

A table showing a list of cities with cells coloured using the city's Dopplr colour if an IndieWebCamp event was held in the city in a given year.

I had never heard of the idea of Dopplr colours prior to the IndieWeb, and a Google Search was not fruitful in returning a page that described the algorithm. I thought I’d write this page to document the idea, and make it easier for people to find the idea.

When I was in high school, I used to keep a pencil or pen up my sleeve. I don’t remember anyone else doing it, or why I started. But if I had a pen up my sleeve, I would at least know I had one nearby for when I’d need one. (I may be mis-remembering whether having a pen up my sleeve actually helped me remember one (oh the irony!). I remember a lot of times that I would forget a pen or a pencil. My organisational skills are better now, although still more on the chaotic side than I would like.)

Now, I keep a pen in a pencil case that is in my backpack: there just in case I need it.

I started thinking about this yesterday because, during a session where I was writing a lot down on my whiteboard, I realised that I had put the whiteboard pen up my sleeve. I haven’t done this in years. I felt a little bit of joy thinking about keeping a pen up my sleeve. Shortly thereafter, I took the cap off the pen and started writing. So as not to forget the moment, next to the words “search engine introspection” (an idea whose essence later became a blog post), I wrote down “pen up my sleeve” – a little bookmark that, every time I look at it, brings back nice memories.

I love that the whimsical and the serious can co-exist on whiteboards. Whenever I see a whiteboard, I like to draw a smiley face on it. Occasionally, I will draw a smiling cat face too. I will do this wherever I am and have access to a whiteboard. Why? This might make someone smile.

Museums and galleries are places we can go to learn about the past, think about the present, and consider the future. Museums are places we can connect: with times, people, and place. We can have conversations in museums. We learn in museums. We can explore.

The topic I have chosen for this month’s IndieWeb Carnival is “Museum memories”.

This month, I invite you to write a blog post about a memory that you have of a museum. It can be any museum: your local art gallery, a museum you visit often, a museum you visited on holiday, a museum dedicated to one of your interests (the sea, video games, transportation, your favourite football team), or a museum at a historic site you have visited.

To get you thinking about this month’s topic, I have a few questions that you can use as starting points:

• What is your favourite museum? When did you realise it became your favourite?
• Do you recall a time when an object in a museum or an art gallery made you feel something? What was the feeling?
• Do you remember when you first stumbled upon a new kind of history or art that you hadn’t thought or known about before?
• Have you met anyone interesting at a museum? A new friend? A tour guide that got you thinking about something in a new way?
• Is there a museum you loved as a child? What made it so interesting to you?
• How does/did the physical space of a museum you visit or have visited affect you? Do you feel a certain way when you enter?

This month’s IndieWeb Carnival runs from March 1st to March 31st. To participate, write a blog post on the topic of “Museum memories” on our website. Then, send me an email at readers@jamesg.blog with a link to your submission.

I will curate all of the submissions I receive on this page as I receive them. At the beginning of April, I will write a round-up that summarises all of the submissions I have received.

Entries are listed below in reverse chronological order of submission. This means that the most recently received posts are at the top of the list.

Entries

• On Museum Memories: Daniel's Story by Bix
• Museum memories by Beto
• Two Museum Memories by Joe
• Two Museum Memories by Jeff
• Museum memories by James (me!)
• Museum memories by Mike
• Museum memories by Zinzy
• The Museum of Possibilities by Jatan
• 美術館による記憶 / Museum memories by Last Encore
• Museum Reflections March 2026 - The WAG-Qaumajuq by Jinny
• The power of museum memories by Ben
• Collections and catalogues: The lasting power of museums by V.H. Belvadi
• Lost in the Met by Nick
• Museums That Aren't Quite Museums by Sammie
• Growing up around a museum by Matthew
• Exit through the gift shop > The fine line between culture and ostentation by Andrei
• Museum Memories by David
• Museums by Noahie
• Museum memories by Elena
• 26W11. The little book in the velvet-lined case by Anthony
• IndieWeb Carnival: Museum Memories by Kristof
• The Antique Telecommunications Museum by Reilly
• Museum memories by Paul

Berthe Morisot’s Girl on a Divan, displayed in the National Gallery, London, caught my eye as soon as I saw it. The painting was the last one I saw in the Impressionist room in the Gallery, but despite having been on my feet for hours before I stood for several minutes looking at the painting from all angles. There was something wonderful about the painting.

Girl on a Divan portrays a woman gazing directly at the viewer. The woman’s expression is relaxed and friendly; the woman has a slight smile on her face. The woman is wearing a white dress with several stripes of colour. The stripes are somewhat incongruent with the white, not having any particular pattern. With that said, the stripes add to the depth of the image, and the colours of the stripes – mainly blue and orange – contrast well with the blue background.

In the middle ground, a shade of blue/green is used to create what could be a chair. This further adds depth to the painting. The woman has one arm raised slightly. Could the arm be resting on the arm of the chair?

I love the use of blue in this painting; the colours are both soft (through the shades chosen) and eye-catching (through the extensive use of blue for the background) at the same time.

The painting is made with sketch-like brushstrokes for the dress, background, and the hair; the face is more relatively detailed.

After several minutes of looking at the painting, I noticed a white dot in the left eye: a glimmer of light. This brought me a lot of joy. First, it took a while to notice this detail! I love “Slow Looking”. Second, after noticing the glimmer in the eye, I thought about how much it adds to the expression of the sitter.

Girl on a Divan is now one of my favourite paintings.

I didn’t include a photo of the painting in this post for licensing reasons. You can see a detailed photo of the painting on the National Gallery website.

When I was building the search engine for my blog, one feature I wanted to implement was syntax highlighting within the search input field. I wanted special operators (i.e. has:noalt, which shows posts that contain one or more images without alt text) to have a different background colour indicating that a given segment of text had some semantic meaning.

Here is an example of the GitHub search feature which has semantic syntax highlighting:

In the above image, the "capjamesg/indieweb-etherpad-archiver-v2" text is set in a light blue font; the text background is in a darker blue. This text is highlighted because it follows "repo:". This part of the query means that I want to scope my search to focus on a specific repository, "capjamesg/indieweb-etherpad-archiver-v2".

By highlighting the text that is part of the search query, I have a visual indication that what I have typed has some special meaning to the software.

Regex101 has a similar semantic syntax highlighting feature in the regular expression form field on their website:

Various characters in the above screenshot appear with coloured backgrounds; each one has semantic meaning.

The status quo

There doesn’t appear to be a semantic way to implement text highlighting in a native input tag. GitHub uses an input tag as the text and a lot of other code; a cursory read doesn’t make it obvious exactly what is going on. Regex101 uses a contenteditable div and gives every character in it a span tag. Both of these involve more code than I would like, all of which I would have to maintain.

I also explored an implementation of an input field that involved clipped background gradients, but there turned out to be many limits: maintaining state while scrolling, glitches in positioning the background, the requirement to use a monospaced font for the input so I can use ch to estimate where characters are in the input field.

What I would like

In my dream browser implementation, I would like to be able to apply the following styles:

• Text colour
• Background colour

I tried to use the CSS Custom Highlight API on an input field but it didn’t work. It feels like this API could be a stepping stone to highlighting text in an input field?

For my use case, which is to style a text input field, I would like to be able to specify the character indices at which a highlight should start and end. I can then use JavaScript with my own grammar parser to determine what character indices should be highlighted, and in what colour (I may want to use one colour for filter keywords, for example, and another for sort keywords).

Security

I suspect there are security considerations for this idea: what if the input field has white text that hides from a user what is in the form field? A malicious site could make it seem like there is no text in a form field visibly even if there is such text. With that said, I wanted to write down my thoughts to start a discussion.

Related discussions

• CSS Working Group Issue #4603

I am using Duolingo to learn a bit of German. I have been using the app for a while and enjoy the exercises. With that said, I am not a fan of Duolingo’s changing application icons on iOS.

I don’t like that an application can set its own home screen icon without my permission. I further do not appreciate that there doesn’t seem to be a way to change the app icon in some states. For example, there is an icon where the application mascot, Duo, appears with a plaster. I find the icon relatively distressing.

This weekend I figured out a workaround to reset the icon (at least on my home screen). Now, I have an Apple Shortcut that does one thing: open Duolingo. Because you can set custom icons for Apple Shortcuts and add them to your home screen, you can create a Shortcut that opens Duolingo using whatever icon you like, including the original application icon.

Here is what my Shortcut looks like on my home screen:

ALT

Two app icons on an iOS homescreen: Pocket Casts on the left, and Duolingo on the right.

Whereas my current Duolingo app shows the plaster icon, my Shortcut shows the icon I chose.

This approach has a limit: because it isn’t the original application icon, I am unsure if you will see the red application notification icons on the icon. This is acceptable for my use of the app; my phone is mostly in Do Not Disturb so I don’t see the notification icons anyway.

Furthermore, your custom application icon will not be used as the icon in push notifications. Indeed, my solution doesn't actually replace the application icon; it only lets you use your own icon on your home screen.

Here is how the Shortcut is set up:

ALT

My Duolingo Shortcut with a single step: Open Duolingo.

To change your app icon, create a new Shortcut with one step: to open an application. Choose Duolingo as the application to open. Set the name of the Shortcut to “Duolingo”. Next, download a copy of the Duolingo app icon and save it to your device.

Then, click the share icon in the Shortcuts application:

Click “Add to Home Screen”:

ALT

The Share task tray with several buttons, including one that says "Add to Home Screen" indicated by a plus icon.

Click "Image" to choose a custom image as an icon:

ALT

The home screen icon editor interface. "Image" is selected, which has three options: Choose Photo, Take Photo, and Choose File.

Click “Choose Photo” if you saved the app icon as a photo. Choose the app icon you want to use. Then, click “Add” to add the shortcut to your device:

ALT

The Duolingo app icon is set in the home screen app icon preview page.

You will now have a Duolingo app icon that looks like this:

ALT

My Duolingo shortcut.

You can then remove the original Duolingo app from your home screen and keep your shortcut.

While this approach is a workaround, it satisfies the requirement of allowing me to use a custom Duolingo app icon.