Nostr notes by David Chisnall (Now with 50% more sarcasm!)

I have used vim since around 2000. I have written five books, a ...

2026-08-28T06:43:42Z

I have used vim since around 2000. I have written five books, a PhD thesis, a few dozen papers and over 150 articles with it. At this point, my higher brain functions are not engaged at all when I use a bunch of common vim commands, they just happen. Documents I wrote with anything else have random :w in the middle.

So I tried NeoVim when it was quite new. Vim that you are familiar with, but better? Great!

Persistent undo is one of my favourite features of vim (introduced a few years after I started using it). My .vimrc includes a couple of lines related to this. One sets the path to be under ~/.cache so the undo files do not clutter the rest of the filesystem (I do the same for the swap files). The other sets the undo limit to something large (10,000, I think) because disk space is *enormous* relative to pretty much anything involving human-readable quantities of plain text.

I don’t often need the persistent undo. But on the few occasions when I have needed it, it’s been invaluable: ooops, I deleted something from this file, maybe last week and one reboot ago, what was it? Undo until I find it, copy it, paste it into the current version. Or, a bit more commonly: I had this working, then I tidied it up ready to commit, now it isn’t working, what did I do?

Vim has kept this working across major version upgrades over a period of about 20 years. I don’t even think about it, it’s just part of Raskin’s First Law: A program may not harm a user’s data or, through inaction, allow a user’s data to come to harm. If vim or the computer crash, or if I close a file and come back to it six months later, my undo history is still there.

The first thing I noticed in NeoVim was that undo didn’t work. I tried opening the file in vim and undo didn’t work there *either*.

Neovim had changed the format of the undo files. It hadn’t upgraded the old one. It hadn’t used a different name for its undo files. It had just noticed the existence of a vim undo file, deleted it (losing all of the data in it) and replaced it with one that vim couldn’t read.

I raised an issue about this and was told that the persistent undo format was unstable and users should not rely on data being preserved in a feature explicitly called *persistent* undo. It had changed once and would probably change again.

And that ended my experience with NeoVim. The authors showed immediately that they absolutely could not be trusted with any of my data. Breaking persistent undo is something I could forgive as a bug, but the attitude that just because something is a persistent file on your filesystem that contains data that you might want is no reason for their program not to delete it meant they had no concept of a duty of care to their users.

Apparently now they’re platforming white supremacists. I am completely unsurprised. Different facets of the same underlying attitude.

One of the saddest things about my time at Microsoft: I designed ...

2026-08-02T13:38:23Z

In reply to nevent1q…8hrw
_________________________

One of the saddest things about my time at Microsoft: I designed a privacy-preserving tip-jar system, where it would be possible for people to embed identifiers in their work (video file, web page, game, whatever) and any viewer (or program itself for programs) that wanted to support the system could record the amount of time spent with it and the user could decide how the reporting was scaled. Content providers would be able to tell whether the visitor subscribed to the system, but not who they were.

Users would pay either a flat fee or different tiers. At the end of a month, content creators would get a report of how many views they got and a payment. The company operating the system would have no visibility into what the users saw, but individual users would be able to get a report of where their money went. The operator would be able to de-list fraudulent or illegal providers so that their revenue shares should be discounted and they wouldn’t count.

The operator would be able to take a flat percentage for providing the payment service and any provider in the system.

Every part of was buildable, including verifiable security guarantees, with the technology available five years ago.

The problem? Microsoft would have no moat. Google would be able to trivially deploy such a system by making it a switch in their ads platform: if you are subscribed to the system, Google Ads don’t track you at all and the creator gets revenue directly. They’d be able to flip a switch and immediately opt most Play Store apps and an enormous number of web sites into the system, so even if Microsoft shipped it first, Google could dominate the market. The chief economist, correctly, identified that Microsoft would be spending a load of money to give Google a new revenue source, because a system like this is a natural monopoly.

My starting observation was that users don’t really like paying for individual things but they do like:<li>Paying a known and predictable amount for stuff that they like.</li><li>Knowing that people who created things they like are paid.</li><li>Not being tracked in their viewing habits.</li>

The system was designed to support all of these things.

Exactly. That's the loss-avoidance strategy. You are given ...

2026-07-30T11:49:23Z

In reply to nevent1q…wkf9
_________________________

Exactly. That's the loss-avoidance strategy. You are given £50K and told that it will be taken away if you flip the coin and lose. Even though a one in two chance of getting a 20x return is a *much* higher average return.

But, rationally, spending money for a 50% chance of winning 20x ...

2026-07-30T11:23:06Z

In reply to nevent1q…j0t6
_________________________

But, rationally, spending money for a 50% chance of winning 20x more is an obvious choice. And, indeed, a game that you would want to keep playing as many times as you could.

There are a bunch of interesting psychological experiments in this space. People are far more likely to avoid something that they perceive as loss.

The thing that makes this really interesting is: would you spend ...

2026-07-30T09:24:08Z

In reply to nevent1q…s9w4
_________________________

The thing that makes this really interesting is: would you spend £50,000 for a go at a game where you had a 50% chance of winning £1m?

So we should all not worry about the Mullvad VPN co-founder ...

2026-07-30T08:11:22Z

In reply to nevent1q…g9v8
_________________________

So we should all not worry about the Mullvad VPN co-founder donating 70% of the entire budget of your far-right party because money from corporate interests has no impact on Swedish elections? Good to know.

> This just means "I am more concerned about getting ...

2026-07-30T07:57:55Z

In reply to nevent1q…8ray
_________________________

> This just means "I am more concerned about getting re-elected than about the climate emergency."

No, it actually means ‘The measures needed to deal with the climate emergency are not acceptable to the industry lobbyists that fund my campaign or will run attack ads against me if I hurt their business’. And that has a very different set of solutions, starting with aggressive campaign finance reform.

The other extreme I often see is: Me: Here is an interesting ...

2026-07-27T09:56:15Z

In reply to nevent1q…6k5a
_________________________

The other extreme I often see is:

Me: Here is an interesting thing about a project I personally worked on.

Internet people: Do you have a citation for this?

Me: I am literally a primary source!

One of the open source projects I used to (many years ago, so I ...

2026-07-18T07:44:01Z

One of the open source projects I used to (many years ago, so I am not really affected anymore) has recently decided to allow AI contributions. One of their arguments is that they judge submissions on quality, not tools used.

Back when I was an active contributor (to the project and maintainer of some other parts of the surrounding ecosystem), I got a lot of submissions that were *bad*.

The thing is, I absolutely *loved* getting these because they always came from super-enthusiastic junior people. Reviewing them and giving feedback often too much longer than implementing the feature myself. The same was usually true for the next contribution too. But that shifted over time. After a few months, I started getting patches where I didn’t find anything beyond the cosmetic that I wanted to change.

Going back further, I *was one of those people* when I started contributing.

Measured as code contributions over the short term, maintainers working with submitters of poor-quality PRs is a waste of time. You’re taking time away from writing code to do something that produces code more slowly.

Measured over the long term, the results are very different. Far more code in that project and closely related parts of the ecosystem has been written by people who I mentored when they came along with some bad code than by me.

It’s never a waste of my time to turn an enthusiastic and incompetent contributor into an enthusiastic and competent contributor, it’s an absolutely critical part of building a healthy project. I wouldn’t have got nearly as much out of F/OSS on a personal level if I hadn’t encountered a lot of people with the same opinion when I was one of the enthusiastic and incompetent newcomers.

But LLMs are not like that. If I spend time reviewing LLM-generated code when it would have been faster to write it myself (which, to be clear, is the case on 100% of LLM-submitted PRs I’ve seen so far), then that’s just a time sink. It’s also taking away time I could be spending helping new contributors who want to actually improve, rather than just take the review comments, feed them into an opaque system, and give me back the results.

> LLMs will stay and people will use it. There's a big ...

2026-07-13T12:26:11Z

In reply to nevent1q…a03m
_________________________

> LLMs will stay and people will use it.

There's a big assumption here, that comes from the previous one:

> It is not that AI does not work, as often said here, it does work for many applications.

That's not the question. The first-order question is: do they deliver more value than they cost. And the answers to that are skewed by the fact that the current services are *heavily* subsidised and customers are paying 10-50% of the minimum viable price. So, if people actually have to pay the real price, how many of those use cases still exist?

The follow-on to that is, given that increasing the price by a factor of 2-10x will significantly reduce the demand, what will the resulting price be? Inference costs won't change much, but the price to users is the inference costs plus the training costs amortised across users (plus profit). Fewer users means that the training costs are amortised across fewer users and so the per-user costs go up. So how many use cases are still economically feasible if the prices go up by a factor of 20x?

But that's just the first order. The second order is: how affordable is it *when you factor in the cost of externalities*. Current LLMs are trained by doing something that, in any other context, would be a clear-cut case of large-scale copyright infringement. If LLM vendors had to actually pay for licenses for the things that they are ingesting in training, the cost of training would go up by at least a factor of 10x.

So there's a *huge* assumption in 'LLMs will stay'. The companies that train them do so on the assumption that they'll eventually be able to recoup that (huge) investment. If you don't train new ones, they become stale quickly. To be useful, they need to include recent events. For software, an LLM that doesn't know about new APIs and language features and does know about deprecated / removed ones is useless. For other uses, the shifts may be slower or faster. But training isn't a one-shot cost as a result, you're constantly needing to do retraining. And that's what needs the massive datacentres full of GPUs.

So, if you want to claim LLMs are here to stay, you need to answer a simple question: Who is going to pay for the training? If no one is, then how will LLMs remain useful?

The entire industry is [spending enormous amounts of money for tiny returns](https://isaiprofitable.com ). How do you expect this to shift to actually making money? Note that the numbers in this page look much better than they are, because all of that $516 B in revenue for NVIDIA is *spending* from other companies (which must be passed on to customers for anyone else to become profitable).

I saw that there’s now a mobile version of Roller Coaster ...

2026-05-04T08:31:59Z

I saw that there’s now a mobile version of Roller Coaster Tycoon (Roller Coaster Tycoon Touch) and I thought it might be fun (one of the Netflix bundled mobile games). A couple of hours of casual play in, it was clear that the game was carefully designed to make it progressively harder and harder to make progress without in-app purchases.

European Commission (npub1m96…ag4y) , if you want to *actually* make things safer online, how about making that kind of predatory practice illegal? Children are particularly vulnerable, but so are a lot of adults. No need for age verification, just an outright ban.

So sad to see a such a respected game series used for this kind of whale farming.

> whatever > [> #> ...

2026-04-28T13:03:06Z

In reply to nevent1q…ler7
_________________________

> whatever > [> #> signal](https://infosec.exchange/tags/signal )> 's reasons are for badgering users for a > [> #> PIN](https://infosec.exchange/tags/PIN )> , it's clearly a design choice they made, because other secure messengers don't do this.

The choice is either:<li>Periodically ask people to enter their PIN, or</li><li>Deal with people complaining that they forgot their PIN and are locked out (or, ideally not possible):</li><li>Provide an insecure way of recovering an account after you are locked out.</li>

The PIN entry UI looks nothing like an incoming message.

> they've been training users to fall for re-register > ...

2026-04-28T09:12:08Z

In reply to nevent1q…p46g
_________________________

> they've been training users to fall for re-register > [> #> scams](https://infosec.exchange/tags/scams )> by constantly prompting users to re-enter your > [> #> PIN](https://infosec.exchange/tags/PIN )> (and the PIN is only necessary because phone numbers are used for sign-up).

No, the PIN is required to reacquire the account if you lose all connected devices. If they used any other unique identifier as the account handle, the PINs would still be required.

One of the weird things in our industry is how many people who ...

2026-04-24T08:37:42Z

One of the weird things in our industry is how many people who participate in standards bodies have never had to maintain something with compatibility guarantees for a long time. If you go to one of the ISO groups that deals with plumbing standards, for example, everyone involved has had to deal with decades-old pipe work and has a visceral understanding of the consequences of standardising the wrong thing and having to change things during the lifespan of a building. But in every standards discussion I’ve been in, someone has proposed things that will obviously cause problems within ten to twenty years and half the group has nodded along and said ‘yup, that will definitely solve a problem I have now’. And saying ‘later we are going to want to do this other thing, and if people have built things with the assumption that putting this in the standard will create then we will have a massive migration headache’ has people complaining that I’m slowing down the process.

And that’s not to single out any standards group or process. I have had this experience in every standards process I was involved in (though, in the ones 20+ years ago, I was generally on the other side of the disagreement).

I don’t think anything Perens has said for at least 20 years ...

2026-04-23T07:38:33Z

In reply to nevent1q…tv5d
_________________________

I don’t think anything Perens has said for at least 20 years has been relevant to any real problems for the F/OSS ecosystem.

A few notes about the massive hype surrounding Claude Mythos: The ...

2026-04-17T10:04:41Z

A few notes about the massive hype surrounding Claude Mythos:

The old hype strategy of 'we made a thing and it's too dangerous to release' has been done since GPT-2. Anyone who still falls for it should not be trusted to have sensible opinions on any subject.

Even their *public* (cherry picked to look impressive) numbers for the cost per vulnerability are high. The problem with static analysis of any kind is that the false positive rates are high. Dynamic analysis can be sound but not complete, static analysis can be complete but not sound. That's the tradeoff. Coverity is *free* for open source projects and finds large numbers of things that might be bugs, including a lot that really are. Very few projects have the resources to triage all of these. If the money spent on Mythos had been invested in triaging the reports from *existing* tools, it would have done a lot more good for the ecosystem.

I recently received a 'comprehensive code audit' on one of my projects from an Anthropic user. Of the top ten bugs it reported, only one was important to fix (and should have been caught in code review, but was 15-year-old code from back when I was the only contributor and so there was no code review). Of the rest, a small number were technically bugs but were almost impossible to trigger (even deliberately). Half were false positives and two were not bugs and came with proposed 'fixes' that would have introduced performance regressions on performance-critical paths. But all of them *looked* plausible. And, unless you understood the environment in which the code runs and the things for which it's optimised very well, I can well imaging you'd just deploy those 'fixes' and wonder why performance was worse. Possibly Mythos is orders of magnitude better, but I doubt it.

This mirrors what we've seen with the public Mythos disclosures. One, for example, was complaining about a missing bounds check, yet every caller of the function did the bounds check and so introducing it just cost performance and didn't fix a bug. And, once again, remember that this is from the cherry-picked list that Anthropic chose to make their tool look good.

I don't doubt that LLMs can find some bugs other tools don't find, but that isn't new in the industry. Coverity, when it launched, found a lot of bugs nothing else found. When fuzzing became cheap and easy, it found a load of bugs. Valgrind and address sanitiser both caused spikes in bug discovery when they were released and deployed for the first time.

The one thing where Mythos is better than existing static analysers is that it can (if you burn enough money) generate test cases that trigger the bug. This is possible and cheaper with guided fuzzing but no one does it because burning 10% of the money that Mythos would cost is too expensive for most projects.

The source code for Claude Code was leaked a couple of weeks ago. It is staggeringly bad. I have never seen such low-quality code in production before. It contained things I'd have failed a first-year undergrad for writing. And, apparently, most of this is written with Claude Code itself.

But the most relevant part is that it [contained three critical command-injection vulnerabilities](https://beyondmachines.net/event_details/anthropic-claude-code-leak-reveals-critical-command-injection-vulnerabilities-e-6-c-1-k/gD2P6Ple2L ).

These are the kind of things that static analysis should be catching. And, apparently at least one of the following is true:<li>Mythos didn't catch them.</li><li>Mythos doesn't work well enough for Anthropic to bother using it on their own code.</li><li>Mythos did catch them but the false-positive rate is so high that no one was able to find the important bugs in the flood of useless ones.</li>

TL;DR: If you're willing to spend half as much money Mythos costs to operate, you can probably do a lot better with existing tools.

Nearly: > it was the amount of detail in the content of push ...

2026-04-16T07:26:41Z

In reply to nevent1q…gsfe
_________________________

Nearly:

> it was the amount of detail in the content of push notifications

It wasn’t the information in the push notification. This goes via Apple’s server and is a one-bit signal that says ‘there may be some messages waiting for you, you should go and check’ (may be, because Signal sends some spurious push notifications to make traffic correlations harder).

The Signal app then gets the message and asks the local OS notification mechanism to display the notification. If the permissions are set up to display Signal notifications on the lock screen, these are also persisted in a database on iOS (I have no idea why. Is there some way of searching them?). If you’re worried about people with physical access to your device reading your messages, I would suggest that turning off the thing that shows them on the lock screen is probably a good idea.

Wow, someone created a Mastodon account specifically to jump into ...

2026-04-12T15:38:29Z

Wow, someone created a Mastodon account specifically to jump into a thread and disagree with me about 'AI' hype. His example of how LLMs are adding value to the economy? They make cold calling more efficient for salespeople. Yup, spam really is the most important use case for LLM fans.

I’ve read a bunch of posts in the last few weeks that say ...

2026-03-22T10:59:05Z

I’ve read a bunch of posts in the last few weeks that say ‘Moore’s Law is over’, not as their key point but as an axiom from which they make further claims. The problem is: this isn’t really true. A bunch of things *have* changed since Moore’s paper, but the law still roughly holds.

Moore’s law claims that the number of transistors that you can put on a chip (implicitly, for a fixed cost: you could always put more transistors in a chip by paying more) doubles roughly every 18 months. This isn’t *quite* true anymore, but it was never *precisely* true and it remains a good rule of thumb. But a load of related things *have* changed.

First, a load of the free lunches were eaten. Moore’s paper was written in 1965. Even 20 years later, modern processors had limited arithmetic. The early RISC chips didn’t do (integer) divide (sometimes even multiply) in hardware because you could these with a short sequence of add and shift operations in a loop (some CISC chips had instructions for these but implemented them in microcode). Once transistor costs dropped below a certain point, of course you would do them in hardware. Until the mid ‘90s, most consumer CPUs didn’t have floating-point hardware. They had to emulate floating point arithmetic in software. Again, with more transistors, adding these things is a no brainer: they make things faster because they are providing hardware for things that people were already doing.

This started to end in the late ‘90s. Superscalar out-of-order designs existed because just running a sequence of instructions faster was no longer something you got for free. Doubling the performance of something like an 8086 was easy. It wasn’t even able to execute one instruction per cycle and a lot of things were multi-instruction sequences that could become single instructions if you had more transistors, Once you get above one instruction per cycle with hardware integer multiply and divide and hardware floating point, doubling is much harder.

Next, around 2007, Dennard Scaling ended. Prior to this, smaller feature sizes meant lower leakage. This meant that you got faster clocks in the same power budget. The 100 MHz Pentium shipped in 1994. The 1 GHz Pentium 3 in 2000. Six years after that, Intel shipped a 3.2 GHz Pentium 4, which was incredibly power hungry in comparison. Since then, we haven’t really seen an increase in clock speed.

Finally, and most important from a market perspective, demand slowed. The first computers I used were fun but you ran into hardware limitations *all of the time*. There was a period in the late ‘90s and early 2000s when every new generation of CPU meant you could do new things. These were things you already had requirements for, but the previous generation just wasn’t fast enough to manage. But the things people use computers for today are not that different from the things they did in 2010. Moore’s Law outpaced the growth in requirements. And the doubling in transistor count is predicated on having money from selling enough things in the previous generation. The profits from the 7 nm process funded 4 nm, which funds 2 nm, and so on.

The costs of developing new processes has also gone up but this requires more sales (or higher margins) to fund. And we’ve had that, but mostly driven by bubbles causing people to buy very-expensive GPUs and similar. The rise of smartphones was a boon because it drove a load of demand: billions of smartphones now exist and have a shorter lifespan than desktops and laptops.

Somewhere, I have an issue of BYTE magazine about the new one micron process. It confidently predicted we’d hit physical limits within a decade. That was over 30 years ago. We will eventually hit physical limits, but I suspect that we’ll hit limits of demand being sufficient to pay for new scaling first.

The slowing demand is, I believe, a big part of the reason hyperscalers push AI: they are desperate for a workload that *requires* the cloud. Businesses compute requirements are growing maybe 20% year on year (for successful growing companies). Moore’s law is increasing the supply per dollar by 100% every 18 months. A few iterations of that and outsourcing compute stops making sense unless you can convince them that they have some new requirements that massively increase their demand.

The other key thing is that money exists to provide liquidity. ...

2026-03-15T11:16:42Z

In reply to nevent1q…0hwk
_________________________

The other key thing is that money exists to provide liquidity. One of the most important purposes of money is to decouple the times when you buy and sell things. If I produce something of value now, I can sell it for some tokens that mean that, at some later time, I can get something I want.

This decoupling is very useful to any kind of functioning economy. If I didn’t sell the thing I made until I needed something, and then needed to find someone who wanted it who had the thing I want, trade would be very hard. But this implicitly means that there’s a temporal aspect to value that ends up influencing money.

When people talk about money being created out of debt, that isn’t a bug, it’s one of the most valuable properties of money. As production increases, so does the money supply, but *holding* money is disincentivised by inflation, which encourages you to buy things you need or invest in producing things someone will need later. The gold standard (and similar standards) tied the money supply to the supply of a single commodity. In the earliest known monetary system, this was tied to the food supply, which worked moderately well because the demand and production of most things correlated with the supply of food, but that worked only for a very simple economy.

I think there are a lot of related issues that mean that this ...

2026-02-01T09:34:36Z

In reply to nevent1q…yhad
_________________________

I think there are a lot of related issues that mean that this can’t be addressed in isolation.

Dismantling the polytechnics removed the best vocational qualifications. The best route into aerospace engineering, for example, used to be through a poly, but when they became a university they had to restructure the course to make it an accredited degree and that destroyed a lot of its value. There were several other subjects where a high-quality course at a poly was a better route in than a university degree.

This has meant that universities have had to take up that rôle as a path to employment. This, in turn, has led to jobs requiring ‘a degree’ but as a box-ticking exercise rather than anything that the job needs or as anything that gives the student either personal enrichment or transferable skills.

I saw some of the worst outcomes of this first hand when I was at Swansea. The Vice Chancellor shut down the Chemistry department because he wanted to get more students into the Media Studies department. The Chemistry department was rated 5 in the previous RAE, the Media Studies department was rated 1. There are some great Media Studies courses, which lead directly to careers in journalism (Cardiff used to have one, no idea if that’s still true. Again, this used to be something the polys did better), but the Swansea one was a waste of everyone’s time.

So there’s a massive push towards academic qualifications as entry to jobs, which means lots of people taking on loans to do degrees in which they have no interest, giving them a piece of paper that prevents them from being deselected but no knowledge or skills that they will actually use in the future. That’s enormously wasteful but also something that can’t be addressed by any intervention solely in the university sector.

There’s also the social aspect. I viewed my student loan as a poorly administered graduate tax: I took the loan and then, if I earned enough, I would pay more tax. But that’s a very middle-class attitude to debt. There are stark differences in how people view this kind of debt and working class people are far more likely to view any form of debt as a negative. So student loans end up skewing the demographics of university attendance away from working class people, at the same time as increasing the number of graduates is advertised as a tool for improving social mobility. That can be addressed by moving to a graduate tax.

I am increasingly convinced that universities fall far short of meeting their requirements both as centres of research excellence and as educational institutions, but fixing either or both of those requires some significant reforms (and defenestration of most university senior administrators).

One small rule change would help here: Companies may not buy back ...

2026-01-22T11:30:44Z

In reply to nevent1q…qnkq
_________________________

One small rule change would help here:

Companies may not buy back stock or issue dividends while they have outstanding loans.

Researchers have discovered a new vulnerability in FreeBSD’s ...

2026-01-21T09:31:08Z

Researchers have discovered a new vulnerability in FreeBSD’s telnetd:

If run in a non-default and unusual configuration[1], it implements the telnet protocol.

[1] Enabled.

Yoga teacher: Just breathe naturally, there are no wrong ways of ...

2026-01-16T18:44:31Z

Yoga teacher: Just breathe naturally, there are no wrong ways of breathing.

My inner monologue: Clearly you have never seen me drink a glass of water.

Can they do that? They don’t own the content. Editors can pick ...

2026-01-15T20:54:25Z

In reply to nevent1q…7jtu
_________________________

Can they do that? They don’t own the content. Editors can pick their own license and a lot use something like GFDL, which has attribution requirements.

For anyone surprised at Moxie Marlinspike pushing an 'AI' ...

2026-01-13T14:32:50Z

For anyone surprised at Moxie Marlinspike pushing an 'AI' thing, remember that the last big thing he pushed was a cryptocurrency integrated in Signal.

Just upgraded for #macOS 26. The UI reminds me of the period ...

2026-01-13T12:00:00Z

Just upgraded for #macOS 26. The UI reminds me of the period around 2002 when teenagers would create 'cool' GNOME and KDE themes with zero reference to usability. Except with more transparency. So like some of the early Compiz stuff, which was far more about 'the technology allows this' than wondering whether anyone would actually want it.

Safari windows are all totally different colours based on the background colour for the page. And they've lost the app icon in the bar (you know, the thing that's *really important* for Mac apps because it's the file proxy that lets you do direct manipulation?), so recognising them is hard.

It's a mess. Apple needs to seriously think about how they ever promoted people to leadership rôles the UI group when they were so clearly unqualified.

EDIT: People always complain about new macOS / OS X. I've been using OS X since 10.1, and written two books about Cocoa programming. I assumed the complaints were overblown. If anything, they're understated. It's so bad in a bunch of totally avoidable rookie errors that anyone who has done one undergraduate HCI course would know to avoid.

#LiquidAss

It's Policing 101, but it's common in Movie Policing: ...

2026-01-12T17:47:05Z

In reply to nevent1q…mz4e
_________________________

It's Policing 101, but it's common in Movie Policing: stand in front of moving cars, shoot towards them, and jump out of the way at the last minute in slow motion as they go past. And these people aren't trained, so they get their training from films.

Maybe someone involved in the movie / TV industry could educate some of the people who put this kind of scene in their shows about the real-world harm that this does?

I’ve taught programming like this, but I’m an increasingly ...

2026-01-11T14:44:31Z

In reply to nevent1q…qlcc
_________________________

I’ve taught programming like this, but I’m an increasingly huge fan of the debugging-first approach that a few people have been trying more recently. In this model, you don’t teach people to write code first, you teach them to *fix* code first.

I’ve seen a bunch of variations of this. If you have some kind of IDE (Smalltalk is beautiful for this, but other languages usually have the minimum requirements) then you can start with some working code and have them single-step through it and inspect variables to see if the behaviour reflects their intuition. Then you can give them *nearly* correct code and have them use that tool to fix the issues.

Only once they’re comfortable with that do you have them start writing code.

Otherwise it’s like teaching them to write an essay without first teaching them how to erase and redraft. If you teach people to get stuck before teaching them how to unstick themselves, it’s not surprising that they stop and give up at that point.

Apparently Samsung is putting ‘AI’ (and ads) in fridges. I am ...

2026-01-10T11:31:23Z

Apparently Samsung is putting ‘AI’ (and ads) in fridges. I am certainly not opposed to innovation and think that a lot of technologies that haven’t changed much for a long time have room for improvement (if you grew up in the US or Europe, try using a Japanese toilet and you’ll understand). But the frustrating thing is that there are a lot of *useful* things that a fridge could do with some computing power and a bit of electrical control that they *don’t* and which don’t need ‘AI’. Some examples:<li>It could talk to a smart meter to run the compressor when electricity is cheap and keep a reservoir of chilled coolant for when electricity is expensive.</li><li>It could automatically close the door if I leave it open for more than a minute and don’t first put it into ‘I am cleaning the fridge now’ mode.</li><li>It could record (with bar code or image recognition - the latter of which you could even market as ‘AI’ because apparently computer vision is ‘AI’ now) when I open things like milk or juice so when a thing says ‘consume writhing 5 days of opening’ and I can’t remember when I opened it, it can tell me.</li><li>It could then warn me if I’ve left something that will spoil in the fridge.</li><li>It could track when I remove / replace things so ‘are we nearly out of X’ is a question I can answer from my phone / tablet / laptop when organising a food order, without having to go and check.</li>

All of these are features I would actually find useful. The last ones require a bit of clever computer vision and good UI design, the earlier ones are just applications of mature technology.

And I would actually pay more for these features (and the first one would likely save me more money over the lifetime of the fridge than the price delta, so *everyone* wins [except oil companies, but that is a feature in itself]).

I could write a similar rant about pretty much every piece of electrical equipment I own. All of these have a load of trivial improvements that could be made if you start by asking the question ‘how do people use this and how do we improve it?’ Rather than ‘how do we add {an app,AI,this week’s buzzword} that marketing wants in the next generation?’

I’m not sure I fully understand this bit of the US legal ...

2026-01-07T19:52:54Z

In reply to nevent1q…77h3
_________________________

I’m not sure I fully understand this bit of the US legal system, but is there a reason that the state’s law enforcement cannot arrest ICE agents for murder and kidnapping?

A lot of the dialog around systemd and Wayland ends up with ...

2026-01-06T09:13:28Z

A lot of the dialog around systemd and Wayland ends up with someone saying at least one of the following:<li>You don’t get to decide what devs work on.</li><li>You are free to do something else if you don’t like it.</li>

And both of these are true. Indeed, the second is a core idea of Free Software. Free Software is about empowering users so that they are *not* beholden to the decisions that their software vendor made and are able to make different choices.

But most people (even most programmers) can’t decide they don’t want to use Wayland or systemd and write something different. These components are large monolithic entities. Even systemd, which is made of a bunch of coopretating daemons, has so much tight coupling between them that you can’t replace one of them without reimplementing 90% of its functionality. And each of these projects is too complex for a single person to create a replacement for unless they treat it as a full-time job.

To me, that really highlights the failure of the Free Software movement. It obsessed over licenses that prevent downstream developers from taking away rights (and making it harder for end users to exercise them) while never thinking about how to design software so that exercising these rights was easy and natural.

In a real Free Software system, option 2 should be so easy that a *large* fraction of users do it. Systems should be easy to shape around users’ requirements and preferences.

To all of the people worried about being 'left behind' by ...

2026-01-05T14:03:54Z

To all of the people worried about being 'left behind' by AI and who believe that AI is as transformational as the web:

I was an early adopter of the web. I had a 9,600 modem. I wrote web pages that had graceful fallback for browsers that predated the img tag. I dealt with the differences between Netscape Navigator 2.0 and Internet Explorer 2.0. I was even guilty of writing web pages that were 'Best viewed in Netscape Navigator 2.0'. I learned about the 10-10-10 rule (as web page should load in 10 seconds on a 14.4 Kb/s MODEM, have all of the key details visible on a 10" monitor, and have no more than 10 points per page).

A lot of the best web developers never experienced this era. We're almost at the point where the majority of them were not even born while I was doing this. And yet, being very late to the party hasn't been a problem for them.

If anything coming out of the bubble is as transformative as the web, you have *plenty* of time to catch up once it's stabilised into something of long-term value.

If it’s legally designated as a weapon, does this mean that you ...

2026-01-02T14:12:19Z

In reply to nevent1q…92wg
_________________________

If it’s legally designated as a weapon, does this mean that you have a constitutional right to own it?

Look, these extreme positions are completely unhelpful. Injecting ...

2026-01-02T12:38:36Z

Look, these extreme positions are completely unhelpful. Injecting bleach into your veins is obviously bad, but only a few people recommend that. And, yes, they're trying to sell bleach and are fully bought into the hype. But the other extreme, not consuming any bleach, is just as unhealthy. Bleach consumption is the future. You just need to do it in appropriate places, such as with a meal. I have one small glass of bleach with lunch and then a larger glass with dinner. I drink it, from a clear glass, so I can see exactly what I'm consuming. Obviously injecting it is a bad idea, but drinking bleach responsibly is good for you, as long as you know what you're doing. Of course, I'm an expert at drinking stuff, in the hands of an amateur it might cause problems. Don't be put off by all of the bleach refusers. They're just anti progress. Anyone who isn't getting experienced drinking bleach is going to be left behind. Just drink it responsibly.

Now, some people object to some of the process of manufacturing bleach. Yes, it does involve dumping toxic chemicals in rivers when you buy bleach from the big vendors. But that's why I use ethically sourced bleach, which comes from a factory in China where it has a certificate saying that it complied with all local environmental regulations. If you're worried about the ethical aspects of drinking bleach, always drink ethical bleach.

I've also recently heard people complain that people drinking and injecting bleach has driven up prices and hospitals can't afford to properly clean wards, leading to an increased infection rate. That's true *today*, but you're thinking in the short term. The increased demand for bleach means that the supply will increase and economies of scale will make bleach prices come down.

I went to a public school and just had a look at their fees. The ...

2026-01-02T10:04:45Z

In reply to nevent1q…pzaz
_________________________

I went to a public school and just had a look at their fees. The fees have gone up by about a factor of five in the last 25 years, well above inflation. If parents can afford to absorb that, the impact of VAT is fairly negligible.

To me, it reflect’s both Carroll’s view that all of his books ...

2025-12-30T11:39:05Z

In reply to nevent1q…meze
_________________________

To me, it reflect’s both Carroll’s view that all of his books were about mathematics and philosophy, his sense of humour (you *did* get what you asked for!) and his slight insecurity about being remembered for the works he regarded as more trivial rather than his serious academic publications. Of his non-fiction, I have only read his book on symbolic logic and I found the writing style quite engaging (though a *lot* more dry than in his fiction), and maybe he felt a monarch should have a stronger grounding in mathematics.

My favourite story about Alice's Adventures in Wonderland is ...

2025-12-30T11:31:50Z

In reply to nevent1q…vc5l
_________________________

My favourite story about Alice's Adventures in Wonderland is that Queen Victoria liked it and asked Lewis Carroll for a copy of his next book. His next book was The Fifth Book of Euclid Treated Algebraically, and he dutifully sent her a signed copy. I am not sure if she read it.

Mitre has just published their [top 25 most dangerous software ...

2025-12-17T07:50:24Z

Mitre has just published their [top 25 most dangerous software vulnerabilities of 2025](https://cwe.mitre.org/top25/archive/2025/2025_cwe_top25.html )

How does #CHERIoT stack up against this list?

5, 7, 8, 11, 14, and 16 are deterministically mitigated with just a recompile.

13 will trap, but is recoverable on a per-compartment basis.

15 is trivial to mitigate with compartmentalisation. [Phil Day wrote about this 18 months ago](https://cheriot.org/security/philosophy/2024/07/30/configuration-management.html ).

6 is mitigated by good capability-based filesystem APIs.

25 is mitigated by our software capability model in the RTOS.

1, 2, 3, 9, 10, 12, 22, and 23 and are not normally applicable on embedded platforms.

That leaves you with a lot more spare brainpower to think about avoiding the remaining seven (4, 17, 18, 19, 20, 21, and 24). The impact of many of these is limited in an environment where there is a programmer model that makes implementing the principles of least privilege and intentional use trivial.

For someone who only plays an economist on the Internet: what are ...

2025-12-16T19:13:38Z

In reply to nevent1q…93mu
_________________________

For someone who only plays an economist on the Internet: what are the likely consequences if non-bank institutions assets are wiped out? Is there something like the contagion that can spread through the banking sector, or will it just be a load of rich people losing money?

How are these people hired. I’d expect people whose previous ...

2025-12-15T09:18:35Z

In reply to nevent1q…wvcw
_________________________

How are these people hired. I’d expect people whose previous rôle was basically the same job as the one they’re being hired for to be asked ‘so how did you improve things in your last post?’ And then someone to judge them based on this. Since most of them have, at best, not done irrevocable damage, I don’t see them having good answers.

Note that this 2.5 figure is a lie. I learned earlier in the year ...

2025-12-15T09:05:51Z

In reply to nevent1q…2jwa
_________________________

Note that this 2.5 figure is a lie. I learned earlier in the year that hospitals are allowed to count reclining chairs as ‘beds’ for reporting purposes. The ward nurse explained this to me while I was with someone with appendicitis who *really* needed to lie down and these were the only ‘beds’ available.

So even that low number is actually an exaggeration of the real capacity. Many of the beds are not actually usable for people staying more than a few hours.

And then the remaining beds are used inefficiently. In a couple of weeks visiting a 4-bed room, I saw multiple people stuck for most of a day (more than a day, in a couple of cases) because:<li>The pharmacy took hours to deliver the drugs they needed to take home on discharge.</li><li>There was no transport available, ambulances can’t even keep up with emergency transport.</li><li>The doctor who needed to sign off on the discharge couldn’t be reached on the ‘instant’ messaging system and there was no process for ‘discharge this patient if these conditions are met’.</li><li>They needed some outpatient care and there wasn’t a carer available.</li>

To he fair to him, there's absolutely nothing in his prior ...

2025-12-10T08:22:53Z

In reply to nevent1q…jv7x
_________________________

To he fair to him, there's absolutely nothing in his prior professional experience that would require any knowledge of human rights law and its history.

It’s a shame that it lists summarisation as something LLMs are ...

2025-12-09T08:26:43Z

In reply to nevent1q…5amk
_________________________

It’s a shame that it lists summarisation as something LLMs are good at, when all of the studies that measure this show the opposite. LLMs are good at turning text into less text, but summarisation is the process of extracting the key points from text. LLMs will extract things that are shaped in the same way as a statistically large number of key points in the training set but they don’t understand either the text of the document *or* your context for requesting a summary and so are very likely to discard the thing that you think is most important. They also have a habit of inverting the meaning of sentences when shrinking them.

At some point, someone needs to start charging Snake Oil Sam as ...

2025-12-04T09:01:37Z

In reply to nevent1q…gqrv
_________________________

At some point, someone needs to start charging Snake Oil Sam as an accessory to the various crimes his product is encouraging.

Aren’t you glad that the Chancellor is encouraging everyone to ...

2025-12-02T09:23:19Z

In reply to nevent1q…36l6
_________________________

Aren’t you glad that the Chancellor is encouraging everyone to invest their savings in stocks just before the market crashes?

RE: https://mastodon.world/@jeffowski/115630556942027240 I ...

2025-11-29T09:57:46Z

RE: https://mastodon.world/@jeffowski/115630556942027240

I boosted this because, a few years ago, I was at MS when our lab decided to organise a collection for the Salvation Army in the lobby of the building. Every gay or trans employee had to walk into the building past an advert, endorsed by the organisation’s senior leadership team, for an organisation that had a documented (recent) history of leaving people like them to starve on the street.

At the time, I was chair of the Diversity and Inclusion committee. I flagged this with the leadership team and not one of them was aware of this reputation. I was, because I am not *completely* oblivious to the world around me (no more than 80%).

I tried (and failed) to get them to institute a policy that the organisation should do some basic due diligence before endorsing a charity. The bar I recommended was to open the charity’s page on Wikipedia and read the ‘controversies’ section. I was deeply disappointed that a leadership team that talked a lot about diversity and inclusion decided to keep endorsing the Salvation Army and refused to institute such a policy to avoid this kind of thing in the future.

quoting
note1rkr…3x4m

#lgbtq #lgbtqia #salvationarmy #redbuckets #bigotry #addiction #recovery

Wait, there’s a BluRay version? They said they weren’t going ...

2025-11-27T21:05:56Z

In reply to nevent1q…d9dy
_________________________

Wait, there’s a BluRay version? They said they weren’t going to do an HD version because a load of the CGI shots were VHS quality and they’d lost the master footage and they’d look absolutely terrible next to HD scans of the scenes they have on film. Did they redo the CGI somehow, or some clever upscaling, or does it just look bad?

The next dominoes in the AI bubble that I expect to fall (if ...

2025-11-27T08:55:24Z

The next dominoes in the AI bubble that I expect to fall (if you’d excuse the mixed metaphor):<li>Insurance companies explicitly exclude coverage of any system using AI and any outputs of AI systems.</li><li>Lawyers in big companies issue advice that using AI systems is too high risk.</li><li>Big companies demand IT suppliers provide an enterprise-management system switch to disable all AI functionality in products, or provide an AI-free version.</li>

The first is starting. A consortium of insurance companies has asked their regulator to approve this blanket exclusion. Their argument is that the risks of these systems are too unpredictable to be able to insure. They can’t reason about systemic or correlated risk if you add a bullshit generator anywhere in an operational flow.

The second has happened in a few places, but is not widespread. Some places are hedging. When I was at MS, the AI policy was basically: ‘look, we give you all of these shiny toys! Please use them! By the way, you accept all legal liability for their output! Have fun!’. One ruling that this kind of passing-the-blame-to-employees-for-correctly-using-company-provided-tools policy is unenforceable and the lawyers will get very nervous.

The third is a consequence of the first two. If your lawyers tell you something is high risk and you can’t buy insurance, you want to make sure it isn’t used.

Windows BSoDs were pretty common back then. My favourite one of ...

2025-11-26T17:02:37Z

In reply to nevent1q…a4gf
_________________________

Windows BSoDs were pretty common back then. My favourite one of these stories is that the very first Xbox public demo had a BSoD. Apparently Bill Gates was quite angry at that and said that Xbox must never show a blue screen in a demo ever again. And this is why the Xbox has a green crash screen.

A few things here: First, I used to fly small planes, but was ...

2025-11-25T09:03:32Z

In reply to nevent1q…pk7q
_________________________

A few things here:

First, I used to fly small planes, but was lucky enough that the RAF paid for it. I looked at getting a PPL and the price of fuel was not a significant part of the cost. It would become one under this proposal, but light aircraft consume surprisingly small amounts of fuel. A lot of it is on takeoff, which you need to practice for a couple of hours, but fuel efficiency improves a lot if you’re flying longer distances, and a lot of the important bits are navigation training. But servicing on the airframe, the requirement to rebuild the engine periodically, hangar costs, and so on are all very expensive.

Second, electric light aircraft (not to be confused with orchestras) are starting to be good enough to replace internal combustion engines for small propeller planes. They’re a similar price for new ones, but there are enough second-hand light aircraft available that the market for new ones is pretty small. I’d expect this to shift a bit if the costs of operating avgas-powered ones went up.

Third, the pipeline for commercial pilots is *amazingly* exploitative. I don’t see making airlines pay for it properly, rather than leaving a load of debt-saddled failed commercial pilots along the roadside, to be a downside.

I proposed this the last time we had a Labour ...

2025-11-25T08:48:55Z

In reply to nevent1q…rfqz
_________________________

I proposed this the *last time* we had a Labour government:<li>Fuel duty on all aviation fuel, to be gradually scaled up to 1.5x the tax paid for petrol, over a period of five years (slow introduction to avoid massive price increases to consumers and to allow planning).</li><li>Triple duty on aviation fuel for private planes, including any plane flying with fewer than fifteen passengers or some equivalent in cargo (perhaps with exceptions for airlines needing to fly a replacement plane after a failure).</li><li>Duty paid on all fuel in the tanks when you land above a safety margin (so you can’t avoid it by only refuelling on one hop of a short round trip).</li>

I still think it is the right strategy.

For those who are skeptical that AI is a bubble, let's look ...

2025-11-24T11:53:30Z

For those who are skeptical that AI is a bubble, let's look at the possible paths from the current growth:

**Scenario 1: Neither training nor inference costs go down significantly.**

Current GenAI offerings are *heavily* subsidised by burning investor money, when that runs out the prices will go up. Only [8% of adults in the US would pay *anything* for AI in products](https://www.zdnet.com/article/only-8-of-americans-would-pay-extra-for-ai-according-to-zdnet-aberdeen-research/ ), the percentage who would pay the unsubsidised cost is lower. And, as the costs go up, the number of people willing to pay goes down. The economies of scale start to erode.

**End result**: Complete crash.

**Scenario 2: Inference costs remain high, training costs drop.**

This one is largely dependent on AI companies successfully lobbying to make plagiarism legal as long as it's 'for AI'. They've been quite successful at that so far, so there's a reasonable chance of this.

In this scenario, none of the big AI companies has a moat. If training costs go down, the number of people who can afford to build foundation models goes up. This might be good for NVIDIA (you sell fewer chips per customer, to more customers, and hopefully it balances out). OpenAI and Anthropic have nothing of value, they start playing in a highly competitive market.

This scenario is why DeepSeek spooked the market. If you can train something like ChatGPT for $30M, there are hundreds of companies that can do it. If you can do it for $3m, there are hundreds of companies for which this would be a rounding error in their IT budgets.

Inference is still not at break even point, so costs go up, but for use cases where a 2X cost is worthwhile there's still profit.

**End result**: This is a moderately good case. There will be some economic turmoil because a few hundred billion have been invested in producing foundation models on the assumption that the models and the ability to create them constitutes a moat. But companies like Amazon, Microsoft and Google will still be able to sell inference services at a profit. None will have lock in to a model, so the prices will drop to close to the cost, though still higher than they are today. With everyone actually paying, there won't be such a rush to put AI in everything. The datacenter investment is not destroyed because there's still a market for inference. The growth will likely stall though and so I expect a lot of the speculative building will be wiped out. I'd expect this to push the USA into recession, but this is more the stock market catching up with the economic realities.

**Scenario 3: Inference costs drop a lot, training costs remain high.**

This is the one that a lot of folks are hoping for because it means on-device inference will replace cloud services. Unfortunately, most training is done by companies that expect to recoup that investment *selling inference*. This is roughly the same problem as COTS software: you do the expensive thing (writing software / training) for free and then hope to make it up charging for the thing that doesn't cost anything (copying software / inference).

We've seen that this is a precarious situation. It's easy for China to devote a load of state money to training a model and then give it away for the sole purpose of undermining the business model of a load of US companies (and this would be a good strategy for them).

Without a path to recouping their investment, the only people who can afford to train models have no incentive to do so.

**End result**: All of the equity sunk into building datacentres to sell inference is wasted. Probably close to a trillion dollars wiped off the stock market in the first instance. In the short term, a load of AI startups who are just wrapping OpenAI / Anthropic APIs suddenly become profitable, which may offset the losses.

But new model training becomes economically infeasible. Models become increasingly stale (in programming, they insist on using deprecated / removed language features and APIs instead of their replacements. In translation they miss modern idioms and slang. In summarisation they don't work on documents written in newer structures. In search, they don't know anything about recent events. And so on). After a few years, people start noticing that AI products are terrible, but none of the vendors can afford to make them good. RAG can slow this decline a bit, but at the expense of increasingly large contexts (which push up inference compute costs). This is probably a slow deflate scenario.

**Scenario 4: Inference and training costs both drop a lot.**

This one is quite interesting because it destroys the moat of the existing players and also wipes out the datacenter investments, but makes it easy for new players to arise.

If it's cheap to train a new model and to do the inference, then a load of SaaS things will train bespoke models and do their own inference. Open-source / cooperative groups will train their own models and be able to embed them in things.

*End Result*: Wipe out a couple of trillion from the stock market and most likely cause a depression, but end up with a proliferation of foundation models in scenarios where they're actually useful (and, if the costs are low enough, in a lot of places where they aren't). The most interesting thing about this scenario is that it's the worst for the economy, but the best outcome for the proliferation of the technology.

**Variations**:

Costs may come down a bit, but not much. This is quite similar to the no-change scenario.

Inference costs may come down but only on expensive hardware. For example, a $100,000 chip that can run inference for 10,000 users simultaneously, but which can't scale down to a $10 chip that can run the same workloads. This is interesting because it favours cloud vendors, but is otherwise somewhere between cheap and expensive inference costs.

**Overall conclusion**: There are some scenarios where the outcome for the *technology* is good, but the outcomes for the *economy* and the major players is almost always bad. And the cases that are best for widespread adoption for the technology are the ones that are *worst* for the economy. And that's pretty much the definition of a bubble: A lot of money invested in ways that will result in losing the money.

We (SCI Semiconductor) are about to hire some folks in the next ...

2025-11-24T10:24:26Z

We (SCI Semiconductor) are about to hire some folks in the next couple of months (probably starting in January, since we're about to hit Christmas):

We're aiming to hire 1-3 FAEs, who can build out the open-source bits of the #CHERIoT software stack (including drivers / various communication stacks), build demos, and work with customers on use-case bringup.

We also want to hire someone else on the toolchain side. Primarily #LLDB + #OpenOCD, but also working with our #LLVM (and #RustC) folks.

Let me know if you're interested!

EDIT: We are a full-remote company. It's easiest for us to hire people in the UK (and one of our investors would *really* like us to hire more people in Sheffield), but elsewhere is possible (though might, for tax purposes, require you to be officially a contractor for a while).

We're also going to be hiring people for our hardware verification and RTL teams soon (more on the verification side than design at the moment, I think). I'm not responsible for them, but I can find out more details if anyone is interested. Our first CHERIoT chip is nearly finished, we're starting to work on the second.

#GetFediHired

It depends somewhat on the outcome. If we see proper ...

2025-11-23T12:14:11Z

In reply to nevent1q…tezq
_________________________

It depends somewhat on the outcome. If we see proper accountability, then it’s money well spent. That means repayment of money gained fraudulently, large fines for all companies involved, and prison for gross negligence for everyone in senior government rôles.

If it’s just a ‘oh dear, lots of people were corrupt and stole a load of money and got tens of thousands of other people killed. Such a shame we can’t do anything about it’ then I can think of better uses of the money.

This post is an example of why no one listens to #FreeBSD ...

2025-11-22T08:47:19Z

In reply to nevent1q…etdq
_________________________

This post is an example of why no one listens to #FreeBSD advocates. They start by not understanding a Linux thing, and then claim a totally unrelated FreeBSD thing was there first and is better.

The soft reboot thing in RedHat is about providing an immutable base image (FreeBSD does not do this) and then a lightweight way of restarting userspace to use it.

The thing you link to is the reroot feature of FreeBSD, which was *a copy of a Linux feature*. Linux has had pivot root for a long time, the thing that is added is an administration layer that uses this functionality for a quick update path, integrated with the normal update flow. This does not exist on FreeBSD.

Similarly, FreeBSD has had jails for ages. Linux has also had shared-kernel virtualisation for almost as long. OpenVZ shipped five years after Jails (and before Jails had things like isolation for SysV IPC and so were actually useful for isolating workloads like Postgres). The value of Docker / OCI containers is not that you can create an isolated environment, it’s that it has a distribution model built on immutable layers and an orchestration model that lets you cleanly separate persistent data (volumes) from the software that runs on them so you can upgrade by simply rebuilding the image and then destroying and recreating the container. And FreeBSD now, finally, has an alternative to OCI containers on Linux: OCI containers on FreeBSD.

If you spent half the effort understanding why people like and use some of these features on Linux as you do telling people that barely related features on FreeBSD are better, then you might actually do some useful advocacy. As it is, you just reinforce all of the negative stereotypes the Linux users have about FreeBSD.

If I, in my professional life, ignored expert advice and it ...

2025-11-21T08:10:48Z

In reply to nevent1q…nlzf
_________________________

If I, in my professional life, ignored expert advice and it resulted in a death, I would be personally prosecuted for gross negligence and my company for corporate manslaughter. If there were evidence of malicious intent, I would be prosecuted for murder.

If it were 23,000 people instead of one, I would expect never to leave a prison cell. I would also not be surprised if a court revoked my company’s corporate charter (the ‘corporate death penalty’) and seized all of its assets for restitution.

When politicians wonder why the public has little faith in politics: this is it. A politician who does this is not in prison. He is not paying any fines. The party that supported and enabled him is still allowed to collect and disburse donations, and is still allowed to field candidates in elections. Neither suffers any financial or professional penalties. And the public debate is over whether the perpetrators should be forced to apologise before going back to collection million-pound salaries. It’s absolutely disgusting that we hold politicians to less account than we do the officers of a small business.

Yes. I have wanted to do polls on several occasions in Signal and ...

2025-11-20T07:30:27Z

In reply to nevent1q…6zds
_________________________

Yes. I have wanted to do polls on several occasions in Signal and resorted to clunky things like posting each option and asking people to react to the one they want (which then floods me with notifications because Signal *still* doesn’t let you turn off notifications for reactions even though they are never a thing that requires an urgent reply).

Moreover, it’s one of the things that consistently comes up in ‘why don’t you use Signal’ questions. Not having polls is a big barrier to adoption just as the lack of stickers was. I have zero interest in stickers, but I do want people to be using Signal so that I don’t need to persuade them to install it when I want to add them. Features like stickers and polls increase the likelihood of this.

Some years ago, I did a fire-warden training course. Most of it ...

2025-11-17T08:42:07Z

Some years ago, I did a fire-warden training course. Most of it was pretty obvious to anyone who had played with fire (for example, aim for the base of the fire with the extinguishers, don’t just use the carbon dioxide ones to fan the flames). But it had one observation that probably saved a lot of lives over the years:

> There’s no such thing as a two extinguisher fire.

If a fire extinguisher can’t extinguish a fire, by the time you have got another one, the fire will have grown big enough that the second won’t work either. Use one extinguisher, and if that doesn’t put the fire out completely then evacuate the building and wait for the fire brigade to come.

It’s obviously not universally true, but the number of exceptions is small and the cost of getting it wrong is measured in lives, so it!s a very sensible rule of thumb.

Yay, the embargo was finally lifted yesterday so we can talk ...

2025-11-14T08:27:45Z

Yay, the embargo was finally lifted yesterday so we can talk about the UK Government funding for #CHERIoT!

They [funded us (SCI) to do two projects, for a total of £7.7M](https://www.ukri.org/news/21-million-backing-for-technology-to-stop-cyber-attackers/ ):<li>Bring <a href="https://infosec.exchange/tags/Rust"; class="mention hashtag" rel="tag">#Rust</a> on CHERIoT to production qualities.</li><li>Build our second-generation chip with a dual-issue core, post-quantum crypto hardware, and an edge inference accelerator.</li>

Not just natural language processing. It’s also the largest ...

2025-11-14T07:54:44Z

In reply to nevent1q…c0h4
_________________________

Not just natural language processing. It’s also the largest public archive of spreadsheets. When I was at Microsoft, a bunch of projects used it. For example, when the TypeScript version of the Excel calc engine wanted to see how good their coverage was, they tried to see how many of the Enron sheets they could correctly calculate (as in, give the same answers as desktop Excel, not give the answer without all of the fraud).

Good news: AI will achieve human levels of intelligence next ...

2025-11-11T08:39:19Z

Good news: AI will achieve human levels of intelligence next year!

Bad news: This will not happen as a result of improvements to AI.

I remain astonished by the fact no one at OpenAI thought ...

2025-11-08T12:33:56Z

I remain astonished by the fact no one at OpenAI thought announcing publicly that they were probably going to default on a trillion dollars of debt might be the thing that causes the bubble to start to burst.

I briefly tried using Microsoft's news app when I worked ...

2025-11-06T19:13:59Z

In reply to nevent1q…jjlr
_________________________

I briefly tried using Microsoft's news app when I worked there. 100% of the ads it showed me were for scams. I flagged this as a reputational risk, but was met with silence.

Until there's legal liability, I don't see this changing.

In my ideal system, they'd abolish the realtime debates ...

2025-11-06T12:08:15Z

In reply to nevent1q…lt59
_________________________

In my ideal system, they'd abolish the realtime debates entirely. They're purely performative and are in no way a good process for evolving legislation. All laws would be discussed in an online forum where MPs could leave comments and have them addressed, in a process more like code review. Once 50% of MPs have hit approve, the law is passed. Changes invalidate prior approvals.

The review system should be publicly readable, so anyone can observe the law taking shape and write to their MPs with specific issues related to the text.

Possibly. Most MPs spend at least some of each week in their ...

2025-11-06T09:53:53Z

In reply to nevent1q…0mxa
_________________________

Possibly. Most MPs spend at least some of each week in their constituency though, so my thought was that they’d go to this week’s Parliament and then home, then to the next location. Moving their staff might be harder.

When they were talking about having to close the Parliament ...

2025-11-06T09:32:37Z

In reply to nevent1q…9z7m
_________________________

When they were talking about having to close the Parliament buildings for renovation, I had two proposals:

Parliament should rotate between constituencies on a weekly basis. Each week, it’s in a different constituency.

MPs must use public transport to attend. If they cannot prove that they used public transport, they may not vote.

I think that would have led to a quite rapid shift in priorities.

Honestly, that might not be the worst outcome. If Labour, the ...

2025-11-05T21:16:32Z

In reply to nevent1q…gmf4
_________________________

Honestly, that might not be the worst outcome. If Labour, the zombie Tories and Reform Ltd are all splitting the vote on the right, that leaves the Greens able to round up a sizeable majority. I don’t care the name of the party, as long as it has the right ideals.

I met a new friend this morning! Her name is Billie and there was ...

2025-11-05T14:29:57Z

I met a new friend this morning! Her name is Billie and there was a little poster up explaining that she is not lost.

She was *so floofy!*.

#CatsOfMastodon

People seem to be bullying trees here now. ...

2025-11-05T14:25:06Z

People seem to be bullying trees here now.

Was there a word limit? It feels like it gets about half way ...

2025-11-04T17:39:07Z

In reply to nevent1q…kgj4
_________________________

Was there a word limit? It feels like it gets about half way there and then has to use two short paragraphs to wrap up, rather than talking about the rest of the financial chicanery and failure to deliver that's going on.

I didn't want to say 'this reads like AI nonsense' ...

2025-11-04T15:01:29Z

In reply to nevent1q…lprf
_________________________

I didn't want to say 'this reads like AI nonsense' because I didn't want someone who wrote something that bad and tried really hard to feel bad. But I did. And it was. So he should feel bad.

It turns out that prose is like code: it takes me far less time to write it from scratch than it takes to fix the nonsense that LLMs extrude.

Ugh, I thought this document was full of nonsense. It turns out, ...

2025-11-04T14:36:48Z

Ugh, I thought this document was full of nonsense. It turns out, it was Copilot slop.

It’s also what happens when you confuse holding elections with ...

2025-11-04T09:18:21Z

In reply to nevent1q…3wj7
_________________________

It’s also what happens when you confuse holding elections with representation.

I said this a while ago on a deeply nested thread, but I think ...

2025-11-03T12:29:29Z

I said this a while ago on a deeply nested thread, but I think people miss a critical part of why Excel is popular. It's not that it's a good programming language (the calc language in Excel is pretty bad, though the reactive programming model is nice) it's that it's an *amazing* debugger. Excel makes it trivial to visually inspect every variable (which it calls a cell). You can debug expressions (which are automatically updated) in exactly the same programmer model as the real expressions. You can create debugging views that mirror some important variables, without learning a different tool.

Build a debugger as nice as Excel and anyone can learn your language.

I’m not saying it couldn’t work, but a few problems to ...

2025-11-03T08:48:56Z

In reply to nevent1q…g3cm
_________________________

I’m not saying it couldn’t work, but a few problems to consider:

Salt water is corrosive. Piping it anywhere around expensive electronics is usually a bad idea.

The heat for datacenters is concentrated on top of chips. The biggest problem is moving this heat *away from* the chips quickly. The second problem is getting it outside. You don’t want to be evaporating salt water on the chip, because you will instantly get a layer of salt forming as an insulator and cooking the chips. You would want to do this on the secondary circuit, but then the temperatures are lower.

Sea water is not just salt water, it is full of alive things (from single-celled up). All of these also residue when you boil the water. But a lot of seaweed has evolved to latch onto areas near fast-moving currants so that it can extract nutrients as the water goes past. So have a lot of molluscs. Basically, anything that has sea water flowing through it rapidly fills up with nature.

Take a look at the published results from the MS undersea data centre experiments for some of the problems in this space.

Understand what people need that keeps them on Windows. Build ...

2025-11-02T10:01:49Z

In reply to nevent1q…4ll6
_________________________

Understand what people need that keeps them on Windows. Build things that address those problems. Don’t shame people for the fact that your tools don’t solve their problems.

To me, one of the weirdest things about academia is how much they ...

2025-11-02T08:42:31Z

In reply to nevent1q…s4rx
_________________________

To me, one of the weirdest things about academia is how much they purport to value domain expertise, yet pretend that the entire field of assessment theory doesn’t exist. I’ve rarely met someone who wrote any kind of university assessment (exam, coursework, project, whatever) who knew the difference between formative and summative assessment, or knew what ‘construct validity’ means. I did some basic item response theory of an exam put together by academics at one of the top universities in the world and found that it did not, with statistical significance, measure *anything* (determining whether it measures the *right thing* is a different and harder problem).

With that baseline, it’s not surprising that cheating is easy. It’s also not surprising that students don’t feel that doing the assessments is worthwhile.

When I was Director of Studies at a Cambridge college, I told my students that they had two goals: Acquire an education in the subject and do well in some exams. It was vital that they did not confuse the two. I would have much preferred for the latter to be a consequence of the former.

To everyone who says ‘install Linux’ as a solution to Windows ...

2025-11-02T08:22:49Z

To everyone who says ‘install Linux’ as a solution to Windows problems (or ‘install LibreOffice’ as a solution to MS Office problems), imagine how you would react if you posted about Linux problems and all of the responses were ‘install Haiku’. Would it work? Maybe, but the person saying it has no idea what you do on Linux, whether the apps that you use run on Haiku (or have alternatives that you could switch to). Neither of you have a clear understanding of how much time and effort the switch would take, or what the network effects are that make switching hard.

If you have actually engaged with them, understood their requirements fully, and have a migration path outlined for them, great. If you’re just being a dick, think about how that makes people perceive Linux users: Maybe that isn’t a great way of driving adoption.

#Linux #Haiku #Windows

I still think about a New Fellows’ induction at my college a ...

2025-11-01T08:34:35Z

I still think about a New Fellows’ induction at my college a few years ago where they had all of the new fellows give a short overview of their research. And there was one whose talk was ‘I’m an aerospace engineer! I work with volcanologists! I build drones that fly into volcanoes! Here is a video of one of my drones flying into a volcano! We send them in to collect gas samples and return, but they also have lots of sensors and telemetry. So we fly them lower and lower until they catch fire! Here is a video of one of my drones catching fire and falling into the lava!’

The guy who went after her was the most upstaged I have ever seen anyone be. He had to follow that with ‘I study French poetry, here is a French poem. It has themes.’ And I felt bad for him because there was no possible talk that could not have been a letdown in his slot. Unless, perhaps, it was ‘I train resurrected dinosaurs to use rocket launchers’.

I did not have Trump killing the Fortran 77 compiler industry on ...

2025-10-30T15:05:13Z

I did not have Trump killing the Fortran 77 compiler industry on my bingo card.

(The F77 compiler industry is entirely subsidised by the DOE, because they have a codebase written in F77 that, by law, cannot be modified unless its revalidated. It can be revalidated only by modelling a nuclear bomb exploding then actually exploding one and seeing whether they are the same. They would love to move it to F90 or newer but can't as long as the test-ban treaty is in effect.)

I refer to it as the site with the sans serif swastika logo.

2025-10-30T12:42:36Z

In reply to nevent1q…h283
_________________________

I refer to it as the site with the sans serif swastika logo.

There might be some confounding factors here. 10% of instances go ...

2025-10-30T07:33:08Z

In reply to nevent1q…s2za
_________________________

There might be some confounding factors here. 10% of instances go down, but the remaining 90% are now posting ‘LOL AWS’ and ‘LOL Azure’ at a frequency that offsets the drop in traffic from the missing ones.

I, for one, am happy that Google is introducing mandatory ...

2025-10-29T11:58:31Z

I, for one, am happy that Google is introducing mandatory registration for Android developers. It sends a very clear message that Google is willing to accept criminal liability for all Android malware, since they are the gatekeepers who permitted it to be distributed. It will be much easier to bring prosecutions for violating things like the Computer Misuse Act when there is a single entity accepting all liability.

@npub1f7r…cfnk @npub1wfp…w2m9 has [written about ...

2025-10-27T13:44:21Z

Princess Nettles Succubish (npub1f7r…cfnk)

Soatok Dreamseeker (npub1wfp…w2m9) has [written about Session](https://soatok.blog/2025/01/20/session-round-2/ ), they don't seem to be solving these problems well.

A lot of Signal's marketing talks about security. Since the Snowden revelations, and the Ex-NSA Chief's 'We Kill People Based on Metadata' comment, metadata confidentiality (specifically: who is talking to whom) is very important.

And that matters in a lot of contexts. If you're going to a protest and you coordinate with other people, keeping that group's membership private matters. If you're organising a union, you don't want it to be easy for someone to leak the set of people planning to unionise. If you're a whistleblower talking to a journalist, you don't want it to be easy for an employer to see who are all of the people who are talking to journalists with phones connected to the corporate WiFi) And there are a lot of lower-stakes versions of this.

And all of the high-stakes use cases are improved by having more people use the same system. This builds an *anonymity set*. This was one of the big problems with Tor: very few people used it for boring things, so simply using Tor flagged you as a person of interest. If you have a communication system that everyone uses for low-stakes things, and another that people use for when they really need the confidentiality, then it's easy to identify people who use the other thing. For example, if you're thinking of unionising and everyone planning to unionise uses Signal, but no one used Signal before they started unionising, it's easy for an employer to get a list of which people's phones suddenly started using Signal.

That said, if all you care about is message confidentiality and you're happy to leak all of the metadata, XMPP+OMEMO does that.

I've hear some folks using it to avoid writers' block by ...

2025-10-27T11:01:13Z

In reply to nevent1q…03d9
_________________________

I've hear some folks using it to avoid writers' block by prompting an LLM and then saying 'that's absolute crap, I can write it much better!'

I can’t speak for fd.o, but I was in a leadership position on ...

2025-10-25T08:25:02Z

In reply to nevent1q…uwpn
_________________________

I can’t speak for fd.o, but I was in a leadership position on another project where we got a similar case disastrously wrong, so I might be able to illuminate how that happens.

The first mistake we made was not to differentiate harassment from conflict resolution. Most of the issues we had between contributors were personality clashes or technical disagreements that escalated. As you say, most of these have both parties acting in good faith. The main thing that the project needs to do is deescalate and get the people involved to talk again. This is absolutely the wrong approach in cases of harassment. There were two key causes of this:

First, (as you mentioned) no one involved had any formal (or, in most cases, informal) training in how to deal with harassment. Most employers offer this, but it’s rarely compulsory. After the initial incident, I signed up for this training with my employer (as did another colleague involved with the same project). This highlighted some of the things we did wrong, but it was quite illuminating who was there: we were the only men on the course who were there voluntarily. Most of the people were women who were there because they had been targets of harassment or bullying and wanted to understand the processes better. The rest were men who had been forced to take the training because they had been accused of harassment (and, from a lot of their comments, I suspect had been engaged in it long term).

Most F/OSS (or other community-led) projects don’t have any formal structure for providing this kind of training. And the work-provided training wasn’t sufficient. There were a bunch of ‘and this is where you need to escalate it to HR specialists (or the police)’ moments, but volunteer projects don’t have those experts. One of the biggest things a F/OSS charity could do to improve the situation would be to hire real experts that projects can use as consultants. Companies that back projects could help out be loaning HR as well as engineers to the projects.

Second, we had very poor visibility into what happened. There’s a natural tendency for humans to trust the first person who explains a situation. In our case, it was made worse because the only thing that happened on project infrastructure (and so the thing that we saw) was an IRC exchange where one project member connected and had a go at another member then left. We didn’t see the backstory, which involved a load of gamergate nonsense on Twitter and elsewhere (and those of us not in the Twitterverse had only a very vague idea of what Gamergate was. I thought it was a handful of people who were upset some game they didn’t like won an award, I had no idea that it was a coordinated harassment campaign). When a lot of the things that happened are private messages, or in non-project spaces, it’s hard to know what the real context is. We saw a load of things quoted out of context that made both people look bad. We also had friends of both people jumping in and defending them and attacking the other.

It really takes *weeks* of investigation to properly handle this kind of thing and dig to the truth. And this compounds the problem of the people dealing with it not having the right training. And, unless they are employees of a foundation backing the project, they also lack the time to do a good job. And, again, the assumption that people are basically decent (which is normally valid) hurts when one of the people is not and is actively trying to subvert the process. The evidence from an honest person reporting what happened and a dishonest person cherry-picking out-of-context comments will look very similar. Unless you personally know the people involved (which brings its own problems of bias) then it’s very hard to work out who is telling the truth. This is even harder when one or both people involved are highly visible in the community, because they will both be publicly sharing a narrative and one is mostly accurate (but only mostly: no one is 100% objective when they’re being personally attacked) while the other is a carefully crafted fabrication, but there’s pressure to respond quickly because both are public and the community is full of people who believe either one and are complaining.

In the last few years, the problem has become worse. A lot of CoC complaints now are malicious. Far-right folks absolutely love baiting people into saying things that look bad when quoted out of context, then deleting the context and reporting the remark. They make a game out of trying to get people kicked out of projects. So the workload has gone up, which compounds the other problems.

I wish I had a good answer for how to improve this.

That doesn’t look like a rich person’s house, that looks like ...

2025-10-24T06:08:11Z

In reply to nevent1q…u9xu
_________________________

That doesn’t look like a rich person’s house, that looks like something desperately trying to look like one. None of the furniture *quite* goes, and it’s all that special kind of bland you get from trying to copy something that a good designer did, without quite understanding why they chose any of the pieces. It’s a common look in generic corporate hotel lobbies, where a load of executives made A Contribution and the poor designer had to accommodate them, while also balancing taste against secure knowledge that everything would get ludicrous amounts of wear and so either needed to be very hard or easy (and cheap) to replace. All while trying to *look* luxurious.

Real rich people either hire a decent designer and let them get on with it, or have a load of totally mismatched things that they either inherited, have some sentimental attachment to, or just happened to like and are sufficiently secure in their own taste that they aren’t bothered by the fact that it’s a mishmash of styles.

This follows the Silicon Valley model popularised by Facebook 20 ...

2025-10-23T07:57:16Z

In reply to nevent1q…0khy
_________________________

This follows the Silicon Valley model popularised by Facebook 20 years ago of opting people into consent for things op because they were in other people’s address books and *those* people consented to sharing personal information. It’s a shame it took regulators so long to stamp on that, it should have been the result of massive fines, possibly followed by fire.

My problem is that I honestly have no idea what that percentage ...

2025-10-20T05:24:12Z

In reply to nevent1q…06zp
_________________________

My problem is that I honestly have no idea what that percentage is. Is it 0.1% waste or 50% waste? I'm pretty confident that it's between those two values, but I have no idea where it is.

And the related problem is that there is no accountability for this kind of waste. When Kier Starmer decides he wants to waste £1bn on an AI buzzword nonsense project, what is the personal cost to him? Nothing (except possible losing the next election and having to live on just the generous parliamentary pension and all of the huge speaker fees that a former PM can charge).

And that's really the problem I have with a lot of politics at the moment: the complete lack of accountability. Politicians can do things that every expert will tell them are bad ideas, and that will have the opposite effect to their stated goals, and there are no personal penalties for them when it turns out that, yes, really, the experts weren't all lying.

I think there are legitimate concerns here. Pretty much every ...

2025-10-20T04:50:28Z

In reply to nevent1q…75ny
_________________________

I think there are legitimate concerns here. Pretty much every major government IT project has been an order of magnitude more than it should have cost, and then failed to deliver because they didn't bother doing proper requirements analysis involving the people who actually have to use the system. HS2 was massively over budget and then the bits that would actually add new connectivity were cancelled. They have done quite a nice job with one of the roads near me, making it much more friendly to cyclists and pedestrians, but it took 18 months to refurbish a couple of miles of road. Even if that had been free, the cost of the disruption was astronomical.

I was visiting a hospital recently and the staff were great, but the infrastructure was a mess. This goes from the physical bits to the IT infrastructure. The main lifts were not working and had not been working for six months due to a dispute between the owners of the building and the NHS over who was liable. Oh, and cardiology was on the fifth floor, and almost no one going there was physically capable of taking the stairs, so there were queues for the one working lift. I saw someone who had been waiting to be discharged for over 20 hours: doctors had approved his discharge as soon as he had the prescription filled, and that happened only when the ward nurse got a porter to make a special trip to the pharmacy and bypassed the systems that were meant to do it. I saw several people staying in beds in wards for an extra day or more because there weren't ambulances available to take them home, or because they needed to have a regular at-home carer visiting and there weren't any available. I saw doctors taking 5+ hours to reply to messages asking if a patient is allowed pain killers because the nurses send them messages on an IM system, and they see those messages only when they log into a desktop computer, which needs them to set aside 10 minutes to reply to a five-second query, and they're overworked.

I saw a huge number of examples of money being wasted because the system was inefficient.

Successive governments have been complaining about inefficiencies in the NHS for as long as I can remember. And each time, they've paid expensive management consultants to make it worse, rather than doing the obvious thing of asking nurses what the top three things are that make them do redundant work and fixing them.

It's quite easy to see that public money is being wasted. It's not so obvious what *percentage* is wasted. When you're talking billions in spending, 0.1% waste has a large visible impact, but doesn't really matter that much.

Safari has a ‘reading list’ feature which is a good ...

2025-10-15T01:06:22Z

In reply to nevent1q…p08s
_________________________

Safari has a ‘reading list’ feature which is a good alternative to opening tabs. It puts interesting pages in a list you never look at, so you know they won’t get lost, but there’s no danger that you would accidentally read them.

Little known fact: If you have XYZZY chromosomes, it unlocks a ...

2025-10-10T15:39:00Z

Little known fact: If you have XYZZY chromosomes, it unlocks a secret gender and you win the patriarchy.

I disagree with a lot of people here, and that's fine. We can ...

2025-10-09T14:32:41Z

I disagree with a lot of people here, and that's fine. We can disagree on what tax rates should be. We can disagree on the best mix of public and private ownership. We can disagree on the amount of government involvement in markets. We can disagree on what privacy rights need to be protected by governments versus which need to treat the government as the adversary. We can disagree on which OS is best. We can disagree on the best way to build a Free Software ecosystem (we can even disagree on whether that's a desirable goal).

But we can't disagree on which people are people. If your perspective is that some people are not people, there is no way that we can still have a meaningful conversation. That's not a 'there are two sides' kind of debate. That's a 'no, this isn't a debate and we can't build a functioning society if you disagree' kind of debate.

Apple originally just did a code dump for every release. The KDE ...

2025-10-09T14:23:52Z

In reply to nevent1q…dcd6
_________________________

Apple originally just did a code dump for every release. The KDE folks asked them if they could be nicer open source citizens and so they created a public svn repository (I think this moved to git at some point?) for collaboration.

In addition to the KDE folks, a bunch of other companies jumped in early. Nokia, for example, did a Symbian port and a load of memory-efficiency improvements. Apple kept all of the ports in their tree.

Did KDE get any return? Well, QtWebKit uses WebKit, and this is wrapped in KDE-specific frameworks for use in KDE apps. KDE gets a complete browser engine, that Apple has put hundreds of person years of effort into, that can be embedded easily in any KDE application. They get JavaScriptCore, an interpreter plus three-tier JIT that Apple wrote almost all of. They get a privilege-separation model that Apple built into the core libraries ('WebKit2') so that it's reusable, so things like mail clients can safely use WebKit to render HTML mail without bugs in the engine compromising the entire application.

Apple also did a few things like rewriting some of the bits of WebKit to use libcurl directly, rather than the CFURL-based wrappers, so that porting it to other platforms didn't require reimplementing bits of CoreFoundation.

I would strongly suspect that KDE is pretty happy with the return on investment that've got from KHTML: I'm generally happy if people contribute more to my projects than I do. If you view the world as a zero-sum game where someone else making money makes you sad, even though they give you something of significant value along the way, then I can see that Apple and Google's use of KHTML code would upset you.

If you are a progressive talking about immigration, please try to ...

2025-10-09T07:47:03Z

If you are a progressive talking about immigration, please try to frame your messaging as follows:

Immigration is, according to a load of studies, an overall benefit to the economy. But we all know from personal experience that ‘good for the economy’ is not always good for individuals. Immigration can suppress wages, at least in the short term and in local areas, because immigrants are easier to exploit than a lot of other groups. They’re unplugged from any support networks they had growing up, they may not be aware of all of their rights (they may *have* fewer legal rights), they may miss nuances of language, and so on. And this means that unscrupulous businesses can under pay, over work, and otherwise take advantage of them.

This is even more true of refugees (who literally don’t have anywhere else to go) and people on work visas sponsored by their employers (and so are dependent on their employer’s good will to remain). Work visas that allow partners to come over but not work places more pressure on the single wage earner and makes them even more dependent on their employer.

By creating a class of people that are easy to exploit, unscrupulous businesses can make you directly compete in the labour market with people who have no choice but to accept conditions that you would find unacceptable.

If you’re worried about immigrants taking your jobs, the solution to this is to provide *stronger* worker rights for immigrants and big penalties for any employer who is found to have violated minimum wage laws. It’s to end loopholes where people act as employees but have the rights of self-employed workers. It’s to make sure that people on high-skills visas can easily move jobs and must be paid a fair market wage.

When someone is demonising immigrants, it’s because they benefit from having an exploitable underclass. The more afraid immigrants are, the easier it is for employers to take advantage of them. And it’s hard for you to compete in the job market against people who have no choice except to accept whatever terms their employer offers.

This is what the study last year on usage at Google showed: ...

2025-10-07T11:27:13Z

In reply to nevent1q…52nx
_________________________

This is what the study last year on usage at Google showed: self-reported increase in individual productivity correlated with decrease in overall measured team productivity. People are able to be wrong faster, which slows down everyone around them.

If you want to understand humans, start by looking at clip-on ...

2025-10-06T07:40:42Z

If you want to understand humans, start by looking at clip-on ties. These are the result of a surprising amount of design work to look like normal ties, but they have the benefit that you can wear them around heavy machinery and, if they get caught, your head doesn’t get pulled in and squashed between cogs. Most humans prefer to avoid painful death, so this has clear benefits.

Now, those of you more familiar with rational species may ask ‘wouldn’t it be simpler to just change the dress code so that people operating the machinery didn’t have to wear ties?’ And, of course, the answer is yes. And, it turns out, factory managers are not complete idiots and realise that having skilled workers killed by machinery (and then having to take the machines out of service to clean the blood and bones out) was not beneficial to profits.

But ties were worn by professionals. People would rather ignore a rule that would literally save their life than violate the social norm that people in their in group wore ties. Being perceived as not being a member of the professional class was a bigger risk than possibly dying.

Eventually, these things shifted and (almost a hundred years after the clip-on tie was invented), more people are happy to not wear a tie as an in-group signifier (they have other ones, don’t worry, humans didn’t suddenly become sensible).

Sometimes they shift because a respected person ignores them. For example, the Department of Computer Science and Technology in Cambridge is informally known as the Computer Laboratory. This used to be its official name and before that it was the Mathematical Laboratory. At the time it was founded, all faculty were required to wear gowns most of the time (in lectures and so on), but there was one exception. For entirely sensible not-being-on-fire-related reasons, you didn’t have to wear gowns in laboratories. Maurice Wilkes really hated gowns, so designated the entire department as a lab: no gowns permitted.

He was willing to violate the norms, and gradually the norms changed.

Note also that he could do this because there was a loophole in the rules (as there always is). Give humans a set of rules that they disagree with and they will search for loopholes. Give them a set of ludicrous cultural norms and they will defend them with religious fervour.

Rules don’t change how people behave. Rules that actually work encode things everyone agrees with and provide you with a mechanism to deal with the outliers who refuse to follow them. Most people refrain from murder because they think living in a society where people aren’t randomly killed is a good idea. Laws against murder don’t prevent it, they give you a process to deal with the tiny minority who think killing people is a good solution to problems.

With very few exceptions, successful laws *follow* changes in broad consensus on acceptable behaviour. They don’t cause these changes.

Part of the problem in the software industry (and one that open ...

2025-10-05T09:08:08Z

Part of the problem in the software industry (and one that open source partially addresses) is that the business models fundamentally do not reflect economic realities.

Software is, to within a rounding error, a *zero marginal cost good*. This means that, once you’ve made one of a thing, the additional cost of another (the marginal cost) is zero. Most things benefit from economies of scale. If you make one bespoke widget, you have a bunch of costs in fabrication on top of the materials costs. If you build a factory to create widgets, the cost of the factory is big, but it’s *amortised* across all of the widgets you make. If you spend $1,000,000 building a factory and aim to sell 1,000,000 widgets, the factory cost contributes $1 to the cost of each widget. For simplicity, I’m going to lump all of the R&D effort involved in designing the widget and figuring out how to build it in the factory costs. They are slightly different classes of costs (they don’t scale with the rate of production, whereas if you want to make widgets twice as fast you will need a bigger and more expensive factory) but they’re broadly similar.

If you somehow increase your market and can now sell 10,000,000 widgets, the factory cost adds only $0.10 to each widget. The more you sell, the lower your fixed up-front costs are as a proportion of your total costs. As these go down (or are paid off by early profits, so the factory is now effectively free), other costs matter more. Factories aren’t magical replicators. You need to feed them with raw materials (or components). You need to pay people to operate them. You need to pay for maintenance and repairs on the factory itself.

Even after the cost of the factory and R&D are all paid for, there’s still a cost. If you’re selling widgets for $10, these costs may add up to $8/widget and, as the other costs are paid off your profit goes from $1/widget to $2/widget. The model where you build the factory as an investment and aim to recoup the money selling widgets makes sense to you, but also to your customers. They can view it as spending $8 on materials and operating costs and $2 renting a factory (which is far cheaper than the $1,000,000 to build their own).

Software does not work this way. The cost of writing software is high. It takes time and skill, and more time and skill to test it. If I wanted to write a word processor, for example, I’d need to hire a moderately large team (tens to hundreds of people, depending on my exact feature requirements) and it would take them a few years to ship the product. Once they have built it, creating a second copy costs (to a first approximation) nothing. If you have a bunch of users who want a word processor, I can sell it to you, and the cost to me is no different if you have one user or 10,000 users.

Business models that make sense for zero-marginal-cost goods are tricky. If it cost me $10,000,000 to make and I think I can sell it to 100,000 people, I can pick a price of $100/user. If you are buying it as an individual, that seems like a great deal: far lower than the cost of building it yourself. But if you are a company or government with 1,000,000 users, you could build it in house for a tenth the amount. Except you aren’t a software company, so maybe you don’t want to. And I really want that sale, so I can give you a discount.

And that’s where it starts to get weird. If I can guarantee 100,000 sales at $100 each, any additional sale I make at *any* (positive) price is profit. If you want 10,000,000 licenses, I can five you a great deal of only $500,000 and that’s still pure profit for me. Wonderful.

But, at some point, I’ve sold my word processor to everyone who wants a word processor, and I’m employing a team of people who are experts in word processors. Because I’m selling software as an immutable artefact, the obvious thing for me to do is figure out some new features people want and do a *new version*.

New versions are great because they cost you a lot less to make than the original (they are incremental changes) but would cost a new competitor more because they would have to start from scratch. Now, my total development cost is $15,000,000, but I’ve already paid for two thirds of that. I had to fix some technical debt, but a new competitor would have to spend $13,000,000 to reach the same feature set, so I can sell this *as if* it cost me as much to make as the first version and customers still think they have a good deal. And maybe you give existing customers a discount on an upgrade version, so they are even more likely to stick with you than go elsewhere.

Eventually though, I hit diminishing returns. You, the customer, look at version 47 and say ‘actually, I can’t think of a single feature introduced since version 42 that I use, why should I pay for the new version?’. And a big customer may think ‘I’ve been buying this for ages and version 23 was pretty much feature complete from my perspective, maybe I should pay someone else to build an equivalent of version 23 that I own and can add new features to if I need them’.

My incentive is to make the program big and complicated so that you don’t really believe you could create an equivalent. The more features, the better. Because *someone* probably needs them, and you need to be able to interoperate with them and so you can’t replace my product with something that isn’t exactly feature compatible.

To address these risks, I you need to *keep* adding features and convincing users that they want them. Spell checking, grammar checking, cloud syncing, AI, and so on. Some of them are actually useful, some I put marketing effort into so people are convinced they need them even if they don’t know why.

And, hey, if that fails, I can always make some changes to the file format so that old versions can’t open the new documents. Oh, sorry, your customers are using version 50 to send you purchase orders? I guess you’ll need to upgrade to be able to read them. Please talk to my sales team, I’m sure we can come to some arrangement.

And all the while, there’s an elephant in the room: bugs. Version 38 had a nasty parser bug that let someone who sent you a maliciously crafted file compromise your computer. Version 39 had one where if you made your company name bold, it would crash. And you do want these fixed.

And now we get to a really tricky problem. If I’d done the level of formal verification required to *absolutely* guarantee that none of these bugs existed, I’d have had to charge $10,000 for v1 and no one would have bought it. At the same time, you (quite reasonably) think you paid for your copy and so should expect bug fixes (and, especially, fixes for security issues): it was sold as a working product, even if the small print did have some disclaimers of liability.

So I propose a compromise. You buy the current version and I’ll give you free bug fixes for three years. But what bugs will I fix? That bug with crashes for bold on your company name is triggered on your corporate letterhead and nothing else. Engineer time to fix that will cost more than I made on sales to you. So maybe I don’t fix that one especially, but it turns out that v40 had refactored that code while adding a new feature and now the bug is gone. So now you need to buy the new version. Unfortunately, the new features also introduced some new bugs.

Okay, this clearly isn’t working. I have a new idea. I’ll sell you a *subscription* instead of a product. For $50/year, you are *always* running the latest version. This is good for me in two ways. First, I don’t have to maintain back-ports of bug fixes to old versions, I just have one supported version. Second, I have guaranteed revenue. It’s good for you because you can predict your spending well in advance.

But, again, at some point you are going to have spent more on the subscription than it would cost to have built the subset of the product that you need for yourself.

So my incentives are the same. I must make the app as big as possible so you don’t think you can replace it. I must tightly couple it to all of my other products, so you can’t replace just one bit. Sure, you *could* replace my word processor, but it talks a proprietary protocol to my cloud back end for collaborative editing, and that uses the same namespace as my file storage, so you lose that if you switch. And the collaboration is built around my proprietary account management system which you use to restrict access to people in your company (and specific people or groups).

I build plugin interfaces that make sure the plugin is subservient to my program, so other people can add value to my ecosystem but can’t become competitors, rather than allowing interoperability with peers where you can select the components you want and that might not include any of mine.

To maintain my business model of writing software for free and then charging you to use it, I must build things that are bad software engineering and user hostile.

A F/OSS system has none of those incentives, which is why it makes me so sad to see them copying these structures. A F/OSS platform could focus on lightweight interoperable components that users can compose in arbitrary ways. A system in which the cost of developing a new component for a specific use case is small enough that individual programmers can do it and other people can afford to pay a programmer to do it is great for a F/OSS ecosystem. Paying someone to add a feature to a massive monolith like LibreOffice is hard (not least because they must be able to upstream it, which requires time and effort from upstream maintainers, if you want to use it long term). Paying someone to write a component for a modular component system is much easier and doesn’t require anyone else to like it.

Component systems were popular in the early to mid ‘90s, but died out because they are hostile to the COTS business model. Microsoft realised that replacing Office with a set of COM components that composed to build arbitrary document classes was completely feasible, but paved the way for competitors to incrementally replace them and displace MS Office’s dominance. The trend has been in the opposite direction, towards walled-garden operating systems with siloed apps, each striving to become the legendary ‘everything app’.

F/OSS doesn’t need to copy this. It can build structures that proprietary software simply cannot replicate because they are directly incompatible with a rent-extraction model for funding development. And, in this world, licenses like the GPL are both unnecessary (someone wants to incorporate my component in a proprietary system? Go ahead, it won’t be competitive in my ecosystem) and harmful (in a system with a hundred individual components tightly collaborating, where my document includes a custom plugin that is a derived work of ten other plugins, what are my legal obligations if I send this document to my lawyer under attorney-client privilege?).

Which is why high (and gradually increasing) corporation taxes ...

2025-10-04T09:37:38Z

In reply to nevent1q…wykd
_________________________

Which is why high (and gradually increasing) corporation taxes were originally proposed. They make each stage of growth harder, which makes companies that are almost big enough to act as monopolies unable to compete with smaller companies that have lower tax rates. In particular, this incentivises investment in the smaller companies (it’s *much* easier for them to to grow), which starves the big companies of capital.

Unfortunately, someone convinced politicians that having a load of megacorps was a sign of a successful economy. It’s depressing to see so many EU politicians say ‘where is the European Microsoft / Google?’ The fact that companies of that scale don’t exist is a feature. The right question is ‘where are the dozen companies that compete in this one market where Microsoft is dominant in the USA?’

Nostr notes by David Chisnall (*Now with 50% more sarcasm!*)