Posts by doublelayer

Windows 11 24H2 can run – sort of – in 184MB

Most of that is going to be the RAM used by the browser. Windows can evict stuff from cache and free up a lot of that RAM, but the browser generally doesn't and happily uses as much as it can.

This is, incidentally, the answer to Richard's question: "The Reg cannot help but wonder why Windows 11 24H2 has the requirements it does, given that the core of the operating system clearly doesn't need them." It has them because they don't want people to use a device with 2 GB, run a browser that wants to use all of that to keep ten tabs open, and blame Windows for why it's awful. 4 GB is not needed for Windows itself, but their recommended minimum for running Windows and having enough room for other applications. They especially don't want people to be manufacturing devices with 2 GB and telling their irate customers that "Microsoft said this was okay" when they call to complain.

AI can improve on code it writes, but you have to know how to ask

"Not knowing knowing what it did makes the article seem speculative and a little boring. (Sorry)."

We do know what it did. All the generations are posted here. Most of the optimization that it did was farming out the calculations to libraries. It did quite a few things that are typical of good software, but almost in the way that people showing off would. The rest of the optimization was adding additional features that nobody asked for, which could be positive or negative. Sure, some longlasting tasks might benefit from a metrics reporting server, but for a task that originally took half a second to execute back when it was allocating a few megabytes when it didn't have to, that might not be the most useful and it was definitely not specified.

Re: I'm not following here

Yes, in the same way that you can do that to an interview candidate. This problem is exactly the kind of thing I've been asked during interviews, where they are looking for me to demonstrate the most efficient way of achieving a goal that nobody cares about. When you're actually working there, they don't care in the slightest how long the computer takes because 0.5s of server time is free unless it's at scale, and instead they're optimizing for lowest writing time and/or lowest difficulty during maintenance. Writing it more efficiently is something the programmer tries to do even though management is mostly ignoring whether they have.

What this won't do so well at is getting bugs out of some code. As demonstrated, as it tried more and more advanced methods, it introduced more and more bugs. It also started expanding outside its brief. For example, the first version creates a function which does the task provided and returns the result, an integer, as an integer. By version number three, it was printing it out to the console with extra text. It wasn't told to do that. Maybe we wanted to use that result in a calling function, but now we're going to have to fix that. That expansion, by guessing requirements, just increases the number of places where bugs can crop up. That massive speedup, by the way, wasn't from the LLM finding a new, more efficient algorithm. It was by using a library to run it outside of the Python interpreter. Of course, if you need this answer really fast, not writing the number-crunching part in Python might have been a starting point.

The bot's output was somewhat impressive to me, but mostly because I've seen much worse code generated for even simpler cases. It wasn't because it did anything particularly difficult.

Apple shrugs off BBC complaint with promise to 'further clarify' AI content

My money's only on the latter. Apple won't face any serious libel suits over this because they can inundate anyone in counteractions until they go away. It will take something they hate seeing, trumpeted by actual media, for them to try to stop.

Now Trump's import tariffs could raise the cost of a laptop for Americans by 68%

Re: Is this madness unbounded...?

I think the only fundamental point is that Trump really likes broad, general, and high tariffs, always has, and has convinced enough people that they will solve their problems. They are now going to find out whether they actually do. Of course, there may be obstacles to imposing all the tariffs he talks about having because he does like to make promises he can't keep, but tariffs are more in his power than most other things he's promised. For example, he has no chance of implementing an income tax offset from the tariff; he can't do it unilaterally and politicians are not going to help him, but he can impose most of the tariffs without external approval.

Dude, you got a Dell, period! RIP XPS, Inspiron, Latitude, Precision

This is called marketing. Marketing specifically calls it rebranding, but you don't have to worry about what they're talking about because what it really means is that they wanted to change something because they hadn't in a while and this is what they picked. They're not trying to make things worse for you to extract more money. In fact, they probably think they're helping by simplifying things a little. After all, was there any clear difference between a randomly chosen Latitude and a randomly chosen Vostro that made the names useful? I don't know of any. Or maybe they thought they were removing an old name so they sound newer, but even then it's a benign change. Companies do this all the time. It wasn't very long since AMD significantly renumbered their processors, which happened because Intel significantly renumbered their processors, which happened because AMD slightly renumbered their processors or because Intel thought that sticking an "Ultra" into the brand name would not look as silly as it actually does, which they probably did because of the "Pro Max Ultra" thing that Apple did, which they did because they wanted to have more expensive tiers of iPhones. It doesn't have to affect our lives very much.

Re: Simple?

Yes it is. Their problem seems to be that they want Apple's branding simplicity. However, Apple makes only a few types of laptops. Understanding the differences between a MacBook Air, MacBook Pro 14 inch, and MacBook Pro 16 inch is not hard because there are only three of them (three models and three differences, they're pretty much the same laptop except for the obvious). Dell has hundreds of models, so it isn't easy to know the differences between all the available ones or even which ones are available. That allows them to serve more markets, and there's a reason for them to do it, but it unavoidably requires making the product naming more complex. Changing the first words in the brand isn't going to make any of it simpler.

What it should mean:

Dell: Word processing, email, web browsing. You should get this for pretty much everybody.

Dell Pro: If you're doing something that needs more processing and memory and want something larger.

What it actually means:

Dell: Cheapest, probably, check that it actually is.

Dell Pro: Basically the same stuff but more expensive.

As with pretty much every manufacturer, the only way to figure out what is in a model is to ignore the brand name and only look at internals and price. Once you've found one that looks to be the cheapest that has what you need, then look at the model number to work out repairability. Sometimes, the brand is related to that, but often, it's pretty random. Often, I've found a model that's theoretically the low-end range with better specs than a more expensive one in the higher range.

Google's 10-year Chromebook lifeline leaves old laptops headed for silicon cemetery

Re: "...The lowest-end hardware, which is a little cheaper than comparable Windows machines..."

I'll have to take your word for how well a low-end Chromebook is, though I have heard complaints which suggest that your observation may not be universal. Since I have used no Chromebooks, I can't counter it with experiences of my own. I have experienced some low-end Windows computers, and it's not quite as dire as you paint it. I have a feeling that many people who have such things do not like them because it is not difficult to use them in a way that makes them miserable to work with. However, I have had the misfortune to use Windows 10 on 2 GB of RAM and an EMMC and, to a lesser extent, Windows 11 on 4 GB of RAM and also an EMMC. I could browse around with that. I could read email and edit documents with it. I could write code on it, which probably wasn't in your list but that is one of the things I did most on that machine. It was far from ideal, but it wasn't useless. I'm therefore not entirely convinced that Chromebooks are a significantly better user experience at the low end, and if they are, I would guess without further description that the difference is that it's harder to make Chrome OS do inefficient things.

Re: End of Chromebook?

Many, if not most, Chromebooks are already at comparable prices to normal laptops. People are still buying them enough for new models to get made. I don't know why, but they are. That means they'll probably be good to continue selling those things for some time. The lowest-end hardware, which is a little cheaper than comparable Windows machines but not as much as you might think, is probably not going to last the full decade for reasons other than software updates. Early obsolescence is still in their toolbox.

Boffins carve up C so code can be converted to Rust

Re: Why?

"So what actually happens say in a production system?"

You see a compiler error and fix your code. Or you see a compiler error, decide you can't or won't fix it, and label it. That's the point of checking for and not allowing those things. It requires a different language in this case, but it's very common for C compilers to complain about certain things that can be compiled but shouldn't be, and it is also common for a production system to operate on the policy that you don't have compiler warnings on your build unless you can explain why they are not a problem here.

Re: Rust, Agile...and then some questions....

Rust was designed to fix a couple classes of problems, not everything in software development. No language can fix or even slightly help with problems 1 or 2. Technical debt is a very broad topic, but if you have and count memory safety issues as a tech debt thing, then maybe it would help you. These problems are almost entirely unrelated to one another.

Re: A.I.

You probably could build an AI* that could do that conversion, but you have two challenges that have prevented it before and will again:

1. This has been around for a long time: readability and maintainability. I have a program here which was originally written in assembly, I think Motorola 68K assembly. That wasn't very useful, and it was not attached to a lot of interfaces, so it wasn't hard to translate to C. So now, I don't need to try to virtualize something to make it run. However, I still don't really understand what's in here and can't modify it. I have C code that produces the same results that the original one did, but modifications means drilling down to understand what every part was for, and that is tricky because I had only machine code to start with. Every time automatic translations happen, readability tends to get lost. Even if the translation is perfect, it can be difficult to impossible to make modifications to it later.

2. Accuracy. This is a bigger problem with modern AIs, but it's always been a challenge. Most software does not have complete, mathematically proven test cases, where if the tests all pass then we're absolutely certain that nothing is wrong in this code. Mostly, we have basic test cases, where if any of them fail then something is very wrong with the program. Often, we don't even have too many of those. This means that when you're translating code, you can't know for sure whether it's even accurate to what the original program would have done unless you run them in parallel forever and set off an alarm if they ever disagree. Modern AI is very likely to add bugs when it translates. Admittedly, most of those are likely to be obvious, ranging from its output simply not compiling to the program crashing obviously, but when it looks like neither of those has happened, that does not prove that the attempt was successful.

* AI: depending on what you intend this to mean. We have explicit rules-based programming language translators around, but they generally don't qualify. I assume you mean an coding-focused LLM, which often make mistakes in their output, especially when the job is big. Translating a large program, and it would have to be large or you could let a human translate it, with insufficient documentation is very likely to be too large a task for such models to complete correctly.

Re: Down vote?

I realize we're getting close to having more people who didn't downvote trying to explain them than there were actually downvotes, but I do have a few comments that could explain part of it:

"If people are using pointer arithmetic, it's for a reason. And Rust won't be able to do a damn thing differently or any more securely about that."

This is an unproven assertion. There may be a reason that could be done better or equally well, or it might be required. It all depends on why that is in there semantically. That level of generalization is as wrong as saying that everything in C should just be rewritten in Rust because it all works.

"You cannot write safe code that interacts with devices, computer buses or the majority of hardware directly. It's just not possible.": It depends what the hardware is, but that's not the case for all of it. Plenty of devices can be used without having to share raw memory as they imply.

"Until we change all the BIOS/UEFI, PCIe etc. specifications to allow arbitrary device detection and communication at bus speed using only, say, JSON or XML or similar languages, you're still going to need to write those bits in C (as they have to make assumptions about what the underlying bits mean and directly access memory)"

Not all the time. Direct memory access is necessary for only some hardware, and it's perfectly possible to handle nonstandard protocols in a memory safe way. It's not as elegant as having everything in a standard form, but it suggests that some classes of simplification are not possible when they are not only possible but sometimes have been deliberately written that way. Even when you do have to do something in a memory-unsafe way, C is not the only option for doing it, though it's often the most convenient.

Re: Why?

I think the theory is that you convert to Rust, and then all the new modifications must be done in Rust so it will be more difficult for people to introduce that kind of bug because Rust will block them. Or you could add those checks into a compiler and refuse to merge something unless your C compiler accepted it. Either way, while there are cases where it would be theoretically useful, I doubt there will be all that much adoption. Among other things, I wonder how much readability is lost with conversion from limited C to Rust, because every other conversion I've done has involved some weird additional syntax which my brain takes longer to parse.

Intel debuts laptop silicon that doesn't qualify for Microsoft's 'Copilot+ PC' badge

Re: So ?

No, because an NPU is not required at all for Windows 11. Right now, it means that they cannot use a marketing badge that nobody cares about. Maybe later, there will be some local AI feature that they won't be able to run, but so far, there isn't even that.

Even at $200/mo, Altman admits ChatGPT Pro struggles to turn a profit

Re: People use it more than we expected?

But you do know that whoever is willing to pay $200 for a plus version is going to use the plus version a lot. It would have made a lot of sense if OpenAI had somehow deluded themselves into thinking that tons of people were going to buy their plan, but not many did, so they got little money for their trouble. The thing those payments are for cost quite a bit to create, after all. What makes a lot less sense is why they're surprised that the people who bought this are using it, and using it so much that they're losing money on that subscription. That's a clear indicator that they did not think through what it cost to deliver the service they were charging for, which is a much more basic mistake.

Re: People use it more than we expected?

He evidently expected that this would be the kind of product where people buy it but don't use it, so you can afford to serve the higher users from the savings on the low users. Like unlimited storage packages, where many users buy them and store 200 GB in them, so you can afford a few 30 TB users. Altman seems unaware that someone who is willing to pay $200 per month probably expected to use the services kind of a lot, and anyone who didn't use them too often saw that pricing and left. Sorry Sam, that price is out of the range for people who might use this and are willing to pay a bit to have the option available.

Amazon worker – struck and shot in New Orleans terror attack – initially denied time off

Re: Wrong type of leave…

You're right, and if something that severe happened, I wouldn't be trying to fill out the form. I'd send my manager an email and tell them to fill out the form, see you when I'm done with the medical stuff. However, I've tried to fill in the forms myself before, and the reason that I've mostly done them correctly is that I take a lot longer to do them the first time I try because I'm being very careful with what I put in each box. That means that, if this worker decided they would fill in the form themselves, I wouldn't be surprised if they made a mistake. They had a better option, making the employer do the paperwork, but that doesn't mean they used it. For that reason, I can't jump to either conclusion about whether Amazon is lying or just unhelpful.

Re: Wrong type of leave…

I'd buy it. Many places I've worked have worker management products with extremely clear forms consisting of about thirty boxes with short labels and no instructions about what is supposed to go in any of them. Some of them need things that the software should be able to but didn't pull from my existing records. Others need codes pulled from a document that's somewhere in SharePoint or Google Drive depending on whether this is a Microsoft or Google shop. Some actually contain data about my request. Even more are just there and can be safely left blank. It always seems easy to fill in such a form incorrectly, especially the first time I've done it. That doesn't mean Amazon wouldn't or didn't mistreat the employee here. Something plausible is also a good cover story if you were at fault.

How a good business deal made us underestimate BASIC

Re: Replace "Beginner" with "casual programmer" and I am 100% on board

When you're collecting data, how do you keep track of it? Do you store it in a location with a name, where you don't store unrelated data? Congratulations, you know as much about files as you'll need while writing software. Unless you're writing a filesystem, kernel module, or something that's trying to deal with disks, all you need to know is that files have names and contain bytes and directories have names and contain files. That last bit is not technically true, but you don't have to care. I'm sure you know both of those things already.

There are a lot of languages that are simpler for the novice than the ones you listed, but they all have limitations, which is why more general and powerful ones as you have listed are popular. Nothing requires you to use them. I would also encourage you to find the many tutorials that are out there for all the languages you name, because we really aren't trying to keep you away from knowing how to write them. There will be many that describe how Python syntax handles dictionaries which could help explain it*. Having a data structure like that is necessary for most programs, so having one provided for you is generally helpful. If you got a modern BASIC, then you'd have to build that yourself, and I'm guessing you'd find that a little trickier than learning the Python syntax for them.

* The simple version is that you name your dictionary and put your key in brackets. If your dictionary is d and your key is "hello", you can set a value with d["hello"] = "world" and retrieve it with d["hello"]. There are other options, but that version is all you need to use dictionaries at the start. Or, if these aren't useful, you could just not use any dictionaries. Python is a lot like BASIC if you just don't use the structures you dislike.

Re: I did comment the other day....

My argument did not rely on DOS being the epitome of anything. It applies equally well to any of the other operating systems that had no security in their design. A lot of the operating systems of the time, especially on the home computers that this article and the comments are talking about, were exactly the same. Mac OS did it. OS/2 did it. RiscOS, BeOS, AmigaOS, all of them did it. This means that, when comparing any of those systems to modern ones, they'll always come up short when comparing their security chops because zero is an easy number to beat.

But your point is correct: not every operating system was like that. Some of them did implement inter-process security. They did not do it perfectly. Some of the places where they had vulnerabilities were due to things that cause modern ones: coding mistakes, resource limits, or just making it easy for admins to mess up. Others were a result of the restrictive design, which is less of a factor with modern operating systems. Their simple design did prevent some classes of vulnerability, but not because the design handled it, but because the features that are necessary to have the vulnerability weren't available. It's similar to saying that my house with no windows is more secure than your house with windows because you can't throw a rock through my nonexistent windows. That is true as far as it goes, but by being able to look through your windows, you are more able to see and respond to possible threats. A system with no networking wasn't vulnerable to attacks over that connection, but it's not because their code was any better.

Re: pot?

"So whats the result if no files, going with some sort of uber-Registry containing it all?"

That's one of the several things I don't get. As far as I can tell, it's objects held by software in their indexed, ready to edit form. What happens when you want to use those objects in other software, I don't know. I would guess that you're supposed to export it from the software into some format that you can import into another piece, and I'm not sure where the data is supposed to be while you do that, but I'm now trying to figure out what you do to promote data interchange when the simplest mechanism is explicitly denied. I think it's up to Liam to explain what you would do where files aren't available, and so far, I've found that part either vague or missing entirely in any proposal he has posted.

Re: English is one of the easiest human languages

My guess is that it will still be English for a long time. The people designing these things have seen lots of examples, and those examples were in English. If they're working internationally, the lingua franca they're going to use is English. For example, the Loongson architecture, which should theoretically be easy to do in Chinese because it's only developed there, still uses English-style mnemonics. You could argue that they copied a lot of that from MIPS, and since MIPS used English maybe that explains it, but I bet it was considered easier to use Latin letters to avoid anything weird with encoding of Chinese characters, and if you're going to use Latin letters and you speak English, why not use ones that make sense there?

Re: A novice does not know the difference between RAM and disk, and they should not have to

And in what form should your stuff be returned to you? Because you can insulate yourself from files whenever you want but before you can, you have to decide on a format. Have a database server and you never have to manually read or write to the disk. You don't have to worry about endianness of your numbers or packing your data. But that only works if you've agreed to use the database's formats instead. You can go one level up and never serialize any of your types because a library is doing that for you, but you still need to limit yourself to things that library understands. You don't have to worry about any of those things, but you can stop worrying only after you've decided how they will be handled for you. Trying to opt out of handling them at all is laziness, the bad kind, and works the same way (I.E. badly) as people who insist that you just solve the problem without being clear on what the problem is.

Re: You had me...

I think we understand that it worked that way. I can even agree that, at the time, that made sense with how to do it. Most of the reasons no longer apply. For one thing, we have better error messages. If they typed "pront" in at the terminal, it still wasn't going to work, but now we'll write out the part that wasn't recognized so they aren't as confused. Whether that happens when they run the file or when they make a typo at the terminal, they're still going to have to learn that the machine won't fix your mistakes for you.

In many ways, all the things you praise about BASIC are maintained by Python, possibly one reason why people keep using that as an introductory language. Sit down at a Python REPL and you can also issue one-line commands and see results. Instead of line numbers, you write your blocks:

for i in [1,2,3]:

print("Hello")

The main difference is that you don't have to re-enter everything because we've decided storing your work to be edited at will is something you're too stupid to have right now. That made sense when there wasn't nonvolatile storage except for a tape or when the editor couldn't be used freely because you had to wait for lines to be printed one by one onto paper. Neither apply anymore, and any student will have used free editing into files many times before they write their first code. The argument that we should go back to those things sounds like someone who thinks the way they did it must be the best way in the world, in which case I have to ask them why they weren't using older methods that real programmers did in the 1950s and 1960s. Let me guess, those were superseded by better technology, but technology stopped getting better when you touched it.

Re: English is one of the easiest human languages

"I also think that agrees with doublelayer's view (and others'), reading between his lines (my bad if wrong) that English is in part, widely spoken for the same reason - its ease of use."

Sorry, it seems we did misunderstand one another a bit. I don't think English is used for its simplicity. I think English was used in BASIC because it was written by people in the US and UK to be sold to people in the US and UK. If there had been a much more straightforward language, it still wouldn't have been used because English is what they spoke. If people were building the early microcomputers on which BASIC would be used in Argentina and planning to sell there, then the language used would be Spanish and people would have lived with it.

English does have several simpler grammatical structures that make it easier to use and understand than other languages. I agree that simplicity is one of the reasons that English is popular in the world today, though to some extent I think the simplicity is the result rather than the cause of that global success. However, I don't think that explains most of it. Most of the popular languages are popular regardless of their complexity. English, Spanish, Arabic, and French are four out of the top six languages by number of speakers because those are languages that spread all over the place and were held there for long enough that many had to learn it. Chinese and Hindi are to some extent an exception, but they had periods of forcing others to learn it as well. Simpler languages failed to achieve success because they had fewer people forcing others to speak the language, fewer incentives for learning it voluntarily, and fewer people expanding it to be more versatile.

Re: pot?

The problem is that we've seen what happens when you blur away the existence of files. You get the smartphone. It's your data, you know. No need to wonder about where it is, Google's on it, you don't have access to the raw data anyway, the apps know where to find it.

And that is wonderful, assuming that you only ever want to do the thing the app gives you controls for. If you want to have some of those files on portable storage because your phone doesn't have enough internal memory, but the app didn't give you that button? Sorry, you can't do it. Do you want to back up your data and restore it to a different device? You could copy all the files and transfer them over, if we let you see them, but since you can't, you only have a one-size backup method which might work or it might simply restore the app, helpfully back to factory settings except it remembers that you asked for dark mode. Maybe the app is doing something wrong? On a computer, I could edit a configuration file to make it do something differently. But even though the same configuration file exists on the phone, I'm denied access to it, so I'm out of luck. This is the kind of stuff that breaks when you try to pretend that files are unimportant.

A lot of the time, you don't have to look at or worry about internal files, but you really do want to have access to yours and having the internal ones is always an option that's useful to have around in case you find a reason to want it. It's necessary if you're going to write good, structured code. Partially it's because it makes organization of something complex more feasible, and partially, it's because portability and reproducability is critical here. Programming involves multiple layers of understanding how the machine works at one level, ripping that away and learning what's under that, then putting back the levels until you have the level of abstraction relevant to your task. Hiding something as simple as where your data is under a layer of vagueness is not helpful.

Re: English is one of the easiest human languages

Could you explain what you wrote, though? In addition to all the valid criticisms of English, it seems to be missing the point that English wasn't used for the sake of simplicity, whether it has it or not. It was used because the people who wrote the language had English as a main language and were planning to have the software used by people who had English as a main language. Soviet programming languages used Russian as their main language, not because Russian is simple (anyone who tries to spell things will find that it's not, not that English should get many orthography points either), but because everyone involved spoke Russian, either because they were Russian or because they had to learn it in order to get to the places where the computers were available.

When dealing with a language like BASIC, language choice is actually quite unimportant. The small number of keywords means that you could use a language you don't speak because memorizing the keywords won't take very long. This is, for example, why all the words commonly used in music are Italian. You don't need to learn to speak Italian to memorize maybe fifty of them, and that's only if you study it for a few years. Modern languages, with lots more library functions, do need to pick a language that the users understand, and English was chosen, again not for simplicity's sake, but because most of the people using and making it already spoke it to some extent and no other language met those criteria.

Re: I did comment the other day....

"someone could write a modern (internet-capable, web-capable) BASIC for the Pi."

Of course someone could. You can make a language with any set of capabilities you want. The question is what you put in it and how you make certain things convenient and powerful. If it's not at least one, the language is doomed, and some languages get close enough to both that they'll get used unless yours does too. BASIC's structure doesn't lend itself well to scaling the number of options, because you either end up creating hundreds of builtin commands or you try to make all your functionality as a library and force it to work in the small subset of commands you're willing to include. That means that, if you want to be able to send TCP and UDP and ICMP packets, you're going to need to plan out a lot of syntax. That's not the biggest problem though. The biggest problem is that you likely also want to receive those, which means you're most likely to build a network stack in a different language that can handle this better and just abstract it out for the BASIC program. That works for students, but it's very different from the BASIC that let you drop to assembly when you wanted.

"The complexity of modern systems is what makes them insecure and takes away their resilience."

I disagree. That sometimes happens, but the simplicity of early systems often also made them insecure. Complexity means that interprocess communication mechanisms can have vulnerabilities that people didn't catch before they were put into production, but before we had those, inter-process communication was unlimited, meaning that those vulnerabilities couldn't exist because everyone was allowed to do that. Before that, inter-process communication was impossible because multiple processes was impossible. So did that make us more secure? No, it did not, because it meant that any process running could do anything it wanted to any piece of data. For example, consider the ubiquitous DOS malware that copied itself onto boot sectors and autoexec files, which could not be prevented because there was no per-file security mechanisms and no monitoring while the process ran. Complexity didn't just add more features, although it definitely did that. It also made the structures necessary to have security in a world where not every line is trusted.

Re: pot?

This is not the first time that Liam has argued that files are harmful. When he wrote his eulogy for Optane, he was unhappy with its death because he hoped it would kill the file system by uniting disk and RAM, a similar approach to what he's said here. I didn't understand it then and I don't understand it now.

Files are useful. They let you store things much more conveniently than an internal structure that something else writes to storage for you. You can find it, copy it, duplicate it, transfer it onto disks or over networks, it's the ultimate form of portability. We only managed without them because resource limitations were so restrictive, but as soon as the machines made storing separate data files feasible, people did so. Once you have separate data files, having code as data was useful in so many ways. Related to this article, early students of programming get two additional benefits from files. The first is that they can review and edit their work, even if their first attempt crashed the computer. They don't need to write down what they're typing so they can see what it was later. The second is that they can have multi-file programs, or in other words they can have a piece of code provided to them which they don't have to type in from scratch. Want to know how it did what it did? Open that file. Want to use it, just type in an import statement as supported by your language of choice, BASIC or otherwise. No need to renumber something or manually type in that code.

Re: GOTO

There is a difference between writing goto in a language where you have better tools and writing it when that is your only tool. There are lots of things in assembly that we insulate ourselves from because it was causing preventable problems. When Dijkstra wrote his famous criticism, he was talking about people who had better tools and he wanted them to use those. He was not insisting that someone write assembly without any jumps.

How the OS/2 flop went on to shape modern software

Re: Developers! Developers! Developers!

"People said the same about WINE 25 years ago. If Linux could run Windows apps there'd be no market for native ones."

And I think those people were right. For two separate reasons, that didn't end up having much of an effect:

1. Not that many people used Linux desktops, so most people didn't bother to make software to work with it, either natively or through Wine. It didn't look like there were a lot of Wine applications because there weren't, but that wasn't because people chose to make native ones. They just ignored Linux users. They mostly still do.

2. The reason why Microsoft writes Linux applications: Wine was never quite good enough for companies to sell things that required it. Sure, some things ran perfectly, but many others did not. That sometimes changed. A company wanting to make a Linux variant might try making a customized version of Wine that ran their software well and then wrapped their application with that version, but that was a much more substantial effort than trusting that anyone who wanted their software enough would do it for them. Had Wine been perfect and unchanging, it probably would have been used. A lot of the software that Microsoft makes for Linux, for example, uses cross-platform runtimes like .NET, admittedly they wrote a lot of the Linux support into that, and Electron, so the effort required to have one is much lower.

Most of the native software available for Linux was written for Linux first. Most of the native software for Windows was written for Windows first. A lot of the time, the authors don't bother branching out if changing the compiler options doesn't do all the work.

Re: I remember reading Letwin's post

I didn't use OS/2, so a couple of those don't sound as daft to me as they do to you.

"Why weren't long filenames truncated? DOS & Windows apps couldn't see anything with longer than 8.3. That was just daft."

Because that was a limitation that is better when you no longer have it. Forcing every limitation on something in the name of compatibility probably wouldn't have helped either, because then it would have been identical to Windows and Windows, being cheaper, would also have won. When the differences make OS/2 better, it makes sense to keep them.

"Why the hell was dragging only done with the right mouse button? What conceivable benefit did that have?"

This one isn't a good thing to me, but presumably the conceivable benefit is that, at some future point, a long press on the left button could be assigned another meaning, making the mouse more capable. Of course, if they never assigned that, it doesn't end up being very useful, but that's the primary reason I can think of for doing it that way, and it has precedent because that's why there were three buttons on the mouse instead of one in the first place.

HMD Fusion: A budget repairable smartphone with modular flair

Re: Pine phone

The European store is attempting to be a more retail experience, with additional post-sales support*, whereas the global one is selling at lower prices with no support included. You can get a more thorough explanation in the blog post when they announced these two stores. You can buy at the lower price from the global store in Europe, though.

* I'm not sure to what extent the support will help with software issues. Pine64 is well-known for providing hardware and letting you figure out software. Consider that before buying their products.

Apple auto-opts everyone into having their photos analyzed by AI for landmarks

For actual creepiness, they're not that different, that is if you believe Apple's statements about how they've worked to make this feature respect privacy. I actually do believe them on this one. However, the important distinction and the reason why there are different reactions is that the facial recognition thing runs on your device and landmark detection doesn't. Landmark detection requires that parts of your images be copied to a server, and they chose to do that copying without asking. Even though I don't think they intended to or are actually doing anything on that server that I am worried about, it isn't something they should be doing without asking the users first. Doing something similar on the device is potentially unwanted, but it doesn't involve copying off data, so it can be on by default.

Honey co-founder's Pie Adblock called out for copying GPL'd uBlock Origin files

I don't think they will face any penalties. While theoretically they could be sued for not releasing or acknowledging, and the developers of the code concerned could try to get punitive damages assessed, that is a lot of work that they will probably not go to. As for why they didn't just release the code as they were supposed to in the first place, I imagine it's a combination of laziness and not wanting people to point out that they didn't write a substantial part of their plugin. As with many such cases, they didn't get to hide that and probably would have gotten less attention if they published that and hid it, but the negative attention probably wouldn't harm them. While they are in the wrong, I wouldn't expect anything interesting to happen as a result.

With 10 months of support remaining, Windows 10 still dominates

Re: Enshittification...

Except, username pun aside, most of the common complaints are not about extracting money. The taskbar isn't as configurable now? We could speculate why they did that, but it's pretty clear that it wasn't about money. It would only qualify if they later charged you to let you have the previous behavior back. Either they somehow thought that the 11 behavior is better or they thought it would take more work to maintain it and they could get away with not doing that work. The same goes for any of the AI features that nobody wants minus those few where you have a finite amount of credits. Those that are merely included are there because someone at Microsoft has convinced themselves that people want that. In my case, and likely everyone else who has posted so far, they are wrong, but they aren't extracting money for those features.

I think the hardware requirements are a much more convincing one. I make this distinction, not to defend Microsoft, but to keep around a useful term. We don't need a special word for things getting worse. People have complained about that forever and people have complained about Windows getting worse with subsequent releases at least from the switch from 3.11 to 95, probably from version 1 to 2 but I haven't seen it. A special term makes more sense when referring to the process that Cory Doctorow described when he made up the term, for example as described here. If the term is used for any change we don't like, then we'll have to use long phrases to get this more specific meaning in there, and that weakens the entire point the term exists for. You can't call attention to something well if it constantly has to be explained every time it's relevant to a discussion.

Re: Error bars

On a similar topic, I wonder whether it might be premature to start analyzing exactly how Windows 11 lost in market share when it was one month and a 0.82% difference. I have a theory: business machines stopped being used as much in December because some people were away, but personal machines didn't have as big a drop. If business machines are slightly more likely to be running Windows 11, and I think that is likely*, that could easily give you that result. If it falls consistently for the next three months, then we can consider it more reliable.

* Businesses replace their computers more regularly, meaning that some of them are likely to have gotten ones with Windows 11 preinstalled. Some of them may have a Windows 10 image they apply to every machine, but those are also the types of companies that would have eventually built a Windows 11 image they could use. Home users tend to have longer and more erratic replacement cycles and often users more willing to refuse updates over and over. I know several people who consistently refuse to install updates, even small security-only ones, and some of them are willing to press the "remind me later" button every day even if there is a "don't remind me" button they could have pressed if they looked for twenty more seconds. Therefore, I think it is very likely that a higher proportion of home Windows machines are Windows 10 ones than business machines.

Re: Enshittification...

I don't hate the term, nor have I expressed criticism of it before. However, I do find that it is being used more generally than it was originally defined. It is intended to refer to specific, long-lasting policy decisions, so when people use it for "I think it got worse", it tends to lose some meaning.

If it got worse because it is a way to extract more money from you or someone else (advertisers, mostly) because that is the way the people running the business have decided to change the business model, that is what "enshittification" was supposed to mean. If it got worse because they have a new UI designer who is not good at their job, that is something else. Even if that UI designer hired more of them and they're all messing up what once made total sense, and they had some programmer friends who don't understand why you don't catch and ignore exceptions unless you have a specific reason, that's not what the term referred to.

Can 4G feature phones rise again on the back of QVGA, thin clients, and remote browsers?

Re: Linux

Invent a way to do that and we'll tell you what's wrong with it. So far, the attempts to do that tend to require you to have a runtime installed, and that was the thing that required a different version for your operating system. For instance, you have something that'll run in a JVM, which means you already have to have one of those, you have JVM overhead*, and you may have JVM version compatibility issues. Or you use the cross-platform .NET versions, which are actually pretty good nowadays, but for a while the non-Windows versions were very restricted or broken and you still get version compatibility issues.

If you have a great idea for how we fix that, it could be really successful. It's not a simple problem to solve though.

Boffins ponder paltry brain data rate of 10 bits per second

Re: The amount of nonsense in this paper is remarkable

"If we go back to the LLM example, know any 20W rated computers (hence consuming 20Wh at full throttle) that will be competitive?"

I can find you a computer that can run in 20 W which can run an LLM, sufficiently quantized, at an output token rate that matches slow human speech. The output you get will be only slightly more crap than most other LLMs. I don't think that's any better comparison to a brain, though. I've known people who can make up plausible fiction on the spur of the moment and keep track of everything, and those who can't string the truth together well if they're asked to do it quickly.

I think whatever we choose, there's not going to be a good parallel to what a brain does. Our limited understanding of how the brain does what it does is part of the problem, and assumptions about it is how we get papers like this that oversimplify things to the point that we can already realize how improbable it is without trying to duplicate their observations.

On the topic of typing gibberish, I'm surprised that you could type it as quickly as a language you speak. I certainly could not, though I type quickly in multiple languages. I somehow doubt that their research participants had been trained in the way you were, though, so I still think factors other than letter distribution were strong factors explaining the speed differences.

Re: The amount of nonsense in this paper is remarkable

I think you're mostly correct, but a few points:

"Someone will have to tell these people that keyboards have a certain lay out for a reason and that subsequently it's not physically possible to type random sequences as fast as sequences in the targeted language of the layout."

I don't think this is why that difference exists. Most keyboard layouts have had a little thought put into them, but yet Dvorak, for all its adherents, doesn't actually speed typing up compared to the less thought-out QWERTY where avoiding lever jams was at least part of the consideration. I think the main differences is that I have a lot of muscle memory. When I want to type "the", I can call on a long history of typing those letters in that order without needing any conscious thought to do it correctly every time. If I have to type "ymx", I don't have that memory. Also, it is common to look at the screen while typing if only by reflex to verify that you have typed things correctly. That's easy when you're copying words, because you can remember the words for a short time. Let's see, I was typing about types of apples, the words say "green-skinned apples often used in", sounds good. If I have to copy random letters from something I'm reading, looking back and forth from one window to another is challenging because the random letters have not been memorized and I'm either comparing them manually, which is slow, or flicking back and forth for by reflex without being able to use that reflex properly, which is going to cause delays as the brain attempts to do something that it cannot do. If we created a keyboard layout by dumping all the letters into the space for letter keys at random, but we then had people only use that layout for years, I think they would be similarly fast while typing languages they know and similarly slower when typing random letters.

"The estimate for the human brain is that it uses 20W a day or on average less than 1W an hour."

Watts are already energy over time; you can't divide them like that. The brain tends to consume about 1.7 megajoules of energy in a day, meaning its average power consumption is 20 W whether it's per day or per second. Which means that you can in fact get a computer that operates with less power, and quite easily. It will be a lot more power-efficient than the brain when solving mathematical problems, but most of the tasks we want to perform daily can't be done at all or anywhere nearly as efficiently on such a machine.

Will 2025 be the year satellite-to-smartphone services truly take off?

Re: Multiple financial risks.

Ground stations shouldn't be as big a concern in many countries, but that assumes that Starlink or any other company in a similar position wanted to help circumvent censorship. So far, they don't unless Musk personally disapproves of the censorship. The problem is that, while they can't easily block the service, they can block local payment to the service. They would have to provide the service for free or only to those who set up their own payments from another country, and it's more profitable to accept more payments and comply with whatever rules the host government has.

Re: Multiple financial risks.

Brazil banned Starlink because Starlink refused to enforce a government order on X. While it's true that the first order they had was against X, their ban on Starlink was directly linked to Starlink's action, not just because they were owned by the same person. Many other countries might be less concerned about that, though.

Re: Battery

That may be, and it's my theory as well, but they keep saying that they're going to offer data services. That could be Musk-style exaggeration which the mobile companies are willing to amplify, or they might think they can actually do it. Either way, I expect that the experience should they try is going to be worse than perspective users imagine. Either the speed will be much lower and people will find that the data isn't very useful, or the speed will somehow be high enough to use and you will quickly add up usage charges while draining your phone's charge while using it. It's up to them to prove these both wrong.

Re: Battery

I think this will be a nonissue for long enough that it eventually becomes a big problem. Right now, when the only thing you can do is send a message, people won't use it very often. They will use it to either send information to emergency services, where they are really not concerned about battery life, or something important enough to fiddle with things to get it out, where they are similarly not too worried. That means nobody will complain about the battery life problem for a while until the day, assuming it comes sometime, when they are able to use data connections from these satellites. When people start to browse a website, get online navigation instructions, or watch a video over these services, I expect battery life is going to start being one of their largest concerns, but it will be a while before anyone lets someone do that. There's a chance that technical limitations will prevent that working well at all, and there's a chance that they choose to charge so much for bandwidth that few ever try it, so we might not even get there.

Eight things that should not have happened last year, but did

Re: To be fair...

No, they are not suggesting that. They are suggesting that it does not, by design, send the collected data to Microsoft. Surely you can see how those things are not the same. At one level, you might just want to know whether negative things can happen as a result of Recall: absolutely yes, probably lots of them. At another level, you might want to know which bad things are most likely and how they could happen, and if you jump to the conclusion that the bad thing is that the data is sent to Microsoft's servers for their perusal, you've named the wrong bad thing.

Re: We all know that 'AI' is not really artificial intelligence.

All that quote suggests is that Steven Pinker doesn't understand a lot of things. I'll give him the order to obey orders thing; understanding that takes more knowledge about how the systems are very imperfect. The not harming humans thing, though, should be eminently simple to understand, at least in the case of making an order that could accidentally harm humans and preferring that the AI think through and reject the command rather than carrying it out.

The job title makes this a lot worse, though. A cognitive psychologist should understand that humans often do things outside of purely rational logic that they would prefer not to do. All sorts of negative behavior is done for irrational or indirect reasons which makes it harder to stop than if it was a simple mistake. Many of the rules we attach to AI, either real software or fictional or hypothetical versions, is designed to prevent that type of emergent behavior from causing big problems.

Screwed by the cloud: Hardware vendors looking for that raison d'refresh

I don't think they're suggesting that the newer machine is more secure by design, but that if you're running eight servers, then theoretically that's eight opportunities to mess up the security of one of them, so you should definitely be trying to have as few servers as possible, and wouldn't you know we happen to have one server that can be like a bunch of older servers in one box. Of course, if you are running a lot of different servers, then you're very likely deploying the same image to most or all of them, meaning you have exactly as many opportunities to do it wrong. At that level, you also probably deploy a bunch of VMs on that server and weird manual configuration is going to be on specific VMs, meaning the number and size of physical servers is irrelevant to security at that level. Customizing each server manually is more likely to happen if you only have a few of them anyway, in which case buying an expensive newer model may be overkill because you aren't using as much performance as you just bought.