Posts by doublelayer

AI girlfriend encouraged man to attempt crossbow assassination of Queen

Re: In Other News ...

If a religious person told you to commit a crime, being clergy probably won't be sufficient to prevent them being prosecuted, always assuming that it can be proven that they actually did recommend it, not that the perpetrator mangled something in order to justify actions they already wanted. If an actual deity was blamed, you are more than welcome to search out that deity, demand proof of incitement, and charge it in court. Should you succeed in finding the deity, I think the rest of the steps will probably be somewhat painful.

You've just spent $400 on a baby monitor. Now you need a subscription

Re: Confused

Structure the sentence to show where the clauses are:

Miku, you see, was acquired by Innovative Health Monitoring LLC last month. The new owner has wasted no time hitting customers (who are used to getting the service for free) with a subscription. The device will still work as a camera if you use it on the local network, but anything else is now behind a paywall.

Does this clear it up? Try removing the words in parentheses. Also, while I don't have one of these, other comments suggest that the other features were more than just access to video remotely. I don't know whether those monitoring functions were of use to anybody, but they were there and now cost extra.

Re: Someone else's computer

I don't think that would help. If I offered a service and wanted to collect 20p per month, I could ask people to pay £2.40 per year or even two or three years at once. People probably wouldn't mind. The reason that this isn't done is that subscriptions tend to cost quite a lot more than that. They're not asking for £6 per month because fees, but because that's £5.80 more than the other one and money is nice. I admit that fees are a problem for particularly small transactions, but most of those are not the size that subscription-sellers want to charge.

Re: Someone else's computer

"So, you're saying there's a whole bunch of people that believe that for one low price they can have something that requires outside services for free, forever?"

Some of the people who bought it may not understand that. Ask a person with no technical background how a video camera could stream information to their phone. I'm guessing you'll get an answer like "it's on the internet, my phone's on the internet, all they have to do is find each other and start sending data". We know that this is really true, but not what those cameras are likely to do because using a central server which connects the two together is easier for the users, but not every customer understands why that has ongoing costs, even though the costs for that are incredibly low. Similarly, the box says that this camera can analyze the audio and video to monitor certain activities, but how does the average customer know that it's doing that by sending the video to a server which runs that software? How can we know that ourselves without analyzing the traffic coming from it or reading the report of someone who has, since although it's more likely to happen that way, nothing prevents the manufacturer from including a more powerful SoC and running the analysis software locally?

People buy a product with the assumption that it can do what the box says it can. They're not trying to work out the implementation details, because if they were doing that, they could build one of the Raspberry Pi-based alternatives that multiple people have described. If it needs an ongoing payment, they expect the box to say that, the way that legitimate products that require one do (it's not rare to see "Requires subscription, free for the first year, £4.99 per month afterward"). This is really not the same as lifetime deals for internet, not that I've ever seen such a thing, since the requirement for ongoing services is not specified anywhere.

Re: Someone else's computer

At some point, we put extra restrictions on people who choose to sell their products to the market. We don't let just anyone build and sell electrical hardware, for example. There are some safety tests they have to pass beforehand and there is consumer protection law that applies after they've sold it. If we removed both requirements, we'd certainly get more and cheaper equipment, and some of it would be just fine. However, some of it would catch fire or otherwise harm the customers, and we've decided that regulation is needed.

A hobbyist making a product should still make a good product. I don't expect a regulator to check them for quality every time, but providing some indication that they're not scamming customers with what's effectively an early obsolescence plan is justified. I'd prefer if we could respond only after they do so, assessing a penalty on those companies that fail a reasonable test of product lifespan, but if that isn't feasible, then taking preventative measures might be justified. If I want to form a company to make and sell something, I already have various standards I have to follow. As long as those are limited to something that provides a real benefit to the public and enforced equally on all participants, it's something society decides.

Re: Someone else's computer

You can still do an in-app purchase for an update or, as at least two apps I've used did, simply stop updating [app] and releasing [app]Pro. In one of those cases, they gave existing users of [app] a discount on the purchase of [app]Pro. Those models are available, and people will sometimes buy updates that way. Of course, it won't be as popular as releasing updates for free. One of the reasons I prefer that is that I've had software which includes, as the update for which I'll now be paying again, a bug fix for something they screwed up in the first place, which never helps my confidence. Still, don't pretend like the option isn't there. Not everyone will take it and you will have to decide how you'll handle users who decided not to update.

Lenovo to offer Android PCs, starting with an all-in-one that can pack a Core i9

Re: The big question is...

My guess: both. I have seen some Android devices made for industrial purchases. They're not loaded with bloatware and locked down so they can't breath. They run a clean AOSP with a couple additions from the manufacturer which actually do something, and you can root them if you want and safely remove the extras the manufacturer added. And also, they get no updates of any kind. If you want to install this month's security patch, you should have all the access you need to install it once you've built the image from scratch because Android doesn't include the software to do this and the manufacturer can't be bothered. So I'm guessing this will look something like that, since the article suggests that most of these are intended as kiosks, not office computers.

I think you're being a bit extreme with all of those comparisons. While Windows NT has some roots in VMS and Mac OS some roots in BSD, including all its roots, you can't call either one of those a clone. Nor was it really fair for someone to call Linux a clone of Unix, but that was much more stated as a goal when Linux was created. For example, while NT borrowed some concepts that originally debuted in VMS, it didn't share code, it didn't get designed for compatibility with code written specifically for VMS, it didn't even get the same commands or calls. That's not a clone. Mac OS was a clone of NextStep at one layer, but otherwise its implementation had a lot of modification and NextStep wasn't really a clone of anything even though it had some Unix-like features. I'd say that most desktop OSes only go back to the 1990s, as the code written then was generally from the ground up.

FEMA to test emergency alert system US-wide today

Re: My only question is WHY?

True, but if someone fires a nuke at Manhattan, they probably don't plan to stop with that one. People in other cities might want to get to somewhere safe just in case another one is heading for, well not them because that's probably too late, but someone kind of near them. In an incoming attack, most people will be unable to figure out exactly where each weapon is designed to detonate, and even if they do all the calculations to get that data, there's little benefit in trying to quickly separate messages for each target rather than telling everyone that you might want to be in a safe place right now and if no nuke shows up near you in the next couple hours, that's great. If we get a full-scale nuclear war, the nukes won't be fired one by one. Tactical nuclear attack is a different story and would require different early warning systems. Let's just agree not to do either.

Re: My only question is WHY?

"I agree it is rather silly to warn us a nuclear strike, it isn't like there is time to get away or if the attack is big enough that there would be anywhere to get away to. But this was an idea from the "duck and cover" days!"

It's not that silly. If a nuclear weapon detonates close to you, nothing you can do will save you from it. However, even in a large nuclear war, there are lots of places where nuclear weapons are not going to be targeted, leaving people to deal with the consequences of weapons that went off far enough away that they aren't vaporized but close enough that they still have immediate problems from it. Those people can, in fact, get into a sheltered location where it is less likely that they will suffer the most dangerous effects the weapon has at the outer range of its effectiveness. Once they're there, they can also be alerted to stay inside for a while to let the initial levels of radiation decay. If they're unlucky and most of them would be, they're probably still going to take more radiation than we'd prefer when they leave, but if they went out immediately, they'd take an order of magnitude more. That's a lot of people who could theoretically be much better off if they have a few minutes of warning. Of course, it relies on people having somewhere to go when the warning sounds and being in the right frame of mind to go there rather than scream or try to go outside to find someone else, but that's on the citizens.

Re: My only question is WHY?

When they tested nukes, they were kind of hoping for that, which is why they did it in the desert or islands that weren't quite far enough away from islands on which other people were living. If someone fired nukes, they would be hoping for the opposite and would design for that, which is where an alert would have a greater benefit.

Re: My only question is WHY?

The only logical thing is a nuclear strike, where an announcement as soon as the attack is predicted, which is still a matter of minutes, would allow some people to get to a safer location. Of course, a lot of those would not actually go there and would start panicking, but that's theoretically a benefit. However, this national test isn't that different from testing in multiple regions one after another. A smaller region getting a local emergency report covers more situations, and if they're testing it, they could just do it all at once.

We were considering using this alert to trigger the secret mind control chips we installed, but we decided not to because they've been acting up. Some of the chips appear to have caused some glitches and made people start saying that they got them in vaccinations, but in reality we already implanted them years before and we've just been waiting until the communication systems were good enough to fully make use of them. At least the screaming about vaccines cleverly hides where the chips really came from, but it's still not great if our mind control conspiracy results in lunatics instead of pliable slaves. Also, we decided to broadcast an activation signal with the firmware update that fixes this in a way that doesn't cause phones to start shrieking. I don't know who decided that people should get a warning when the control starts up, but give us some credit, we shut that idea down fast. We're competent evil overlords, after all.

To any sarcasm-deficient readers, this was all completely true; I really do have mind control chips in everybody, vaccinated or not. Don't make me use it. Insert evil laughter here.

To any non-sarcasm-deficient readers, let's all do this. It gets boring having to state the obvious.

Cat accused of wiping US Veteran Affairs server info after jumping on keyboard

Re: I dunno

I have no problem believing it, because the confirmation dialog probably sets keyboard focus to the yes button when it opens and the space bar is really big and easy for the cat to hit by accident, pressing that button. I've inadvertently pressed space on a button I didn't want pressed with my fingers and I was using the computer at the time. A cat can manage the same.

I wonder if I've had the only cat who didn't step on keyboards. She didn't mind walking on my desk while I was working or even getting in the way, but she was always very careful not to step on the keyboard as she did it.

Big Brother is coming to a workplace near you, and the privacy regulator wants a word

Re: Interesting note

Before I start, I've never had this happen, never known someone who said it happened, and I kind of doubt that many companies would want to do it. However, if we just assume that a company is going to, your suggestion is not likely to prove anything about it. For example, you suggest that we prove that only the company had the information necessary to do this. The problem is that they don't. Depending on your country and personal finances, probably every bank or other financial business you've dealt with has had all that, places where you rented things may have had enough, and any subcontractor who checked about you would have had access to enough data. Any of those places could have stored data longer than they needed to and been hacked, and it's not even that illogical to assume that's what happened and they're scared of fines (if you live in a country with a regulation that could fine them for doing so).

If a company decides they want to harass you, they have lots of options. It wouldn't be very hard to decide to delete the emails reporting information about you right after sending them or to use a system that's easier to hide. The only way that this works easily is if they're completely incompetent on how to unnecessarily retaliate, but I'm guessing most of those won't bother trying, so if you're facing retaliation, they probably know at least some things about how to get away with it.

Re: "or offsite [...]outside work hours.

And what kind of misuse justifies what level of surveillance? This isn't about things like device management which can prevent you installing software that hasn't been approved. In many cases, it's not even more invasive software that's still designed for technical protection. Usually, we're talking about software that's supposed to check whether you're doing your job or not, but since it isn't smart enough to know that (your manager is, funnily enough), it uses various unreliable proxies like whether it thinks you're typing enough at the right times. What is the misuse you think needs protection, and how extreme can a surveillance measure get before you have a problem with it?

Kaluma squeezes JavaScript onto the Raspberry Pi Pico

Re: Why, $deity, why?!?!

"I have terrible difficultly understanding how JavaScript became the defacto standard in web browsers - it's a case of the lowest common denominator IMO."

Because it started small. HTML is a pretty weak language for the smallest of active tasks, and people doing web design in the 1990s didn't want every change to a page to require a round trip to a server over a slow connection. The first uses of JavaScript were to do relatively simple tasks to a page's content, and at that level, you don't need very much. You don't need execution speed or a lot of types or a coherent dependency system when most functions will be at most twenty lines and involve showing or hiding some blocks or putting dynamic text into them. Any language that includes all that stuff would probably get in the way, because you can easily write HTML in any text editor, so the last thing you wanted when writing a basic client-side script was to need a massive compiler and a big set of header files you'd have to have stored somewhere. Then it advanced slowly from there, with people using slightly more complex blocks.

It also won because the alternatives at the time and later were horrible. Most alternatives required a large blob to execute at all, a blob which frequently got updated. Sure, if you had the choice of writing a moderately complex client script in Java or JavaScript, you might prefer Java. However, if the question was helping users install the right version of Java so your thing could execute or forcing yourself to write two hundred lines of JS, you wrote the JS. A language like JS, I.E. without external dependencies, without a lot of base code that had to be incorporated, and without significant version pinning problems, probably would have done better. The other options did not check those boxes, so JS was adopted.

"Do you think they could fit a JIT into 264kB of RAM?"

No, but that wasn't their point. They weren't trying to say that it could have been done better but wasn't. They were saying that, even though some runtimes have been made super fast by spending a lot of money, it won't do users of this one any good because the result of that spending is not available to them.

By the way, while a JIT and then an interpreter for the remainder probably can't be usefully fit into that memory, you could precompile things to a lighter version, which would probably help over a straight source interpreter.

Re: Hmm.

From your description, I expect that some of them will try. My guess is that they'll get one of the boards which take an RP2040 and more flash chips so they have some more space to put files and then start experimenting with how much they can import. I believe this from a simple thing: your article says they've gotten Node and NPM on it, and someone will try to use it. You can hardly write anything in NPM without pulling in a useful library which provides some function you don't want to write, does it well, and comes with thirty seven dependencies that nobody is clear what they're for and why they're there. So yes, I do expect that people will try to run that on that system and that they'll either run out of RAM or end up choking it by repeatedly loading and unloading parts of an unnecessarily large dependency tree.

You can write something small in JS, and some users will. I just don't imagine that it will stay that way for everyone, just the people who make a functional result when they're done.

Russia to ban all VPNs – again – says senator

Re: Everyone's at it..

Probably your ISP's DNS server, as that's what this ban has been when talked about in the past, although that hasn't stopped the Russia adherents from pretending that's a governmental ban. If you're posting here, you should already know at least three ways of bypassing that and I'm not talking about VPNs, but changing the IP address to a different resolver which has other benefits anyway. As usual, my UK-based endpoint can resolve and load RT's site just fine.

For the moment, it sounds like they're limiting this to app stores that can install the VPN on mobile devices, not the network connection to endpoints altogether. I wouldn't expect that to make a dent, but I built my own, and probably most average Russians trying to avoid the censorship won't be doing that.

However, if they do decide to block connections, they may have more ability to do so than China. China wants business to happen, so they allow VPN traffic from ones they think are corporate and only go after ones that appear to be public with moderate success. Russia has cut itself off more than China has and may well decide that, since we in the west have mostly stopped employing Russian residents anyway, they might as well have all Russians work for Russian companies which can have VPN endpoints located inside Russia. It won't help their economy, but nearly nothing they've done in the past year and a half has and that hasn't stopped Putin any of those times.

$17k solid gold Apple Watch goes from Beyoncé's wrist to the obsolete list

Re: No doubt

Yes, they can replace the batteries, but I'm referring to the stresses that that might place on the unit compared to the smaller ones placed on a mechanical watch or one that uses very little power. Opening the case much more frequently to replace a battery that drains every day is going to add to the maintenance cost, and if you ever get a battery that swells, it could cause damage that's much harder to repair.

However, maybe it would be better to talk about the fact that that watch only has 2.4 GHz WiFi, and we will eventually get to a point where that is not as often used. I already didn't bother having a 2.4 GHz signal for my home network, so it wouldn't be able to connect if I had one, and I doubt I'm the only person to have done that. Of course, when it stops being able to connect to the internet unless you set up a connection method you otherwise wouldn't, it's not entirely useless then either, but it's a bit less functional. This is pretty common when it comes to devices whose main point is to be a computer.

Re: No doubt

People can melt that down to get at the 75% of it which is supposed to be gold, and I'm sure that if it wasn't, we'd have heard the story by now. What I don't know is how much gold there really is in the case, because they only said how much of the gold-colored metal is gold, not how much of that metal is on it. It could be a thin shell around the rest of the obsolete unit. Then again, we're talking about the people who spent $17k on a newly released product that wasn't expected to last for decades, so those people probably don't care enough to get full value from their junk.

Isn't this the point, though? You decide what features you want to have and select a device that meets those requirements, including any tradeoffs required to get there. For instance, my wrists are devoid of any kind of watch, because I've considered the options. I don't need a watch to tell me the time because my computer's right in front of me and it has a clock and my phone is nearby and it has one. The traditional time-only devices aren't useful to me. The list of features in smartwatches are also not that useful to me because, as before, my phone can do most or all of that and I don't mind getting it if I decide I want to talk to someone or I have a calculation that I don't want to do in my head.

I find the comments comparing smartwatches to time-only watches a bit confusing, because people appear to be comparing things only on where you keep it during use, not what it does. On the same basis, I should tell anyone who has a mobile phone that they're stupid because it gets much less battery life, even if it's one of those that last a week, than my solar-powered pocket calculator. Sure, you can call people with that one and my calculator only has four arithmetic functions and that percentage key for those who don't want to type /100, but they're similarly-sized devices that are both stored in the pocket, so they must be comparable and expected to work the same way for the same amount of time.

Re: No doubt

I'm guessing that anyone who bought a solid gold first generation product has enough money and obsession with Apple that they've already bought a newer one and don't care that this one has become worthless. I'm sure someone will buy it for the gold value, though I don't know how much gold they can have put in it without making it a bit too heavy. Either way, Apple isn't going to need to make them functional.

I have to disagree a bit about your comparison to a simple watch. These things aren't the same kind of product, so it's not surprising that one lasts much longer than another. A watch that only tells the time doesn't use as much power as one that has a WiFi radio and GPS receiver in it, so purely for battery reasons, the smartwatch will last a shorter time. Apple should have had more software support than they did, but I don't think anyone was expecting it to be working usefully in 2050 when WiFi will look very different and the things we used to use with the device no longer exist.

ASUS's Zenbook S 13 is light, fast, and immediately impressive

Re: Apple power plugs

I would have the same concern, so I got some magnetic USB-C power cables which I can connect to charge my computer. The magnetic tip protrudes a couple millimeters away from the case which isn't ideal, but otherwise they've been working pretty well. You have to be careful to find cables which will agree to pass the voltage you need, because there are some magnetic USB cables that are 5V only.

Apple power plugs

Your praise for Apple's power adapters confuses me. Yes, they have the option to simply plug the transformer into a socket without a cable going between the two, but this also means that the transformer can easily block other sockets or cause other problems. This gets worse as the laptop's power requirements increase, because they always make the transformer longer to deal with this. The one that only puts out 45 W can probably fit neatly in a few places, but the one that makes 85 W not so much. It can easily block nearby sockets, is too big to fit into recessed sockets, and can be heavy enough that you don't want to install it on vertical sockets in a few countries. If you have any of those problems, the solution is a cable included in the box (at least it used to be) that goes from the socket to the transformer, exactly what you didn't like about this one. Is a box in the middle of a cable really so big a problem? I doubt it's as massive or heavy as the ones for gaming laptops, so if it's as small as some of the ones I've seen, it makes relatively little difference.

X Social Media sues Twitter 2.0 over alphabet soup branding

When does a raw deal become your fault

From the article: "Yaccarino has really been given a raw deal."

I guess so, in the sense that her job is a pretty awful one and it's not likely to get better, but a lot of us knew that when she joined. She should have as well. Eventually, that stops really being a raw deal and starts being a bad decision to accept that deal. I don't think she was out of alternative options before she accepted this.

Google doubles minimum RAM and disk in 'Chromebook Plus' spec

Re: Weren't Chromebooks supposed to be low-end-use-as-terminal-only ?

They gave up on that a long time ago. It made them seem even more pointless to me, because now they're no longer cheaper* than a laptop that can do more things**. Then again, I was never going to buy one that could only do the Chrome OS things, so maybe it was inevitable that they'd realize that dumb terminals weren't very popular when you could have a smart one without much difference.

* There are cheap Chromebooks the same way there are cheap Windows laptops. You can run Windows 10 on a dual-core Celeron or Athlon with 4 GB of RAM and an eMMC, and it won't even be unusable, but you still don't want to to the extent that buying used is a better idea if you really have that restriction in budget.

** Yes, I know, Chromebooks have a Linux VM, just like every other OS in existence. I prefer when I can select my own versions of Linux and, if the need is there, Windows. I'm not particularly interested in testing a bootloader that might be unlockable when I can buy lots of computers whose bootloader unlock procedure is pushing the right function key when it turns on and selecting a different disk from the menu.

Musk's first year as Twitter's Dear Leader is nigh

No, PayPal the product already existed under the PayPal name. The company wasn't called that yet, but it wasn't called x.com either. Musk brought in the x.com name along with some cash because he threatened to overwhelm PayPal the product by spending a ton of money to get users away from it. When they got rid of him, they made the company name PayPal to match the product they offered already. Somewhat like Twitter, Musk found someone else's idea and slapped his pointless name on it.

Free software pioneer Richard Stallman is battling cancer

Re: An inspirational guy

Make sure that the -rf flag combination is incredibly important to catch out people who typo rms to rm so that it causes the maximum possible damage. Maybe -f could be "full output", while being careful that the default non-full output is useless. After all, what better tribute to our long history of CLI tools but making sure that some of the command decisions are either confusing or make accidents a bit too easy?

Re: +1

I don't know, but when posts of mine have a couple inexplicable downvotes, I tend to ignore them as either someone pressing the wrong button, someone annoyed at something I said elsewhere, or someone who simply didn't want to explain why they disagreed. I don't think it's the time when someone blindly downvotes all the posts as I can see a few that have none, but that's always a possibility as well. I tend to ignore downvotes unless there appears to be an unusual number of them, in which case I start to look for a reason someone objected to what I said as it will help me better understand our opinions.

Raspberry Pi 5 revealed, and it should satisfy your need for speed

Re: Would love to upgrade my RPi 400.

They could, but for a few reasons, it wouldn't be any better. I've covered some of them in the comment you replied to. For one thing, look at a Pi 400 teardown. It's not easy to take apart the 400 without breaking something, meaning manually swapping the board is not a process people will want to do too often. It's possible, but the last thing they need is to have a lot of complaints about the board breaking when someone tried it.

The second reason is that there's basically nothing that gets reused. I have a Framework laptop which has swappable motherboards. When you buy a new motherboard, you're keeping a battery, case, screen, audio system, ports, and quite a lot of stuff really. If you swap the motherboard of a 400, the only thing you're keeping from the old one is the keyboard, and it's a pretty basic, cheap keyboard, and a plastic shell which are really cheap to make. There's little advantage to the user in keeping those things when you could have a new keyboard and your old 400 still works.

The third reason is that the new Pi is more powerful. I could see them making a 500 with a faster storage facility using the new interfaces on the board, but if they did that with a 400 motherboard, there would be no space for the new hardware and possibly a heat problem. Redesigning it to attach up the new interfaces would also let them change the design somewhat. For example, one of the complaints about the 400 was that the USB 2.0 port was on the left side and some people wanted it on the right for a mouse used by right-handed users. If they made a new version, they could easily move that or even have two of them, and since they'd have to redesign just to add the things the Pi 5 can bring, it would let them find some new design options.

Re: Lost the plot

I'm aware of the Windows tax, and I tend to avoid paying it when I can. Still, that's not the same as locking down a computer, where they could have done a lot more things, probably quite successfully. I think Gates might have tried to do it had he remained in charge. He was more willing to pursue a monopoly and more aware of the technical requirements to do so. However, he didn't remain in charge and the locking down didn't happen. Microsoft had important roles in several areas of firmware standardization where they could have put locks. Those locks aren't there.

But maybe that's not for lack of trying, maybe Intel just wouldn't let them. Let's consider the hardware Microsoft makes. That's the area where they can do whatever they want to lock them down. It wouldn't be hard at all for them to make sure that a Surface would only ever run Windows. After all, they're writing the BIOS, firmware, everything. I've had a Surface running Linux, and the steps to do that were plug in USB drive and select it in the boot manager. Microsoft could have also been sneaky and made sure that drivers weren't supported under Linux if I did that. No, everything worked perfectly out of the box. Microsoft even developed some phones, first with Windows Phone and then with their two-screen Android things. We're talking about a market where nearly everybody is locking them to only boot the manufacturers' version of the OS, not even letting you change that without hacking. Did Microsoft gleefully adopt this strategy where it wouldn't be noticed among the ten other companies doing the same? No, both the Nokia phones and the Surface Duo have been unlocked and a frequent target of people hacking them to install other operating systems, from the people who managed to run Windows 11 on a Duo to people using them to test mobile Linux.

Microsoft doesn't do everything well and I have plenty of complaints about them. Painting them as a constantly evil force in defiance of years of experience does not help.

Re: Would love to upgrade my RPi 400.

"I wonder if they'll offer a drop-in motherboard upgrade for the 400?"

No, they won't. There is no part that can be reused. The only thing you could possibly reuse there is the keyboard, but the rest of the system's all in one piece. Even if they made a 500 (or whatever) mainboard, you'd probably have to cause some damage to the case to get it in, and the mainboard would cost about the same as one in a plastic shell with a whole new keyboard on top. This is probably a good thing, because if they do make a 500, they have a chance to improve on the design of the 400. At the very least, they'll have other interfaces that it would be useful to provide somehow but the 400 case has no provision to do.

Re: For my pumpkin soup...

It depends what experiments you want to run. If you're just interested in learning how to use Docker, you can do that on your computer without having to have much RAM in it. If you want to create distributed systems with it, then your laptop will not be great no matter how much memory it has* but Pis would work pretty well. If you want to run something that needs a lot of RAM, then unless it already supports using distributed nodes with less memory, the Pis won't be able to run it well and you'll want to get more RAM in a single, powerful device to run that.

* Yes, you can experiment with distributed systems by creating a bunch of VMs on one machine and treating them as nodes, which will work as a simulation, but it won't give you the real thing like a cluster of SBCs would.

It all depends on what your motivation is. It sounds like your motivation is "I want to be part of the Pi movement". That is not necessarily others' motivation for buying a Pi. They might just want a small computer for a certain task, and for that motivation, there are many non-Pi computers in existence which they'd consider. If they're not specifically attached to the Raspberry Pi's products as the only option and only consider things that they can run Linux on, it is much more logical to compare options for price, performance, power consumption, etc.

I'd also point out that "the Pi movement" is kind of vague. What counts here? Does someone else's SBC count? It's still ARM, Linux, open source, low-power*, easy prototyping, but it's not the Raspberry Pi's product. What if the Raspberry Pi company built something completely different? Would that be part of the movement. The term is so vague that it kind of sounds like the argument of someone who only ever buys Apple products and always will just because Apple made them, but I'm guessing that's not what you intend. I may not be a participant in the movement simply because I don't understand what it is.

* The power consumption of boards like this may also be the reason people are considering X86 boxes in the same category. Yes, the Raspberry Pi will be consuming less power than refurbished boxes, but not so much less that it changes how you would use it. If you include others' SBCs in there, they usually consume more power than the Pi does.

I didn't know that, and it's good to hear if your summary is correct in all cases, because it addresses a large sector where the Pi has been unsuitable. Specifically, the Pi has always run pretty badly when you need it to be powered by a battery. Power consumption at idle wasn't low enough to avoid quickly draining a battery, but the only lower power mode was to completely shut it down, meaning that resuming from that would require waiting through the whole boot process and you either needed to build in an RTC or be confident that you'd always have networking so you could have the clock working. This is why, for example, I never bought one of the Pi-into-a-laptop products; they'd all advertise three to four hours of battery life and deliver two. I did build several battery-powered Pi projects myself, and in order to get them to last long enough, I ended up having to put in a battery so heavy that the portability aspect was sacrificed.

Re: Lost the plot

Oh, that's why every Windows machine I've ever owned can be erased quite easily and have my choice of Linux installed on it? If Microsoft was going to lock down computers so I couldn't do that, they've had about three decades to try. The only times they got anywhere close were the Windows RT ARM devices (dead) and the current Windows on ARM devices (you can still install Linux without them blocking you, but the Linux probably won't work because ARM).

Mozilla's midlife crisis has taken it from web pioneer to Google's weird neighbor

Of course such things exist, and you don't even have to get that advanced when you can have one central CSS file which is linked in to every other page and use a variety of template pages that get pasted in. The trouble is that, with each change you make, someone has to learn more of a markup language that, while pretty simple for those of us who use significantly more complex languages, is still precise and takes some persistence to use properly. Consider the difference if someone, let's call her Alice, has built a website for someone else, yes it's Bob. Now, Alice is busy doing something else, and Bob wants to add a new announcement to his website. How does he do that?

If Alice has written it in hand-coded HTML, the instructions probably look like this:

Write your new article in HTML. Save it as its own page, then modify the home page file to link to it with a summary. Make sure to link in the CSS file on your new page so they look the same. Then run this script which will replace the template lines with your standard navigation, header, and footer content. Once finished, upload all the new files to the server and test.

If Alice used Word Press, the instructions look like this:

Log in to the backend. Click new page, write the content in the boxes. If you need links or anything like that, there are buttons to add them right next to those boxes. Check the box to mention this on the front page, and if you want, adjust the summary. Click preview, and if it all looks good, click submit.

Which would Bob, who does not know HTML yet because Alice was doing that, want to do this time? For that matter, if Alice isn't the only typical maintainer, which one does she want to do if she doesn't want to see someone else mangle the scripts she's written for this? In most cases, Alice wants to have a CMS that has a GUI editor, perhaps because she's gotten tired of Bob calling her every time he wants a new page on the site because he can't be bothered to learn HTML and CSS to do it himself, but he can manage to type in some boxes. However, Alice doesn't necessarily have time to hand-code a CMS for Bob, so he ends up with one of the existing ones with a bit of customization on it.

I'd like for this to be different, but it's not. I know this from experience. I have written many backend systems in more complex languages to get around the basic options of HTML, but I do that for my own systems or internal things for my job. The thought of having to do that for other clients as a web designer sounds like a job I would dislike very much.

Even this has a reason, and it's one I'm forced to accept and to notice in other areas of my work. I write software, not websites, but there are still plenty of cases where I have to write some code but it feels like I shouldn't. Usually, it's some form of data manipulation (take it out of this database, do a little branching to reorganize it, put it in a json file). This doesn't take a lot of mental effort, as I don't have to do anything particularly efficient, nor invent any new paradigms, nor know much about the system, basically no creative effort and little technical knowledge involved. Since I can and have done all of those things, it feels like someone who doesn't have the knowledge that I and my colleagues have should be able to do this simple bit. Unfortunately, it is usually not that easy.

The same is true of HTML. If you want to have a page with some text and some links, it feels weird to have to hire someone who knows how to make a page. Of course, one can just learn some HTML and make a functional page, but it won't make a visually appealing page, nor will it make later modifications easy when each page on your site has to be built from hand. To build a low-end CMS for your own site takes more knowledge of some backend language to construct the pages as well as the frontend HTML and CSS. I've done it several times, but I've used either PHP or Python (and once C, but I wouldn't recommend that). People don't want to learn that, nor do they want to hire someone to do something comparatively basic. Word Press, while not really the right tool for the job, lets someone design some kinds of site without having to learn to code properly. Yes, the resultant page's code will be ugly and the maintenance requirement is high, but they don't care about the former or know about the latter. From their perspective, their site looks good enough, they spent a lot less time than the coder would have, and if they need something bigger, they'll deal with it then.

I don't know, but I have heard lots of complaints when PDF support wasn't in browsers. One typical complaint was that there were a lot of PDFs online and they didn't like downloading them. The other one was typically targeted at Windows by Linux and Mac users who thought that an operating system that didn't ship with a PDF viewer was substandard. Oddly enough, when Microsoft added PDF viewing to Edge, they didn't change their minds; one might think they didn't like Windows for a different reason. So yes, people did ask for it, just not us although since I don't edit PDFs, a reader in the browser works well enough for me.

Re: Self-reinforcing

"Though it's a brave company willing to potentially lose 3 from every 100 customers."

Not brave at all. Some companies are really obsessed about making sure that every customer has a brilliant experience, sometimes to a level where their workers suffer from it. Most others don't bother with that, even if it means that some of the customers won't want to stay customers. A lot of tactics to save money can deal with a 3% drop if they save enough, and businesses take that gamble all the time, sometimes to their benefit and sometimes not. In addition, most companies that don't work on Firefox assume that the users will just use something else, and often they're correct. My family uses Firefox (I didn't make them or even ask them to, so it's not just us technical people), but if something doesn't work, they'll temporarily use Chrome rather than make a big deal about the problem. Of the Firefox users, what proportion will really refuse to use something that doesn't work well in it?

We can, but it won't help. Plain HTML lacks too many features that people like using, which is how we got scripts in the first place. HTML version updates come too slowly to entirely clear them out, although the update to HTML 5 helped a lot. For better or for worse, and often it is worse, users have come to expect that a webpage can update itself, not require you to refresh it every few minutes to retrieve updated information. They know you can make a table where you can click on a column to sort it without reloading the whole page. They know you can have an upvote button that doesn't reload this entire forum topic when you do it. And they know that you can put these things together and get, for example, a collaborative word processor that doesn't need any program installed, even if that word processor is going to be much more annoying than the local ones. This means that, if we tried to reverse the course, those users and the companies who make the products will be asking why they can't have those things anymore and we'll have to have a better answer than "because I hate JavaScript and it means browsers are harder to write", because they won't care about either of those completely true statements.

EFF urges Chrome users to get out of the Privacy Sandbox

Re: Better than cookies

While I didn't vote, I take a rather different view. Topics is better than cookies in the sense that me cutting you on the arm is better than me breaking your arm. I shouldn't be doing either, and you can act to prevent me doing them. Arguing that the cut is the better of two evils is not very relevant when you have the option of keeping me from touching your arm at all.

This is even more true when you consider that topics does not necessarily mean that any other tracking has stopped. It's just a new method, not as strong or dangerous as the ones that exist. Theoretically, people might eventually stop using the tracking they were already doing in favor of this one. I've already discussed this possibility in a different comment and concluded that I don't expect anyone to do that. However, it's easy to demonstrate that it has not happened yet, even if it eventually could.

Re: Is there a local database involved ?

Sure, we could probably compile a new version of Chrome that always says we're interested in the same three categories and get everyone to use it. I was going to suggest 1, 2, and 3, but 2 and 3 don't exist because whoever made this list may not understand how serial numbers work. If we're going to do that, we'll have to decide pretty quickly what our three magic ones will be, or we'll end up dividing the users unnecessarily.

The problem is that it won't really help for three reasons:

1. Anyone who is willing to use that version is likely not to be using Chrome in the first place. How much pollution can we cause if we're all using Firefox or even one of the non-Chrome Chromiums which don't have this yet?

2. Google will still have all the data they ever did and will be using that, not the new topics thing, to target their ads and tracking.

3. Advertisers won't give up on any of their tracking to use this. Honestly, it's pretty useless from the advertisers' perspective and they won't have any reason to use it. For example, there's a topic ID for tracking people interested in vacation services, but there are no topic IDs for where they want to go. If your business is advertising to people for a certain destination, that's next to useless. Paying to advertise your vacation package to Australia will be much less likely to work if you're targeting vacations in general than if you can target people who recently visited a site about travel to Australia (or, since they're not very good at tracking despite the effort, someone who recently did anything at all related to Australia). I'm sure they'll have some people look into the system to see if they can use it to track the users more, but they're not going to abandon their tactics for something less useful.

PhD student guilty of 3D-printing 'kamikaze' drone for Islamic State terrorists

Re: The cops also reportedly discovered at the home an IS application form,

IS were well-known for having an unusually good bureaucracy and documentation for things that other groups might not write down, such as exactly who they stole what from and whether the person from whom it was stolen was murdered, enslaved, imprisoned, or escaped. I'm not surprised they have application forms, given that their recruiters were probably quite eager to collect as much personal information on possible supporters as they could.

Israel and Italy have cheapest mobile data out of 237 countries

I don't think we have anything close to the stuff we would need to evaluate that. Take the maximum prices. I'm sure there is someone willing to sell you a gigabyte of data in the US for $83.33. If there isn't, I'm now willing to do so and I'll also sell plans at the UK's top price as well. In both cases, it's really easy to make a profit on that. What the study doesn't mention is how many people actually use that plan. For example, I've seen some ridiculous data prices sold by companies aiming at the hobbyist IoT market (think a 4G device designed to connect to a Raspberry Pi), but it's not marketed to people with phones and most hobbyists who want that will build their own and use a more normal provider. Should that plan be in the list, moving the averages? It might be better to weight it by number of customers, but that could leave you tracking the prices paid by people who didn't shop around, not the price you could actually get if you spent a few minutes looking. You could also use the price of a plan that you would view as optimal after doing some comparison shopping, which would probably be cheaper but means you could accidentally base your calculations on a price that's only offered temporarily. Now repeat all these questions for over two hundred countries. It's not an easy question to answer.

Re: Yes, but.

It doesn't always work that way either, for example in countries that subsidize the price of bread. A few countries have done this so they can do some damage to the economy without having that particular statistic become the focal point of a protest. There are metrics you can use to try to compare purchasing power, but they're not going to be simple.

Re: Market pricing factors

A coverage map will indicate otherwise. Yes, in rural areas, one network might have towers where the others do not, but that can happen anywhere. They all cover the entire country, and with a quick search to check those maps, they all have some claim to be better nationwide (fastest, biggest, most 5G by some unspecified metric, other unverifiable claims). I think your allegation of non-compete agreements is simply incorrect, but since they'd be illegal anyway I'm not expecting you'd have a citation.