Posts by doublelayer

AFRINIC election annulled after ICANN writes angry letter to African regional internet registry

The problem with option A, which is probably why they didn't do it, is that it's not easy or cheap to bring a bunch of representatives from small companies in lots of countries in for a vote. Proxy voting has a purpose, and I'd note that you could get the same result with a one representative one vote system. Specifically, some group that stands to benefit and has enough money to arrange bulk flights to Mauritius submits fake documents saying that each one of their passengers is the single accepted representative of one voter in this election. More expensive on them, but still within their capabilities. Remember that these people were submitting fake powers of attorney, which are not minor documents to forge, so they're not trying a small crime here.

Theoretically, there is no reason this has to touch a Chinese court. China doesn't control the routing systems. It would be easy for the rest of the world to see that Afrinic has withdrawn the addresses and no longer honor the previous users' routing announcements. It might not happen that way for fear that China would decide to continue using those and we'd get some fracturing of the IP system, but it could be implemented that way.

It's not a lack of demand. It's a lack of funds. ISPs in Africa are often more starved than most other countries for IPV4 space, and it means it's hard for things like local hosting to obtain the blocks they need, with consumers often relegated to horrendous CGNAT setups. But they also have low budgets, so they can't get those addresses in a bidding war. That's one reason why they have their own RIR in the first place. Sadly, if the people running the RIR are willing to sell those outside the region they're supposed to be in, it breaks down just as much.

It's a very good thing they didn't use that method this time. The people who showed up in person didn't submit their proxy vote and decide they could go to Mauritius after all. Their proxy vote was fraudulently submitted by other people. In a similar way, people who chose not to go but went to the effort of submitting a real proxy vote ended up with two votes, and since they weren't there, there was no way to verify which one was correct. This was more than a last-minute vote change and it needs to be handled more severely.

Microsoft dangles extended Windows 10 support in exchange for Reward Points

The article covered this: because you have to pay them monthly to have enough space to store your backup, and the easiest way to get that is to buy Office365, and if you do that, maybe you'll use Office365 and then you'll want to keep having it even after you replace your computer, and that would mean plenty of cash going directly to Microsoft's bank accounts. Or maybe some people will forget they have it or not remember why and keep paying for that OneDrive forever, the way I see some people with Apple devices who have higher storage subscriptions for iCloud but don't know why or for how long. No, it's not about AI training. Sometimes, the secret profit plan is less devious and more obvious than the theories.

Tesla Robotaxi videos show Elon's way behind Waymo

Re: Gosh

"I mean they produce more cars than Waymos fleet every few days."

And, if this was a question of how many cars they could put on the streets, that might be relevant. They're only operating ten. That doesn't seem to be their problem right now. When Waymo wants to scale up, they can buy more of the cars they attach stuff to and step up the manufacturing of the hardware they attach as well, and yes, that would take lots of money, but if they have a reason to make so many, they'd probably have the money to spend on doing it.

And, if the problem is with driving quality, manufacturing capacity is completely unimportant until that's fixed. For both of them.

Re: Gosh

Maybe, if we thought the information in the article indicated things were working well, there would be more positive assessment. Care to explain why those things aren't symptomatic of major problems? Incidentally, if you read the comments on articles talking about other self-driving car efforts, you'll find that there is a lot of skepticism of all manufacturers around here. When commenters who are already less pleased with the idea than the average are faced with a manufacturer doing it with less capability than those who have come before, you'll get even more negativity. You are, of course, welcome to explain how it's more advanced than we think it is, although I should warn you that if your argument goes something like "It's more advanced you idiots, why don't you just admit it", that's not going to help.

Re: Maybe not drugs

Given that this is his third, no fifth, no seventeenth, arbitrary number reference and every single one of them included a 4, a 2, and a 0, in that order, with no digits although occasionally a decimal point between them, no.

Just say no to NO FAKES Act, EFF argues

Re: Does privacy cover this?

Not all of that is related to privacy legislation, and some of it is not correct. For example, the existing laws that forbid you from using pictures is related to restrictions on what can be done with a "likeness" which is much more similar to the "property right" the EFF is complaining about than privacy, since privacy would forbid the taking of the picture in the first place rather than just some possible uses of the content of the picture later on.

And Facebook not using user posts in the EU... that ended and the regulators are fine with that. There may be a chance that NOYB will get a judge to block that, but it's not looking great so far. Maybe a better privacy law could do what GDPR is failing to do, but this is the US which has nothing other than a few patchwork laws in a couple states that have large carve-outs and are not well-enforced in the first place. If they had privacy legislation that covered this or even required similar permission as you'd need for a model release, then the EFF would have a point that existing laws covered it. Since those don't cover it, the EFF's argument has a hole in it, and I wonder if it might work better if they suggested to remove the dangerous parts of the law rather than abandon it entirely. The protection that law is written to provide is necessary and not available elsewhere.

That goes for other countries as well, where existing privacy legislation will probably not be sufficient to deal with AI training on personal data. The logic that Facebook's successfully used to get training on text allowed under the GDPR is the same logic that would be used to train on images. We can hope that a court decides that's a big enough problem that they'll block it, or we could make it official and not gamble on the result.

Does privacy cover this?

The EFF's complaints about what's in this are quite correct. Vague statements about what is covered make this law dangerous, and virtually unlimited requirements to take down maybe almost anything when maybe anyone asks you to is a perfect way to get bad consequences. Focusing on the existence of tools that could theoretically be abused is looking in the wrong direction when focusing on those abusive uses is more important, especially when there are legitimate uses for the tools. They're not wrong about what this law does badly.

However, I have to wonder if they are wrong about what is needed to accomplish this goal. Privacy legislation would be very welcome, but I'm not convinced it would solve this problem. Privacy legislation might prevent a company from getting a bunch of pictures of someone without consent and then making a fake avatar of them, but there are lots of ways of doing that which wouldn't be covered. The problem is that AI companies have acted as if most laws don't apply to them, and unfortunately, they've just gotten a judge to say that they can use data for whatever they want with the only proviso being that they had to obtain it legally. The fact that that counts as a partial win against AI companies speaks volumes, since normally, obtaining something illegally would be an obvious problem and nobody would need to question that part of it in the first place.

So let's consider how a company could legally obtain pictures of someone who doesn't consent to them being used to create a fake video of them. Several really easy things come to mind. Maybe they're an extra or actor and agreed to appear on camera, but not to having that footage used for literally anything later on. Or what if they posted video or photographs of themselves to a public social media website? In both cases, they voluntarily chose to put their likeness online, and that would remove most of the privacy protections around it. Do we really want that to be the only bar preventing someone from using that data without their consent in such an invasive way? To prevent that, the privacy law would have to specify that any subsequent use of the data had to be agreed to beforehand, and any data produced in violation of that had to be removed. Or in other words, the privacy legislation that would fix this problem actually looks quite a lot like this legislation does.

If the EFF is so confident that some other law can fix this problem, they need to do a better job of explaining how. Until then, they only have half of an argument, which makes it a lot easier for anyone motivated to dismiss and discredit them.

Iran cyberattacks against US biz more likely following air strikes

Re: Shocked

"My understanding being the US hit the nuclear sites."

Places where they make nuclear weapons are military things. That's especially true when they're run by the military as well, so there are military members there, but even if it only employs civilians, that's a target. But if you don't think that is correct organization, basically any time one country deliberately bombs another one, it's a war. The place in Qatar that Iran bombed had US soldiers in it. They are at war and have been at least since that first bomb fell. Of course, since this isn't a neatly organized declared war, they can stop being in a war unofficially by just no longer dropping any bombs on each other; you don't need a formal peace treaty to end a war if both sides decide they're through. Maybe that will happen. It won't have stopped this thing having been a war.

And even though it's definitely a war, you can still support it. Nothing says that you can't support a war. Nothing says you can't support the aggressor in a war, although since it's usually the less popular choice, there are numerous ways to describe things such that the other side is the aggressor and you could do that with the US-Iran war, and actually quite a bit more easily than with some others. You can have and defend the point of view that war was a reasonable and appropriate reaction to previous events if you feel that way. Just know that you will have to have that argument if you support this, because war is what they did.

Re: Shocked

Whatever the purpose, bombing someone counts as a war. Israel and Iran have been in a war with one another for a while, and the US and Iran are in one now. You can argue that these weren't the first attacks of the war and that it's existed for some time, or you can argue that it just started last week, but either way, what they're doing now is called war. That doesn't mean you shouldn't have a side here. You can support one of the belligerents in a war, or you can support neither of them, and your opinion can change as events do, but it won't stop it from being one. Directly dropping bombs on military members from the other side is one of the most obvious ways to determine whether it is one, and both countries have been doing that.

Re: SSDD?

There's often a degree of restraint involved, where the attackers still do things they find useful, things where they don't think they'll get caught, or little prods in reaction to something, but they deliberately don't do something against someone who might respond by escalating. That restraint is unlikely to be seen after large bombings, so I would expect that some things may be different. Of course, they could still have some restraint if they think the retaliation might be more bombings. So far, I do think there has been said restraint, I don't think the level of activity in the last few years is the most they can do, but I don't know what they will decide to do now.

Huawei chair says the future of comms is fiber-to-the-room, which China has and the rest of us don’t

Re: FTTR? Really?

Of course some people might need or want that increased speed, and they can and will pay for as much fiber as they can take. There are various places that could use that much. For example, some places that do research on massive datasets and move them from big storage servers to big processing computers and back again would quite like to have fiber between them. However, we're not discussing whether it will be used anywhere or saying that anyone who thinks they need that speed is mistaken. We're saying that most users do not and therefore will not pay for the development of the technology.

And that is true for two reasons. Most residential users do not need that much bandwidth because they're not moving massive files. Streaming video won't use that much bandwidth either, and after a certain point, even large files download so quickly that people no longer care how quickly they moved. And those researchers I talked about in the last paragraph: they're not using it either. Oh their storage and compute systems have fiber between them just like I said, but instead of installing fiber between the server room and their lab, they put both things in the server room and they access it from a computer in their lab which only has to show the interface of an application and, at most, some real-time high-res video, which can work just fine from the copper line that was already in the building.

We will never stop demanding more performance from our technology, but most things do eventually reach a level where fewer and fewer people need more. I have 32 GB of RAM in my computer, but a lot of people don't need that, and I don't often need any more than that. Some people do, so I can't say "32 GB ought to be enough for anyone", but neither should I expect the average consumer or business user to drive down the prices of 128 GB DIMMs through mass demand. Maybe some software will come along that needs it and is used by everybody. Maybe memory makers will keep finding more efficient ways to manufacture it and I'll buy it because it's about the same cost as the lower capacity so why not have the extra capacity. But maybe neither of those things will happen and I'll stick with 32 GB and my family and friends who don't run as many VMs or big dataset crunching will stick with their 8 GB and 16 GB, because not one of them has asked me for more RAM in the last decade.

Re: An eventuality..

I disagree. Your comment demonstrates why in one way, and there's another problem with it. Here's one:

"Folks were pre-wiring CAT6 10yrs ago for the impending 10GbE rush in homes"

And how many of them have 10 Gbps uplinks from their homes? Few can get it. One of the reasons is that few who can want it. I actually saw it as an option once for a residential connection where I was helping a friend get connected, although that required unspecified additional installation charges. Did I buy that for him? No. Did I buy the 1 Gbps connection with no additional charges? No. I bought the 100 Mbps connection, because he lived alone and didn't routinely upload terabytes, so he would notice the hundredfold speed increase not at all, but he would notice the higher monthly bill every month. To some extent, I expect that hardware will improve and faster service will become the default because it's no cheaper to provide anything slower, but that's far different.

There's another problem with FTTR, assuming it means what it looks like and isn't the translation mistake people have described, namely that it doesn't work the way most users do. Most users don't want to plug their hardware into every room. They may have a few things they want cabled in, and the rest of their bandwidth is consumed over WiFi. Fiber going to a box that can handle it, attached to a few meters of copper, is not going to have such a speed difference over sending fiber those extra meters, but it's especially not going to have a speed difference for all the devices where WiFi is the bottleneck. Having fiber to each room would make it easier to have good WiFi coverage in the same way that having copper going to each room would, and in most cases, with no difference in the experience from the user.

Re: Going beyond 10Gb/s requires fiber for now

"The "build it and they will come" argument is dumb."

That is what I said in my last comment, but it is still less dumb than saying that they're the first to bring fiber lines to individual units in a multi-tenant building, which has been done in new builds and retrofits in many countries for at least a decade. I don't think we're going to see massive increases in bandwidth needs from residential users, and they'd get happier users by bringing existing quality to the many buildings that don't have fast cabling yet. Of course, there are datacenters that will need fiber going everywhere, but they tend to already have fiber going everywhere.

Re: Going beyond 10Gb/s requires fiber for now

As I said in another post, there are two options:

1. You are correct, FTTR is a mistranslation, and they are lying to claim superiority because that's often done in buildings with fiber, either actually delivering the fiber to each flat and putting something fiber-capable on the end of it or bringing fiber to a local network closet near the flat and having a short copper line from there into the actual flat.

2. You are incorrect and they actually do mean bringing fiber lines into the building and connecting them to each office or room of a house, in which case they're right that we are not doing it and maybe they are, but I don't think we'll turn out to need it any time soon.

I don't know which one it is, but number 2 is the less stupid argument, because it is very similar to the arguments for 5G and 6G they have been making for some time. The theory goes that we will eventually need more bandwidth for something, and the places with the technology to provide it will be better situated than the ones that stuck with cheaper technology that becomes a bottleneck. That was true before; houses with some internal ethernet cabling are nicer than ones that don't have that and need a mesh WiFi setup to the one location. 4G data was much faster than 3G and enabled some uses that 3G wasn't very good at, and before that, 3G was capable of data traffic at a moderate, usable speed whereas 2.5G data traffic was often either too slow or too unreliable. The problem for Huawei is that the cheap and old version of the technology we have now is actually working rather well, so to justify replacing it, we will need something that uses more bandwidth.

Re: FTTR? Really?

If that's what they mean, then they're wrong to say that China has it and other countries don't. That's still a home, whether there's one or more rooms in it. What FTTR suggests is that in something like an office, the inter-office cabling is fiber as well. There are only a few situations where that would be noticeably different from older copper cabling, but it would be noticeably more expensive.

The AIpocalypse is here for websites as search referrals plunge

Re: khtml -vs- Gecko

That's how user agents work. It's a mess of annoying cludges from history. That's why pretty much everybody's starts with "Mozilla/5.0". Sometimes, I think we'd be better off if we decided to simply discontinue using the user agent altogether. It had a point, the point is broken, just hand out the document for the URI requested.

Their point, however, was that the bot did not decide to identify itself as one. The bot is masquerading as a browser, at least by user agent.

Re: Some of hese comments seem to miss the point

Perhaps you are unfamiliar with the concept of a patent, which means there is a public document out there describing how the thing works, in a public database. That's what you have to do to get a patent. I think the company concerned probably didn't want that to be easy or they would have put their patent numbers on their website like others do, but it's still out there and available to me. Eventually, I can find that patent and read it to learn what makes their thing different from others. I was hoping to avoid having to do that manual search by using a tool which is supposed to be better at searching than I am and was probably trained on a bunch of patents including this one (from 2016) since they are public documents. It wasn't, hence why I had to do that search, which took a while, but now I have the details I was looking for.

And guess where I got that patent file eventually? I tried going to a few patent offices first, because they store the databases. Their search systems aren't very good, though. That means I ended up finding the patent number in question at...drum roll please... Google patents. You think they might have trained their own AI on that? The point being that this is yet another case where AI will try to answer a question it has no ability to answer and prove to be useless.

Re: Some of hese comments seem to miss the point

When I do searches, I don't know the answer. A wrong answer is not any better than a useless answer. In the case of your pepper grinder example, I don't expect the AI to give me a correct answer tailored to the model of grinder I have, but instead what tends to work on most grinders. Twisting the top would have been my guess to open it, and then I expect there is a hole where I put in the new peppercorns. If that's all the AI summary says, it is no better than my guess. So just to check, I put the prompt "How do I open and refill an OXO pepper grinder?" into some LLMs, and that's what I got, paragraphs to tell me to open it and put the peppercorns into the hole, which is located either on the top, bottom, or side.

Meanwhile, if I ask a question about something more specific, the AI summary is more often wrong than useless. Either way, the answer is so frequently wrong that, when it is right, I can't count on that. Meaning I have to look at results to try to figure that out which is what I was going to do anyway. Maybe you're getting better results, although the people I know who trust LLM results have often found that following those instructions has given them results they don't appreciate.

Re: Some of hese comments seem to miss the point

If you're warning us that they can always make the AI answer worse, granted. However, it's already annoying for most of us because the answers are often wrong or useless. For example, I recently deliberately ran an AI search through Google and Perplexity because I wanted to know how a certain technology works, but all normal search engines sent me to pages that had no interest in the technical details. What I really needed was to go to the effort of finding the patent for the technology or some introductory material describing the mechanism and get a summary of that, but that wasn't very easy to find in search results (including the word "patent" in the search query just found a lot of pages telling me that the mechanism concerned was in fact patented).

Seems like something that search-equipped AI agents should be able to do. They can find which pages have that information and distill the information for me, right? And I expect that they may get the information wrong, but at least Perplexity cites sources, real ones that actually exist, right? So I run the query and the LLM happily tells me that the mechanism works "electromechanically". Which is true. Yay, the LLM didn't make anything up. Of course, that's hardly any more detail than saying the thing was built out of atoms of some elements mixed together, or in other words, the AI search result was even more useless than the search results I had before, but it was pretending not to be.

In other cases, the answer is simply wrong, but people who assume that, since it looks plausible, it is probably correct end up acting on it anyway. Ideally, you get the correct answer to your query, but you're probably not getting it as often as you think.

Re: Be careful what you wish for

And you've gone too far the other way. The law and the politicians who called for it, both the Canadian version and the Australian version which was almost the same, specifically said that they had to do something about the problems of linking and quoting. If they had just stopped at quoting, maybe they'd have had a point, but they called out linking from Google and large social media as a problem requiring compensation. They were surprised when the response to a law that was passed because linking is bad was that the people they spent months saying should stop linking or pay for doing so decided to stop linking.

The problem was stronger on the social media side, where people, including the social media companies themselves, had an incentive to copy the entire text of a news article and put it on that network so people stayed there instead of going to the paper's website. Sometimes, to try to make their act less obviously illegal, they'd only copy the first four paragraphs, thus removing important context without cutting it too short that anyone was motivated to leave. Obvious copyright violation though that is, I can see why someone might decide another law is needed to do something about it. It is, as you will have noticed, entirely separate and opposite to linking, which would be the good thing they could do instead, and that's what many search engines were already doing, quoting a couple, possibly cut off, sentences for context but requiring people go to the site to read the whole thing.

And yet, the laws were written so badly that the only way to interpret them was that linking was the problem. What the backers of the laws wanted was to take money from tech and give it to newspapers, and that's why they included linking, but the result is that the law effectively tells you that including large newspapers in search results is a bad thing and you shouldn't do it.

Huawei's latest notebook shows China is still generations behind in chipmaking

Re: Still generations behind

"A 2 year old machine without AI crapware is probably faster than the latest one with it, and all the better for not having any."

That is irrelevant to any discussion of power efficiency unless that AI is absent from the old machine and completely un-turnoffable from the new one. Neither of those is true; a two-year-old machine will get any builtin AI functions the new one does and both of them can be disabled equally easily. If, at some future point because they haven't yet, they make it impossible to turn off something, they're not going to make it easy to turn it off, just for people with a different processor than yours.

Since we're talking about desktops or laptops, imagine their relative performances when wiped and a suitable Linux distro installed. No AI to complain about. Then, the actual power efficiency numbers can be considered. In most cases, the machine that's two years older will be just fine, and the difference will be small enough that it's not noticeable. That might be a little different if you're running ten thousand of them, but even then, having to buy new hardware to get that minor benefit is likely to mean you accept the costs of the less efficient processing rather than the costs of ten thousand new machines.

That isn't as true when the machines you compare to are a lot older. I've known people who were running 2009-era machines with a Core 2 Duo in them which were running at an average power consumption of about 120 W. They were getting about as much performance out of them as a Raspberry Pi 4, and possibly not needing all of that. A Raspberry Pi 4 can run in 10 W. So in that case, the efficiency savings would be noticeable and would cover the cost of the new hardware. For example, at the average UK power price of £0.245 / kWH, a power saving of 110 W, and a purchase price of, let's say, £60 for the Pi and some accessories I assume you don't already have, the time to recover costs is three months.

Re: Weasel words

I'm not sure what you're getting at, since the entire point is that TSMC, Samsung, and Intel had 7 nm earlier and now have newer methods, though of course they still use 7 nm. Also, it wasn't comparing all of Asia, but China, and specifically SMIC, to arrive at the "generations behind" claim. TSMC usually has had the most advanced manufacturing for the last few years, so that they're further along than Intel is not really a surprise.

We could also debate the relevance of that, because many cases don't mind that the chips are built on 7 nm tech. Still, if you need lower for some reason, TSMC can do it and SMIC can't yet, hence they are actually behind. Which itself is also not a surprise because until recently, China was buying many of their advanced chips from TSMC, so we've only recently seen the results of throwing a lot into getting that capacity.

Former US Army Sergeant pleads guilty after amateurish attempt at selling secrets to China

Some of them probably were, but some of that was also spammers trying to do it cheaply and not having great English skills. It also really depends on what kind of phishing this is. The kind that says you won a prize, but you need to follow some steps by talking to them is the kind that benefits most from deliberate errors, because they're sending that to millions of people and can only talk to so many to get payment details from them. The kind that gets sent to you looking for you to click a link and enter a password benefits a lot from not having those errors because they benefit from a very brief level of trust. A lot of phishing can work well without those deliberate errors, and a lot of the people using it as a tool aren't good enough at their job not to make them.

UK students flock to AI to help them cheat

Re: “um…?”

The US and UK have different versions of the pint. They're both an eighth of a gallon, which doesn't help because the two countries also have different versions of the gallon. And before some UK person insists on lambasting the US for changing everything, the US pint is the older one, and the UK one is the one they made up in 1824. And before some US person lambasts the UK for that, the US decided to have two measurements of volume, one for liquids and one for solids (if anyone tries to measure gasses in pints, I'm going to pretend they're not talking) call them both a pint, and have them be completely different. Anyone outside those countries can look down on both systems.

"But LLMs as a tool to turn an argument I've both reasearched and made, and as importantly organised, into long-form text - seems, in principal, no different to using a calculator."

I'm not sure how you arrive at that conclusion. There appear to be many major differences, all of which are incredibly important to education and most of which are similarly important to any other use anywhere.

In education, teaching writing can be as much the point as teaching the information the student is writing about. History essays are testing both of those skills, and literature classes often care or at least should care more about the writing of the essay than the analysis of whatever Shakespeare was trying to say there since the students have little experience of Shakespearean language or context and are unlikely to come up with anything interesting. Even in something as simple as a lab report, the writing is important as well as the scientific information within. Producing something adequate but being unable to do it again is exactly what that education intends to prevent.

LLMs also have the nondeterministic problem. If I use a calculator, then as long as I put the right numbers in and it isn't broken, I get the right answer. If I get the wrong answer, I did it wrong. I can put correct information into an LLM and still get wrong data out. I frequently get new evidence for this. For example, I had a situation where I needed to describe something in about two sentences, but I found it tricky to get the useful information in in less than a couple paragraphs. So I decided to see what an LLM would do with the task. Instead of taking my paragraphs and distilling something out of them, or even selecting two sentences at random from them which would have been better, it invented new incorrect information and gave me that. That's a bigger problem for those who accept whatever an LLM produces without question, but it isn't comparable to a calculator if you have to do that.

Re: Erm, but how many are getting away with it?

"I have never either agreed with, or seen the sense in requiring coursework to be submitted electronically rather than written in longhand"

Mostly because it does nothing whatsoever to prevent cheating, since it's really easy to take cheated work, whether LLM-generated or plagiarized from another source, and write it on some paper. In the past, it did make it harder to check, because comparing students' work for similarity was easy to automate, but LLMs can now generate new cheated essays for everybody so that is less easy. Still, there's no benefit, as far as I'm concerned, in making people handwrite versus type, and since most people will be expected to write by typing later on, that works for me.

If, instead, we are trying to prevent cheating, we'll probably have to worry more about where students do work than how they write what they do. It's much easier to prevent cheating when they're doing the work in a constrained and observed environment, so more tests and less homework seems like the logical start, but there are many courses that wouldn't handle that well and it involves spending less in-class time on teaching. Still, the other approaches that have been suggested haven't struck me as likely fixes. I don't hold out much hope for teachers managing to phrase assignments in a way that an LLM can't answer, I don't think AI to detect text written by AI is going to work out very well, and how they turn it in has no effect at all.

If there is a flood of Windows 10 machines which can't be upgraded to Windows 11, install Linux on them, an editor, remove the browser and maybe as much of the network stack as you can, and have the students type on those. Nothing says that writing has to be on paper to avoid LLM output.

LibreOffice adds voice to 'ditch Windows for Linux' campaign

Re: 430 million

"But if Linux Mint acquires a quarter of the market it will shake Microsoft to its foundations."

I agree, but let's phrase that in a different but equally correct way:

"But if Linux Mint acquires half a billion new users and doesn't lose them within the next six months, it will shake Microsoft to its foundations."

That's not going to happen. Even if you try your absolute hardest, it's not going to happen, but most involved are not trying their absolute hardest. I am not either. I'll install Linux if requested, and I'll offer it if asked for options. People aren't asking, and when I suggest it, they aren't requesting I install it. They have asked me to install Windows 11 though, and most of the time, the computer they have does support it. This is why I'm less certain about that number of unsupported computers. There is a large number of machines that can support Windows 11 but are still running Windows 10 (let's call that X). To know how many machines in active use are going to be cut off, we need to know the value of X, but it won't show up in surveys of OS usage so we can only estimate it, and the most reliable way to estimate it is to have figures about which processors and TPMs were used in machines sold in the 2017-2020 period which I don't have and I don't really want to read.

Re: 430 million

Of course, and those will work great. However, they are not as common in businesses as they are among home users, and businesses are more likely to do something organized about the end of support, whether that's installing Linux, buying patches for Windows 10, or getting new hardware. Home users may do all of those things too, but two other options are quite likely: break the checks on the Windows 11 installer to install that, which is much easier for them than it is for corporate IT because they'll only have to run manual steps on one machine rather than hundreds and no security audit is going to complain at them for having done it, or just stick with Windows 10 with no patches because they don't know or don't care about the risks.

If you want either group to switch to Linux, I think you'll have to be more active at getting them to do so given the many other options they'll consider. If you want to switch enough people to Linux that Microsoft really feels it, I think you have to convert many businesses.

Re: 430 million

I don't know where that number came from, but remember that many devices that don't run Windows 11 have broken since they were bought. Basically every machine released in the last seven years will support Windows 11, and businesses often replace user desktops more regularly than that. Some businesses might do a three-year or four-year replacement cycle for some reason, despite the fact that they're getting rid of perfectly functional equipment, but even those who sweat their hardware see a lot of attrition due to damage and will have bought plenty of machines since that date. That means a lot of computers that can't be updated are no longer in use by the places that would be most worried about it and therefore the most likely to switch to Linux if they had to to keep updates coming.

Remember also that Microsoft has extended updates for Windows 10. You just have to pay for it. Businesses who care about updates and don't want to replace their hardware have that choice as well. Probably a few of them will take it. I still predict that there will mostly be a wave of people updating to 11 at the last minute and another set of people who never update, even some who easily could but just don't want to, even as security updates end. If you want a lot of those people to switch to Linux, then you have to shout loudly, as various organizations have been doing, and make the process easy, which none of them have really done, and you have to do it quite soon. Some people will take this opportunity to try Linux and find they like it, but that's not going to be 430 million people even if I'm wrong about that number of computers being an overestimate.

Re: As I've said before

The complaints are that you can't use that with the keyboard-shaped 500. You can do it with the 5B, but when the 500 was designed, they didn't put an M.2 connector on the board and they didn't expose the connections to attach such a hat.

My solution to this would be to not buy a 500, buy a 5B instead, and plug in a keyboard. But then, I still don't understand why people like the keyboard units more than the SBCs, so there may be a reason why that's not their chosen solution. To anyone who has that preference, I'm still curious why and would appreciate hearing your reasons.

Techie traveled 4 hours to fix software that worked perfectly until a new hire used it

Re: manuals

It's often quoted as from him, but it seems to have a somewhat clear spread from non-famous initial usage in the addiction recovery area three decades after Einstein died. Unless those quotes were made up too, it also makes more sense in that context* than it does in modern usage, where there is insanity that doesn't involve mindless repetition and mindless repetition which can better be described as stupidity than insanity.

* The first quote used "insanity" to contrast with "sanity" which was being used as a synonym of sobriety or at least a lack of addiction and related negative effects on life, and the repetition concerned involved the repeated acts of someone with an addiction even though they knew it was harming them. Suddenly, it actually makes sense to me how that quote would have come to be which makes a lot more sense than IT people dealing with frustrating users. Incidentally, although I previously knew that the Einstein source was misattributed, I only read about this initial usage about five minutes ago.

Re: manuals

I agree that it's not a good definition of insanity*, but I do think we need a word for whatever that is, because I get frustrated by that far too often. People who, because they don't know what to do, try the same thing over and over even though they have no reason to believe that anything has or even could have changed so it will work where it didn't before.

If it said "try again later", fine, keep repeating it. If you've changed something else that should unblock it, okay, run it once more. If it's exactly the same, pressing the same button on the theory that it will eventually do what you want instead of that thing that button does is a bad idea, and I wish I hadn't seen people use that approach.

* I also have trouble coming up with an actual definition for insanity. Defining mental illness is hard enough, but defining that unclear line where it goes to that higher level has often proven challenging to me. Usually, I think that the core part has to be dissociation from reality where they either invent nonexistent things or fail to recognize existent ones, but that too can be hard to define. It's a good thing I'm not a psychologist.

European consumers are mostly saying 'non' to trading in their old phones

The more likely versions are:

Trade-in site offers you £5 and sells your used device for £120. Nobody buys it because it is not worth that much.

Trade-in site offers you £5 as a discount so you buy the new device from them, then tosses the device into the bin.

Trade-in site offers you £5 sight-unseen with a little footnote about determining device condition. That makes sense, because you can't tell them everything about the device you're sending in and there could be a problem that reduces the value. They get it, immediately toss it into the bin, and tell you it is actually worth £0, but are you going to complain about such a minor loss? Are you going to pay for the thing to be shipped back?*

The alternative that gets closest to the place you're describing is where you post it on a marketplace and find a buyer, and except for any fees you might pay, the money they pay is the money you receive. Great, let's do that. The problem is that you end up spending a lot longer describing, photographing, posting, and responding to questions about a device that people probably don't want to buy because a fair price is about £25 and the lemon problem applies. That is why, in my example above, I was happier to simply hand the phone to someone who would use it than to post it for sale and deal with this work for an uncertain and small reward.

* The trade-in for value that vanishes situation is one I've experienced personally, although in that case I was expecting it. A friend's ereader had been broken, and I was buying her a replacement. They had a trade-in program and offered me a small amount for the old one. It asked me to describe damage, and it had a field for "screen broken" which I selected, but they still offered me a nominal value. To not my surprise, they determined on receiving a ten-year-old device with a mostly nonfunctional screen that they were no longer going to offer me that discount, even though in this case they had all that information up front. Of course I didn't complain, because I didn't want the broken thing back and I didn't think there was anything better to be done with it. I also wouldn't be surprised that they do the same for functional devices. It's an easy way of taking competition off the market.

Re: Do you really need that expensive phone?

The dumb phone finder has filters for both WhatsApp and Signal. If that is your primary problem leaving the smartphone, you might find some candidates there. I have not used it or the devices it recommends as I appreciate many aspects of the smartphone and don't want to leave it, but it seems like these people have categorized a lot of devices.

Exactly. In my limited experience, the only way to get an even slightly reasonable number for trade in value is if you bought the thing a year or two ago and it was a flagship model. I don't buy flagship models because they're way too expensive, and if I did, I'd keep them much longer than that. In fact, I keep my cheap phones a lot longer than that. By the time I don't need one*, no organized business is offering much for it. In some cases, they might have a little more resale value if I sold it myself, but at that point, I end up having to decide how much it is worth to go through the effort of posting it, responding to questions, and hoping the person it gets shipped to isn't trying to get a free phone.

* Most of the time, I stop using a phone when it's doing something unfixable, so that isn't usable by anybody except for someone looking for parts, and I don't know how to find them. On occasion, I have had a phone that worked but I didn't want it, for example the case where I needed a temporary device and when temporary ended, it was still usable. Stuck on an old Android version of course, but it actually got some security updates. I still couldn't easily sell it and ended up just giving it away to someone wanting a phone for a child.

Australia finds age detection tech has many flaws but will work

Re: Data collection...

"you are suggesting that "We should just sit here and suck it up"........and do absolutely nothing to protect our own privacy!"

Am I now? No, I'm suggesting two things to everybody else. One is that we focus on more things than our own privacy, since the privacy of others is actually important no matter how selfish we are. One of the best ways to avoid surveillance is to be one of a crowd. That means making sure everyone can use encryption, because practice shows that lots of people will use it if they're allowed to and that protects us all. The second thing I suggest for others is to actually think about how well their personal privacy methods will stand up to attack, because they are being attacked, and modify their plans. You have instead chosen the complacency that your methods work (somewhat) and would still do what you describe under legislation such as this (no), and those who are worried about their own privacy should be thinking further and making plans to keep it.

But, if you've decided that I'm calling you a criminal for seeking privacy when I've been condemning privacy-invading legislation, the rest of this clearly isn't getting through. In that case, you're right, your system is perfect and you'll be fine, no need to bother worrying about anybody or anything else.

Re: Data collection...

The government will be happy to decide what documents count. If a birth certificate counts, then yes, go ahead and use that. If it doesn't, then take your birth certificate down to the people who approve passports and get one of those. Either way, you will end up using paperwork the government permits for the purpose. You write as if using papers to establish identity is something governments don't know how to do. Pretty much everywhere does it, and they have their list of acceptable documents for the act you're doing. If you want to do the act and you don't have the accepted documents, then you have to go get one of them in order to do the act. Of course, the ideal document acceptance list for this particular interaction, visiting a website that has no need to identify users is "none required and none accepted if the user wants to provide one", but that's not going to happen if a law like this gets passed. Australia's considered three alternatives and since identity verification is the only one that's not really easy to circumvent*, that's where they're going to land.

* One of their other options is AI age verification, which will be really annoying and risky, but depending on what images you have to put in to be verified, it probably won't be too hard to spoof it with a quick sample from someone who looks older or even an AI adult video generator. It will also be a nonstarter because it will get a lot of the answers wrong in the opposite direction. The remaining one is trying to use a credit card, which is a lot easier to steal than valid identity documents, can be used frequently, and won't even be noticed much if they're just checking validity without charging it.

Re: Data collection...

"Only if "the system" knows that the specific identity in question is fake. How would "the system" know this?"

We appear to be in a loop. Consider again, over and over if necessary, what you would do if they told you to scan your passport. You have four logical choices:

1. Scan your passport.

2. Scan someone else's passport, knowing that simply obtaining it has made you guilty of identity theft and you can go to prison.

3. Scan a fake passport and hope they don't have the technology to immediately identify this. Most likely, they can get that if they need it and they can make your fake a crime too.

4. Refuse to use the service.

Now you are correct that some people have indeed successfully committed identity theft before, and not just people with some degree of legal cover; people in your hypothetical position, I.E. criminals, have succeeded. Maybe you'll be one of them. Others have ended up enjoying some time in a room without internet access because they tried. Some of us might think that, if you actually commit it, you belong there. It's also much harder to obtain identification documents for someone than it is to obtain a burner phone or pseudonymous (happy now? I use the right word) email address, even if you are overestimating how well those work. Other than that, you have no problem with the course of action you suggest. I am worried about what happens to everyone else, which evidently doesn't concern you, and at this point, I'm not too worried about what befalls you as you clearly have a plan you like.

Re: Data collection...

And you write as though you missed the part about verifying age by looking at identification documents and denying you access if you don't. Your burner phone (incidentally, I'm guessing you're not in Australia because Australian phone law is stricter than UK or US regulation) and emails will do no good if the website you want to access with them says "photograph of your passport or you're not getting in, and by the way a photograph of someone else's passport is identity theft and a photo of a faked passport is fraud". You can take some actions, and doing so helps, but two things remain the case and should be taken into account when we consider what laws should be:

1. Anonymity mostly works when the people you're interacting with are content to allow it. Only a few things work when the rest of the system involved is determined to block you.

2. The things you do to remain anonymous are difficult or expensive and many others don't do them, either because of the difficulty or because they don't know that you can.

Chances are that what you think will protect you isn't as strong as you think it is, but even if it's perfect, we should oppose these laws for others and so that the next push doesn't break the protections we use.

DHS warns of sharp rise in Chinese-made signal jammers it calls 'tools of terrorism'

Re: "illegal key fob jammers"

That's not a jammer. Jammers transmit to prevent something else from usefully transmitting. Your tin is a passive thing to prevent a transmitted signal from getting to where you don't want it. They're completely different and your one is legal.

Re: Don't panic people

Some of those are real concerns, for example if there is no coverage in an area, that will be a problem for emergency comms in the area. Some of them are not. For example, as long as they use radios, a jammer can jam them. It doesn't matter whether those are cellular frequencies or some other frequency band because the jammer hardware is almost identical, so unless they have massively multiband radios that can hop around to whatever band is not being jammed right now, someone with the hardware can jam emergency radio transmissions if they are willing to put in the effort. Criminals jammed traditional P2P police radios as a tactic before cellular was even an option.

Microsoft testing PC-to-Cloud-PC failover for those times your machine dies or disappears

I don't think it's about data. Mining that is not easy. Mining it and getting anything useful is very hard, since most will be random business files of no import to anyone, possibly including anyone in the business. One reason I think this is that, if I've interpreted the article correctly, they're relying on using OneDrive to store user documents, which they've been pushing for some time and occasionally gets used as a sort of backup system*, and so the data they'd be syncing is already on their servers anyway.

To me, the explanation is much simpler. If you can convince someone in IT or IT management that this works well as a backup solution, that's another item in the Microsoft bill that gets paid without question forever without anyone thinking about whether they're using it. If they actually use it, then they may start using more cloud desktops which means plenty of money per month for what you could buy for about three months rental. If they don't use it, then Microsoft gets paid for merely having the ability to provision one if requested which doesn't happen. Either way, it's easy profit.

Europe slams online tat bazaar AliExpress for dodging obligation to stop dodgy traders

I've seen lots of obviously dodgy tech items on there. Even when something is plausible, it's mostly random whether you get something the manufacturer is able to make really cheaply or a quick imitation of what that product might look like. I needed a Bluetooth audio to 3.5 mm audio device, and I got one from Aliexpress for less than it would cost me to make the plastic enclosure of such a thing, let alone getting it shipped. It's been working very well ever since. On the same visit, I was advertised a phone with a terabyte of storage for about £50, but I'm pretty sure that wasn't going to be the same story if I bought it. A lot of the time, you're deciding for yourself whether the product being advertised is actually plausible, in which case it's probably real, or if it's too good to be true in which case it definitely is. Eventually, someone will get smarter and start advertising fake products using descriptions and prices that make it look like a possibly real one and the platform will die, but for the moment, it's a dangerous area full of fraudsters that can be successfully navigated if you're careful.

Regulation would be welcome, but like Amazon and eBay before it, I doubt they'll do enough to have the effect they'll need to. Policing a market where people can list their own products without review is hard, and the platforms that have that challenge have mostly responded to it by deciding to not bother. Regulators of those platforms have decided to mostly ignore this. Until both things change, this situation will not end and it will be up to the customer to be as careful as they can.

Firefox is dead to me – and I'm not the only one who is fed up

Mozilla undoubtedly has more people than truly needed to accomplish the goal, but you can't make a comparison like that. Seven people have not yet built something, but you're implying that their 2026 alpha will arrive on time and work, neither of which is proven or provable. I've aimed for a release date before. It doesn't mean I achieved my goal each time.

Also, even if they do have a release by then, there are lots of things the alpha will probably break on, which is not a surprise, because it's an alpha, but how long do you think it will take to get from 2026's alpha to full, production-quality, tested release. Also, it's often quite a bit easier to build something from scratch than it is to build the same thing among a bunch of older code which you have to avoid breaking, to make sure everything works on all the platforms that run the same code, or at least subsets of the same code. They also develop more things. Does the Ladybird project intend to develop desktop and mobile browsers? From a quick read of their website, I'm not convinced they are writing anything other than the engine, but Mozilla does build all those other things. Having another browser engine would be great and I hope they do hit their targets, but I'm not going to jump from "I want it" to "it will definitely happen and be perfect".

When Ladybird has a functioning release, we can compare their team size to Mozilla's browser engine people. Until then, the comparison has insufficient information. Even when we have that information, comparisons from a team building one piece to an organization building many will be invalid.

Re: They keep removing features. When was the last time they added one?

Not all the things you listed have actually been removed. Some of them are still there, so I must conclude that you didn't look for them.

The problem with adding features is that they'll often only appeal to a specific subset and anyone outside the subset tends to be a little annoyed that they spent effort on them. For example, they did add a new feature relatively recently, and it's one I quite like, namely translations a la Google Translate but using offline models. I use that feature a lot and get privacy benefits from it. But if you don't, then I could understand why that doesn't seem to matter much. The article author has a similar thing in Pocket, which I have never used and was rather unpopular when Mozilla spent money on it, but now it's unpopular when they stop. I don't think all the things Mozilla does are valuable, but I have to consider that maybe there are others who benefit more than I think from the things they've developed.

I've got plenty of complaints about Firefox, but I have always ended in the same place, which is that they are the best browser of the realistic choices available to me. Everyone else is either more limited (Safari or Chromium-based things), or they built their thing around what Mozilla made. If I use Librewolf, I'm still using most of Firefox, and the project would not exist or would not work well if Mozilla wasn't building the things they do. Of those who bother to do it, Mozilla is doing a better job than the rest of them in my opinion.

Re: Fork it

What do you mean by that? Because I can still change the search engine, either the default one easily or to a different one on the fly for a search. Admittedly, they did annoy me by removing one of the methods to do that, but that's just a UI change and I can still do it, just not by the keyboard-friendly way I used to (typing the first few letters of the search source and pressing tab used to do it, but now it's less convenient).