Posts by doublelayer

Azure startup credits don't apply to Claude via Azure AI Foundry, reader finds – after $1,600 charge

Re: Scammers heaven

I agree about who should be handling the complaint. That's definitely Microsoft instead of Anthropic, since Microsoft charged for the service, gave the credit, and wrote the terms of what the credit covers. That's Microsoft's bag. I also agree that they probably should do something to make their customer service people more reliable. If they make them simply say "read your contract" instead of explaining it, it would likely frustrate users who can't get confirmation, but it would prevent problems like this. Microsoft either needs to make that the policy or find a way to ensure that clarifications of contract terms are correct.

While I would also issue a refund in this case, whether they're legally required to is not as clear, hence my original statement. I, like you it seems, think that would be the useful thing to do when there was a clear mistake by Microsoft staff, but that doesn't mean, as the original post argued, that they have a legal or contractual requirement to do that if they feel like charging more.

Re: Scammers heaven

Why is this person defined as the victim in this case? I haven't said I think he's wrong, I've just pointed out why the original argument about contract law doesn't do what the proposer thinks it does, but you seem to have decided that I think they should lose. Perhaps you can argue why they're definitely in the right here? It might help to understand your point. I'm not really convinced by either side right now since those terms were available but he did have a reason to believe the customer support, so maybe you could even convince me, though accusations of victim blaming for explaining what the contract said aren't likely to do so.

Also, the acronym is SLAPP, not SLAP. It's not unique to the United States (the UK is well-known for making such legal actions very productive, especially for slander and libel, for example). And it has nothing to do with this dispute about what terms should apply when the contract said one thing and the company's employee/contractor said something different. I'm sure you have a point for bringing it up, but you could have been more convincing by saying it.

Re: Scammers heaven

It's really not that clear. What was the contract change? The documentation always said that third-party products aren't covered. It was the support person who got it wrong. If the problem is contracts changing, then the original terms, that the credit is usable on Azure products from Microsoft only, are the original ones and, following the unchanging law, this guy is on the hook.

If, instead, you're arguing that companies should have to pay for something if their endorsed staff say it, that's a defensible position, but it's very different from what you said and it is not clear law in many countries. Informal agreements with staff members can be enforced, but it's much more likely to happen if the staff are senior or the negotiation long and provable, with a mistake from a support person usually not getting that far, especially when the question was about what the existing terms were rather than asking for exceptions.

GitHub infuriates students by removing some models from free Copilot plan

Re: Economics lesson for life

Or read the comment correctly and actually believes the statement. I've seen similar things argued. Not (so far) that GPUs aren't subject to development, but that a datacenter packed full of six-year-old GPUs would still be very valuable even if my weird theory that AI didn't make money turned out to be correct. I did ask them why, but since they knew less about computers than I did, they needed me to tell them what else people use GPUs for.

Medical equipment techs beg for right-to-repair lifeline

Re: delays in receiving parts, service keys, manuals, and other necessary repair materials

Both suggestions so far would not work very well and the lawyers would likely make things worse. The reason is that it's not a crime to have a machine that isn't fixed yet. It could be a contract violation if the medical facility has a support contract, and they could sue. But if the hospital doesn't have the funds or has a better use for them than paying a greedy company, the company isn't responsible for that or the consequences of the machine not being available. Patients can't sue the company for charging too much for a repair and not providing the resources for someone else to do it when neither is against any laws.

I did say the lawyers would make it worse, and here's how. In most right to repair situations, the law requires that locks are undone so people can repair the machine without assistance or permission but don't require that the manufacturer support the result; after all, they're not responsible if I try to put something back together and do a bad job. This means that any negative consequences of the repair can be held against the person repairing the equipment or the facility that let them. Which would be easier for a lawyer? Option 1: suing someone you can say disassembled sensitive medical technology without knowing what they were doing, causing harm to a patient who was cared for by the faulty equipment. Option 2: suing a company you can say charges too much for a part which does cost something and good luck proving how much so you can show why the price is too high, with the patient harm being indirect because they weren't using the equipment because it wasn't working. Lawyers go for the easy cases and right now, option 2 are impossible cases.

Right to repair in the medical context is much more complex because it will require much more than it does in most others. None of the things that are required, for example producing and providing methods to verify that the repair was successful and that the machine is still safe to use, are in the laws yet. You have to put them there before any of this will work, and assuming that the problem is easier to solve will lead to lawsuits and likely more difficulty actually accomplishing it.

AFRINIC accuses litigant of trying to ‘paralyse’ it

IPV6's designers don't like the DHCP/NAT thing. That doesn't mean you can't do it. You easily can and many do. Pretending that IPV6 blocks you from doing it reinforces the impression that those who argue against IPV6 are operators too incompetent to implement it. If you have complaints that are actually true, as in things that are blocked or worsened by the design rather than things that work perfectly well but someone opposes, they would be more convincing.

Users protest as Google Antigravity price floats upward

Re: Do people think this should (will) be free?

I would have expected a price increase too, one reason I didn't use it, though the "it produces crap and occasionally decides to delete everything for no reason was a bigger one.

Your response doesn't work well though because Google hasn't done a normal price increase, they've done a bait and switch. Would you feel the same if all the services you pay for suddenly did the same? For example, instead of your ISP announcing a price increase for the same service you currently have, your connection suddenly drops for no reason, then the ISP announces you've run out of netcreds and you should consider a new plan where, whenever you run out of netcreds, they charge you some amount for some more even though you don't know what netcreds are or how many you need to move data, they appear to have invented the concept yesterday, and they won't tell you anything useful? This is not a normal price increase, and users are more justified in being angry about what happened than if they had had a normal one.

Re: Do people think this should (will) be free?

Did you miss the fact that the people complaining already pay for this? They're unhappy that:

1. What they used to get for their money is no longer available at the price they were previously charged.

2. Google pretends it is, but doesn't make clear what you can actually buy and at what price, so even if they are willing to pay more, they don't know how much they have to pay to get what they had.

Is your point relevant to either of those complaints?

DR-DOS rises again – rebuilt from scratch, not open source

Re: Why?

It's a question worth asking. If the answer is "no reason, it makes me happy", it provides information. If the answer is not that, then there might be things the unacquainted user would like to know which convinces them to use it or, if source was available, help out.

But also, the analogy fails. Some people work on old vehicles because they have historical value, so making them operate can let people see how technology existed back then. That quite often is for the general public when they're put on display or maintained by museums or it can be for the benefit of a smaller group. The analogy in this case would be building a new car, though one with few features, and designing it to look like an old car. You can do anything you want, but if you have another reason, it might be compelling to someone else, so they ask to see if there is one.

Chardet dispute shows how AI will kill software licensing, argues Bruce Perens

Re: Back to the future

I was only trying to argue against one of your prongs because the general anti-copyright one is a well-argued subject on which there's little useful to be said. Praising AT&T's monopoly while suggesting the much smaller monopolies granted by copyrights and patents are a problem seems contradictory, and you backed that up with several theoretical snarls such as "the "creator" of knowledge enjoys no general legal rights anyway" (they have copyright and patents, so presumably those are non-general rights, but I don't care what you want to call them) or "The writer has no property rights if we do not grant them any such rights" (true, but tautological, and since we currently do grant them rights, then pointing out that we could change the law so we don't would be a suggestion, not a premise, and boils down to "changing laws in a democracy is possible").

Hence, I skipped that prong and jumped to the examples of support which, as I stated, didn't work the way you claimed they did. Academic tenure was designed for a different purpose and helped the creation of public goods only by chance, whereas direct grants do so by design. AT&T's monopoly was mostly not authorized and eventually ruled illegal, and what few monopolies were actually given by government were not those that led to its contributions to technology which were much better funded by monopolies they obtained themselves and eventually were deprived of by courts. If you dislike copyright, I don't need to argue the point, but I don't think your examples do what you think they do.

Re: Back to the future

Academic tenure doesn't do what you're saying it does. It is granted by a private institution, the university, at their whim, whether the recipient delivered lots of value or did not. The intent was insulating them from threats to the university from donors who didn't like what someone was doing. It didn't protect loads of people who did useful things for the public. Things that did include direct grants for research and incentives for certain types of research, development, or investment.

Copyrights and patents are another attempt to do the same thing by allowing the creator of something to benefit from it for a period that ends and with public access (stronger for patents but copyright has exceptions and requirements that do the same thing). If you think those are mistaken, you can propose an alternative, but academic tenure isn't it. Maybe you have something which will build on the concept, but I haven't heard that part yet.

Re: Prompts?

But using that logic, everything is deterministic as long as you have the power to set everything to exactly the same state. In practice, when people use the words, they're talking about a specific system where they either do or do not have that power. In this case, they're talking about a cloud model which is not available to run on their own machines, where the cloud-run version does not let you set the temperature, and where the cloud-run model has system prompts which change without notice, the user can't see, and the user can't change. That system, as controlled by them, is not deterministic; it intentionally involves random input which they do not have the power to change. It would be deterministic if run by someone with more privileges, but it's not. They could run a different model which gives them the permissions to do that, but then the result would be different. Therefore, when they say that the process isn't deterministic, it's correct from the perspective of that user or anyone attempting to reproduce the results outside a small group of mostly employees of the model makers who have the ability to do that.

Re: Software patents as well?

So? Remove those clauses, change the name, and stick in your own. Now it works again. Nothing stops you making the license you want. Lots of things will prevent that from being accepted by others generally, and the FSF will surely post a warning if anything substantial uses it, but you can still make it and use it. Many have before you; see the FSF's complaints list for several examples.

Linux PC vendor System76 tries to talk Colorado down over OS age checks

You are entirely correct, which is yet another reason I strongly oppose these laws. I never said different. My arguments have been to describe why building from source is not the result of the law nor the biggest problem since binaries would be easily available from sources outside the jurisdiction of the laws. For anyone within that jurisdiction, especially developers but quite possibly also any users or downloaders, there would be problems, and I expect those problems to intensify in the next version of the law. The difficulty of building from source isn't one of them. Harassment from ill-informed law enforcement very much is.

I think we're talking about two different things. My point was that, while these laws stay local, it doesn't require a black market, underground binary system. As long as the place accepting donations is based in one of the 43 states without such a law or one of the many countries that hasn't followed the lead those states, the UK, and Australia are setting, then they can operate all the normal facilities, including a professional website and distribution servers.

You're correct about the harms of this law, and I am worried because I expect that, after those who wrote it find out how ineffective it is the first way they wrote, they'll write a stronger version which will have more of the effects or try to get it passed elsewhere, both of which could have the effects you're talking about. However, it is still worth considering that as New York goes, the rest of the country and world doesn't have to follow and maybe we can use that to our advantage.

But you only need one person to do it for a binary to exist. There are a lot of distributions. There's likely to be several who aren't based in one of those states and don't feel like following this law when they don't have to. That is why either this law will have many people ignoring it without consequences or it will lead to even more invasive ones trying to detect and stop them.

Except what would easily happen in the short term is that someone living outside a place where laws like this apply will recompile it because they refuse to run such a thing, then they'll put that online and people inside affected areas can download that version. What are you going to do? You can't punish the person who made the modification or published it; they didn't break your law because they're not resident there. Either you give up or you come up with another requirement, for example that any ISP block access to software that doesn't follow that requirement somehow, and the problem gets worse.

Swiss e-voting pilot can't count 2,048 ballots after USB keys fail to decrypt them

Re: Time and again

It's the key on those drives, not the votes, they evidently had multiple copies, and nothing suggests the problem is hardware failure. If you write the wrong key to the media, the choice of media is not the problem.

Polish cops bust alleged teen DDoS kit sellers – youngest just 12

A person on a median salary used to have an easier time buying a house, but not without saving aggressively to do so. The need for a down payment is not new. The reason I phrase it as expensive houses rather than wage compression or stagnation is that housing is somewhat unique; other expensive things are as or more affordable to a median income earner than they used to be, but houses are so big an investment that their relative costliness is the most noticeable. The causes are very different, since there are a lot more people wanting houses, the people who bought them on those median salaries still want them, and not that many new ones are being built and those who build them are usually building for wealthy buyers or those selling their house to buy a more expensive one.

What the higher salary buys you is more ability to have what you want. If you want something big and expensive, you still have to do without other expensive things in order to save to get it. Expecting that you can have a lifestyle that spends most of your income and still buy more wasn't any more realistic before, and people have been finding that out to their displeasure for some time. People earning hundreds of thousands have ended up in disastrous states because they felt so comfortable with their high salaries that they started spending on all the expensive things they technically could afford. Why not take a holiday in Japan, and since it's so far from the UK, why not spring for first class tickets there which cost £8,000 each. Then they are made redundant and have the same problem you're expecting for others, but that wound is self-inflicted. It's a lot easier to experience that problem the lower your salary is. Almost everyone could find a thing to do with more money, but as the value increases, the level of hardship they face if they don't have it decreases.

I think you're taking one thing a person can do and assuming everyone does it. I have myself and know others who have earned much more than I needed to to afford my expenses and saved the remainder specifically because I wanted to have savings for a large expense. Yes, I did take a better rental, but I continue to cook instead of eating out so I would have more savings. It's working. People who earn more and choose to spend it are making a choice with benefits and consequences, and they have the freedom to do otherwise.

I see how, in our world where housing is expensive so you have to do that for a while, especially if you have other nondiscretionary expenses like student loans, it can feel difficult to ever afford it. If it feels difficult, I see why some people just abandon the hope and spend the cash now. It's not actually true of the high end of your salary band, who can buy houses if they do what people have always had to do and save large chunks specifically so it will become an option. There are many people at the lower end of your band for whom this won't work, and that's the newer part, but that's because houses are a lot more expensive now, not because wages are compressed, and it makes a huge difference between £30k and £100k.

It sounds like the difference is 230%. Let's do an experiment: let's go to the 30k people and see how many of them think a 230% raise is not a meaningful difference. Then we can go to some 99k people and see whether they'd mind taking the 69.7% cut so those few who think there is a difference can test it out.

There are people who earn more than £99k, and given your dismissive attitude toward a £130k salary in general terms, I think you have considered that, but even within your own arbitrary range, there are very noticeable differences.

That depends on your industry, skill set, and a bunch of other things where we don't know your personal circumstances, but for all of us, there comes a point where we're not likely to improve our salary by switching jobs because companies are just not willing to pay much more for our skills. Perhaps you're at that point, though if you haven't tried, you may not be. So? There's no law that you can always keep increasing salary by switching. If you want more, then it's time to find out what you can do that people are willing to pay more for, most of which is costly in time or other resources to you which is why it's valuable in money to someone else. Maybe it's gaining a new skill, accepting a job with a less pleasant work task, or moving to a different place, but if you really want more money, then you may have to sacrifice something to get it once you've exhausted the opportunities to increase salary for relatively free.

As for your absolute descriptions, I don't think your description of what a salary increase in your field can buy is accurate, but even if it is, what difference it makes is a better room, more pints, or that much more savings toward a larger expense later on. I do wonder how your actual salary compares to the general public's, very relevant to any claims of subsistence level.

The discussion so far has been about punishment levels or the reward of a job, and that's also a valid discussion, but I have a different question for you? What talents? What talents do they have which would be worthy of a job? What job could they reasonably be given if we accepted your premise?

Nothing in the article suggested they wrote software, just that they sold it. The talent therefore is selling stuff to others on the internet, and given the typical buyers of DDOS services, that's not likely top-level salesmanship either. Even if it turns out some of them did write it rather than the much more common finding it from someone more competent, it's DDOS software. You could write that in a few minutes. The ability to write that does not indicate the ability to do anything more complex. Sure, the technical ones could take their skill, learn more about tech, and eventually have skills that a job would pay for, but what makes you sure they had any technical ones and that they already reached that level?

Microsoft Authenticator to nuke Entra creds on rooted and jailbroken phones

Do you mean to ask why people root their phones? Sometimes, it is to use apps that would otherwise not run, for example network blockers that use more direct control of the network stack than non-root allows. Other times, it's less to do with an application and more with control. For example, if you have a device that came with bloatware, having root means you can remove it. Without root, you can at most disable it. Root also lets you back up and restore all your files, not just those apps choose to put into common storage, since one of the places that the Android designers choose to hide from the typical user is many of the directories holding user data.

There are a lot of things you can do with a phone that don't require root, but it's a level of control that people who bought and own devices should expect, so those who want it often go get it.

Re: Hmm, is the data on a personal device Microsoft's to delete?

The authenticator app does not create keys; the remote server creates the keys and the authenticator app copies them. There's not really a law defining who that data belongs to, but if someone tried to answer it, it's more likely they'd decide the server guy does and the phone guy was authorized to have a copy. Combine that with the fact that an owner of a rooted phone can copy the keys before the app deletes them but they probably won't work if Microsoft implements this competently, and the fact that the software doing the deleting is what the user chose to install (even though they didn't have other options if they needed that kind of authentication) I think a claim against Microsoft for this would fail.

AI has made the Command Line Interface more important and powerful than ever before

Re: GUIs designed for humans ?

And not too long after the linear sequences, we found that two-dimensional arrays of those symbols, sometimes with lines and spaces to delineate between chunks was useful, and more symbols made it more compact than less symbols. That's why we have been making things closer to GUIs and benefited from doing so. GUIs aren't about the pretty pictures. They're about showing more information so it's faster to give specific commands. Instead of having one command that lists resource names, one that provides details of a resource from its name and another command that takes action on a resource from its name, the program lets the user browse the resources and see things about them when they're taking actions. In simple cases, that just saves the user having to remember the name and type it accurately (not hard if file names are manually set by users who pick good ones but more annoying when they're UUIDs or that colleague who thinks report-final-Oct2025-updated3Feb2026-draft-complete is acceptable).

CLIs are great for many things where we want automation or bulk actions, but GUIs are great for far more than graphics manipulation.

SETI admits its search for alien life may be too narrowly focussed

Re: Do we know what to search for?

True enough. However, they didn't say that we shouldn't look at radio, they asked whether we have other possibilities to try. Are we searching on radio because we think that's a likely place for evidence to appear or because it's easy? Would we be more likely to get better results if we searched on something that's harder but more likely to carry useful information? There's even a benefit in naming things we can think of but can't detect so we can start figuring out how to detect them.

The other side of the "do the easy stuff first" coin is the "searching under the streetlight" problem*. SETI has limited resources and, at least from every participant or adherent I've heard, really eager to find something. It might be useful for them to add a "where are we not looking and could it be worth starting to look there" research topic to the many different "why didn't we find what we want to" ones.

* For those who have not heard it before, Alice comes along and sees Bob searching the ground near a streetlight for his key. Alice asks Bob where he last had them, and Bob indicates a place quite a way from the streetlight. Alice asks why Bob is over hear, to which Bob responds that he might as well start here because it's much easier to search with the help of the light.

Fake job applications pack malware that kills endpoint detection before stealing data

Re: do not open random files from strangers

The problem was and still is the hiding of extensions. The last extension is always what is used, but when Windows and Mac OS hide these by default, it's harder to see immediately what the file manager will do to open the file. Not impossible as there are still file type description strings and icons, but I think hiding that is unhelpful. I wish I could blame that alone for user gullibility, but I don't have enough faith in users to actually think so.

And on the world of 'ix', the files are indeed not executed because of a string of characters. They're executed on a bit. Which is also set by the user, so if someone opens an archive which maintains file permissions and the person who put the file in there set the bit, the file will be executable. Remember that this attack involves sending someone a container file, in which are other files, and relies on the user clicking to open the container and then clicking a contained file to execute it, then accepting permissions, at least for all the kernel driver parts. It is possible to do that on things other than Windows really quite easily.

Re: This is a Microsoft failure

Ah, an autorun exploit. Those are just unforgivable. Let me get my pitchfork. Though while I'm doing that, you wouldn't mind pointing out that exploit, since in fact there isn't one in this campaign?

And drivers. How terrible of Microsoft to have buggy drivers. We really must put a stop to this vulnerable code written by programmers in Redm... I mean Sautron and Eden Prairie because that's where Adlice and IObit are respectively based, and it's vulnerable versions of their drivers that are installed by the malware, if the user has granted it the permissions needed to do so. But I'm sure you read the report to figure out what they were talking about drivers before making your accusations, because you definitely have a basis for your complaint instead of having a company you dislike and an urge to find reasons for them to be to blame whether you understood the facts or not?

Re: do not open random files from strangers

Because guessing at the proper software to handle a file from a few bytes defined in a database is so much more reliable. Most of the time, it's identical to the file extension method because it uses whatever the creator of the file set it to use. If I had to choose one, the file extension method wins out because a user can be taught to recognize them and they can be easily displayed, whereas guessing from magic numbers requires that anyone manually check those before opening anything, hopefully with the same database the file manager is using.

Neither method does anything about people who don't know what an ISO is. We could either try to train users more or we could try to build software which insulates them from this complication. Let's not pretend we have a solution if we don't, though.

US state laws push age checks into the operating system

Re: An actual solution?

Which is less bad than what they are going to do but breaks in several places, as, admittedly, do most of the worse options. For example, you still don't have much which ties that UUID/password to a person, so there's nothing stopping:

1. A child copying an ID of their parent.

2. A child copying an ID of someone else.

3. A child trading one they copied to other children.

4. A person obtaining multiple UUIDs from different suppliers, if any notary can do this, they should be quite cheap. If they're not quite cheap, we have another problem. Then they sell some of these to children.

Number 4 would likely be quite illegal. The problem we're facing is that people passing these laws aren't willing to stop at responding to illegal situations. They are trying to prevent anything bad from happening, which is mostly impossible, and whenever it is proven that something bad can still happen, they look for yet another technical measure to make it impossible with privacy being the only thing they're willing to give away to make that happen. Your solution will be too much for those who oppose verification altogether and too little for people trying to obtain it.

If you insist on making this a nationalist thing, would it really hurt you to learn what other countries are doing? The UK has significantly more invasive ID verification. Australia has stronger restrictions based on age with much more than "please enter your age" to back them up. Both of those were on your list of people who are supposed to fight against such things with you. Did you miss this, or is your generalized problem with one country causing you to ignore the clear differences in the specific complaint you're making?

You've got more problems. So far, the laws concerned apply to California and Colorado, not the rest of the country, let alone Canada or any other country. That's not the whole thing, as several more states also have laws trying the same thing in different ways. It would be entirely legal for an operating system provider to ask for ages in California and not do so in Canada. The US federal government couldn't even make Canadians do that any more than China could, but also, they haven't tried. You have plenty of things you could complain about that would be real. Why make up a fake one that collapses so much more quickly?

Re: > What "problem" has to be tackled anyway?

Your arguments are either weak or fallacious. Children should be protected from trafficking and sexual exploitation, and both those things are already illegal. How does age verification enhance either? Of course, it doesn't. Therefore it is not helpful in proving the need for, benefit of, or harm of not doing age verification. Your choice to put it first suggests you have nothing better.

Then you argue that we can't question this because some countries have legislated this. That argument is bad in several ways. The discussion at hand is whether there is a problem and what it is, and that is entirely relevant to if, not how. People in those countries are free to question the purpose of the laws that have been passed and to encourage their repeal. Also, we're talking about laws in the United States which do something related but not identical, and people in any of the other states do not have such a law yet. If comes first, since if the decision is not to pass such a law, how you would implement it is not relevant. You cannot declare a question everyone else is debating and you claim to be debating irrelevant just to win.

Brits fear AI will strip the human touch from public services

Re: worried that AI will dehumanize public services

I don't think I agree. The problem I have is that, if I honestly try to put myself in most of the positions alone, I don't think the improvement would be noticeable. When putting ourselves in place, it's tempting to imagine that we have massive power to change policy and take actions, but consider how it works from the real positions where that is limited. Let's consider, for example, someone applying for a benefit for which someone has to meet eligibility requirements.

If I was the clerk processing this, my job is to make it easy for people to apply, right? So I should be human instead of robotic, help people understand the forms, maybe complete them to reduce their paperwork burden. Wouldn't that help? Of course, but it would also take longer than telling people to do it correctly themselves. There aren't many of us clerks. If I do that, I provide good service to a subset and everyone else has to wait for any result because assistance time is inversely proportional to processing completed forms time. Also, that's the person who is faced with people who are trying to defraud the system, and they're blamed if they don't identify and reject them. So I see why low-level employees don't seem very helpful.

So maybe I'm the manager. I can make the system easier so it's more efficient. Let's make better online docs and forms so it's easier for people to complete them without manual assistance. Except I can't hire programmers to build the forms right; I don't have that authority. I can request funding and put it out for bids, which means I first have to try to get someone to write a spec that an outsourcer can implement even though the people doing that aren't familiar with writing specs and then I face companies who specialize in building these things for as much as they can get and are very good at that job.

So maybe I'm a senior director and I'm going to solve this with a big hammer indeed. If we allow lots more hiring, then wait times will improve, they can hire some computer people, and I've created some jobs which is nice. And then the politicians and voters complain about the administrative cost and that there are people who aren't working, so I'm told to cut budgets. And the point is that the politicians and voters aren't wrong, the lack of a big internal programming team isn't wrong, and the clerks not completing everyone's form isn't wrong. Every one of those countervailing pressures exist for a reason, usually cost-efficiency or fraud prevention which leads to cost-efficiency, which everyone also wants.

There are lots of improvements possible, but making them requires getting everyone to want things to improve and to work together to do that. This also runs into some people who don't much care that they're not good; they can't improve them anyway and they're doing just fine the way it is, so they don't help either.

Re: 'There's a naive techno-utopianism in Whitehall'

This is related to many of the problems, but some parts are misleading or incorrect. For example, that lots of captains thing doesn't mean what you're implying it means. In a normal running navy, there would be lots of people with the rank of Captain who are not the highest-ranked officer aboard a ship. In fact, the officer that is commanding HMS Dragon is a Commander (one level below Captain). It is perfectly normal for someone to be ranked as Captain and have duties on land related to logistics, tactics, administration, or review, and no navy could operate without them.

Unfortunately, the problem of reducing inefficiency is not one we have a good solution to. A lot of people don't care, others do care but aren't competent to do things, some do them anyway and make things worse, and making everything work together takes more sustained effort and ability to make changes than most are willing to do.

Google embraces third party app stores and payments to put Epic Games case behind it

Re: … but isn't GrapheneOS built on AOSP?

It is based on AOSP. "Annexing" doesn't make any sense in this context as it is an open source project that Google already wrote most of and anyone can use without Google's permission, but I think I know what you're trying to say, which is that Google is allowing parts of it to atrophy and replacing them with updated versions they don't release the source for. That is causing and will cause more problems later.

Unfortunately, the Linux-based mobile operating systems have another problem: they mostly don't work. The few people working on them are doing what they can, but mobile operating systems are not easy and they don't have much resources. These are still mostly tools for people willing to accept that many things will be worse than every other phone because they're that into open source, and there are not enough to keep most hardware attempts in business which is why Motorola isn't interested in jumping aboard. Motorola had to choose between potential problems later or definitely problems now, and they probably never even considered the latter.

Apple's budget-friendly MacBook Neo is bursting with color and compromise

If this is the first time you're learning that Apple machines are always more expensive than comparable machines, at least if you shop around a little bit, then welcome to the world, new child. Perhaps it's worth it for something Mac OS does better. Perhaps someone is willing to pay for the pretty Apple logo. Or perhaps it's "idiot tax" as El Reg coined. What it shouldn't be is at all surprising.

By the way, the original MacBook Air in 2008 had a 64 GB SSD or 80 GB spinning rust and 2 GB of RAM. That was not enough for many people then, and this won't be enough for many people now, but why pretend that they were the same specs? It does not help make points when it's so easy to point out that you were either unaware of the points you were relying on or you were lying.

Re: Why would it need a SIM?

I've seen many which put limits on the tethering, which can be an important factor. The cheap plans I use don't tend to limit tethering, but they also don't come with lots of data. When tethered, I'm using things that tend to consume more, so those cheap and unrestricted plans will run out quickly, and one downside they tend to have is that data above the cap is more expensive. For example, the plan I currently use reduces the speed to 64 Kbps when the data cap runs out, but even if I wanted to continue tethering and use that speed, it also blocks me and only lets phone-native traffic use that slower option. Then again, if you had a plan like that, having the laptop connect directly probably wouldn't be any better and a better though more expensive plan would probably be the best solution.

Re: Is it 2012 again?

Let's be honest here. Phones with more than 8 GB of RAM exist, but they're far from normal. Plenty of people use low-end phones with 4 GB and they are fine. You need to get to flagship and recent flagship level at that to exceed it. The Google Pixel had at most 8 GB until the Pixel 9 last year. The iPhone didn't get 8 GB until the 15 Pro and the non-Pro 17s still have 8 GB. Most phones do not have more, in fact most phones in current use have less. Laptops are more variable, but there are also a lot of 8 GB laptops around there, both ones you can purchase and ones that are still in use because they're just fine.

I'd want more as well, but you underestimate how mostly fine 8 GB is for a lot of tasks. Browsers will use lots of RAM, but they can also detect when it's low and reduce their usage acceptably, though some browsers do a better job than others. Based on your reference to 2012, I'm guessing you tend to need more RAM than the average, and so do I, but it isn't what everyone needs.

HR may have to cajole and soothe reluctant employees to get them to use AI

I'll give LLMs some credit here, rare for me, but they deserve it. It can help in that situation because they can identify some things from a mass of text more quickly. If you could reject something on finding a sufficiently serious problem, a program which can quickly eliminate a subset of the inputs lets you spend more time on the rest. If you have to identify and point out all the problems that exist, then it's less useful.

Re: Clarification needed

Literally, it means that HR is supposed to join with managers in convincing people to use AI because senior management wants AI used. In practice, that won't help much, because people who refuse to use AI generally do it because:

1. It doesn't work, in which case maybe tech could help by improving it (maybe, let's be optimistic for once), but HR would be useless.

2. The management has no idea what it's supposed to do, and the individual workers haven't come up with anything, in which case HR will be useless.

3. The workers fear that they will be blamed for incorrect results or delays in performance caused by using bad AI tools, and HR could try to reassure them, except that most people wouldn't trust HR's opinion on whether or not there's going to be blame flying around and how they can try to deflect it.

4. Workers fear that successful AI will put them out of work and that there's a chance the AI would be successful, so they're deliberately trying to encumber it, in which case HR's encouragement won't change their mind. HR threats might, but probably not the way HR wants them to go.

Cyberwarriors elevated to big leagues in US war with Iran

Re: Operation Epstein Shield

That is how war works, true. The other way that war works is that, if someone isn't shooting at you and you shoot at them, then they're more likely to shoot back at you. Iran can justify their actions through the laws of war if they want. The question is whether their actions help or harm them in battle, and we've not seen too much help so far.

Turns out most cybercriminals are old enough to know better

Re: Not sure I believe these stats

That's probably true, but we would also probably want to weight by severity to get an accurate picture. I expect a lot of young people launch DDOS attacks; it's cliche for a reason. But most of the time, these are not very damaging and the people doing it have little knowledge (all something that basic takes) or even less (there are so many ways to pay someone else to do something even that basic for you). If we're identifying meaningful cybercrime, I would try to include more detail about the consequences and complexity of the crime, information that any survey from law enforcement probably doesn't have much access to.

AWS says drones hit two of its datacenters in UAE, urges users to move resources to different regions

Re: Is this what took Amazon offline on the 27th??

I doubt it. The attacks didn't start until the 28th in most time zones, and retaliation took a bit longer. Amazon the store is unlikely to be hosted in or rely on the UAE and Bahrain zones unless you live in that area and possibly not even then. If Amazon was down, it was probably something else.

Re: Sprinklers!?

If the building burns down, then the GPUs are probably not in great condition, and even if they did somehow survive, the fire brigade are not going to be worried about saving them when they try to keep that fire from taking out whatever is next door. It would take some work to make sure you only use water when the damage it's going to do is already inevitable, but it's probably still worth doing.

Re: Isnt this a little weird`

I'm not sure why you're surprised. Datacenters are generally built to withstand normal problems like power failure. Even then, depending on what's in the datacenter, they may have an acceptable reaction to those failures, for example safely shutting some things down rather than keeping everything running regardless of how long grid power's out, while other DCs do want full failover on that. They are almost never built for staying active when being fired at. If you have the ability to specifically attack them with military drones, you'll be able to do damage they aren't equipped to absorb without anything going wrong. Rather few things are built to absorb this because they usually rely on local militaries to defend them if a war breaks out and can often assume that wars won't break out because people try not to install expensive things in likely war zones.

SaaS-pocalypse chatter is doomster pr0n. It would be nice if enterprise IT were boring again

Re: Funny

I'm interested to see if that turns out to happen. A lot of places could have avoided many SaaS products, except that maintenance and development were too hard for them that they ended up accepting the SaaS vendors' relatively lackluster versions of both. There was no reason why they couldn't hire some people to build that themselves or at least hire internal teams to develop around some core they chose rather than have the external provider build it for them. Factors other than technical practicality were part of this, and those will likely still exist for the LLM-written version, even assuming that the LLM-written version actually works which is far from proven.

One problem that is likely to be around, though I don't know how much damage it will cause, is that such things often get new parts bolted on. LLMs aren't great at making incremental changes. Even if something works well at the start, it still has to last the ravages of time and little mandated changes to succeed. We have two competitors who are both quite bad at achieving greatness, so there might be a more interesting competition than LLMs normally face. I'd still give the SaaS vendors the edge here because they do have the lock-in factors already. The first step in porting an existing implementation to something new is seeing if the LLM can reverse-engineer the database all the important data is in.

PCs and phones to get more boring and expensive in 2026 thanks to memory drought

Re: What is a PC ?

Most of those don't make sense to me. EFI settings not staying I can see, though it has little to do with Windows, but the fast startup feature hybernating when you don't want it to is not new and has no effect on whether Linux can be installed. If you're erasing Windows, it doesn't matter whether it stored a copy of RAM on the disk before you erase it. That only matters if you're trying to write the Windows partition from in Linux, and switch that to "mattered" because Linux has been able to deal with the unclean NTFS filesystem for a few years now, so even that can be done.

As for function keys that don't work, that's unfortunately common in my experience because a lot of bootloaders only seem to check for newly pressed keys at some stage they don't announce so you have to press the key repeatedly and hope you've hit the right time. I don't know why that is, but that's been true of many computers I've had and is far from new. It's still much less effort than it takes to boot to something else on a Chromebook, some of which involve disassembly or downloading unverified scripts and hoping they work or at least will detect if they don't before they write too much.

Denizens of DEF CON are 'fed up with government'

I don't see anyone suggesting he did not. I see people questioning what level of significance it should have, with opinions ranging from "remorse and subsequent opposition mitigates it somewhat" to "it is unforgivable no matter what happened afterward". Both of those are valid positions in moral philosophy. Neither makes much difference to a comparison which never stated that Franklin was a person of perfection whose life should be copied in every possible respect.

Open source devs consider making hogs pay for every download

Re: Isn't this announcement / discussion

The problem is that the stuff they're talking about doesn't require a login. It's open source, so anyone who wants can have a copy, so no identifying information is collected, so if someone requests a thousand copies, it's not always easy to know who that is or, depending on how they did it, to determine that it was one person requesting a thousand times rather than twenty people requesting a thousand times between them somehow. Throttling can often work, but naming and shaming is often more difficult than the reward would justify.