Posts by doublelayer

Shots fired – literally – over proposal to build datacenter in Indianapolis

Re: Pre-emptive self defence?

"If he'd said the bullets hit near where his child/him/his partner was sleeping, then the statement would have been more accurate, and no less shocking."

Unless he and his partner and his child happened to be sleeping on the opposite side of the building, in which case it would be less accurate but a meaningless difference to the general danger of shots being fired at the house. The relevant part, that the bullets were going through a part of the house frequently in use, remains the same, even though the early morning meant the chance of that part being in use during the shooting was lower. Someone having an early appointment or difficulty sleeping could easily have changed the circumstances in a more tragic direction. I don't understand why you consider this difference to be as important as you do, but your minimization of it, suggesting that you would have to be outside to be at risk for instance, is tripping my own "why is that the point you're making" alarm.

That could be part of it and does seem to have happened at times, but there's also two more options, both much more sympathetic:

1. The politicians misunderstand this and think there are more jobs available. The companies wanting to build them are easily able to claim so, and politicians aren't experts on what a DC needs.

2. The politicians correctly understand this and think the jobs are worth the cost, possibly because they misunderstand the costs for even more sympathetic reasons. A couple hundred jobs is still worth something, and the DC providers are often good at pretending they'll cover all the costs of powering their new site. If they're covering the cost and the only downside for the residents is an ugly building, that is a much lower cost to pay. The fact that the promised payment is contingent on contracts between the DC and the local power provider which actually will pay for all the power costs assuming everything goes well and no plans change any time in the next ten years is more difficult to prove and may rely on information politicians don't have since they're negotiated between those two directly.

And for that matter, the citizens who oppose it generally don't have any more information about whether there will be problems from the proposal. They are likely using similarly limited information and coming to the opposite guess. It happens that, with modern GPU-heavy proposals and move-fast-hide-everything logistics, that guess is more often correct, but the same people object to lots of things that don't have negative effects on similarly knee-jerk reasoning.

Re: Pre-emptive self defence?

Bullets can go through doors and retain plenty of momentum. Your suggestion that the child would need to be outside for the comments to be correct is flawed. The early in the morning aspect reduces the risk somewhat, but by the time we're getting to that level, I think we both know we're not talking about useful differences. If a child as a potential victim makes this any worse, then regardless of the time of day, that is relevant.

Apple's chips are the core of a new landscape, but its biggest win is Windows

Re: If only Linux was as simple...

What use is that comment? If you're going to respond to anyone reporting problems switching to Linux with comments that either say "your problem isn't real" or "I don't have your problem so I can't be bothered", you can't act surprised when Linux doesn't get adopted because anyone who has any problem will decide that it evidently doesn't work for them and abandons it. These kind of problems are very annoying to solve because, even if I want to spend the time on each one, I have to track down the exact combination of hardware and software involved to identify what went wrong or whether it was user error, find the fix, try to do something so I don't just tell each user to run these opaque scripts and maybe it gets fixed long-term, then start again on the next report. I don't really want to do that, so I only rarely try.

Maybe we should just not care whether the general public adopts open source software. That's definitely easier. But if we are going to do that, we should stop pretending we still care and recognize the loss of the network effects that their presence could provide.

OpenInfra General Manager talks sovereignty, governments deploying tech 'kill switches'

Re: Another Puzzled Old Person Here.........................

Those ways are not quite as different as you describe. I don't think there are internet kill switches in Washington, but if there are, why wouldn't they have had them on private lines? Did they suddenly become interested in the ability to take things down in 1995? That's especially true because a kill switch on the internet or big chunks of it would cause a lot more damage than individual ones on private lines, so if they wanted the ability to target individual things rather than black out communications altogether, the private lines were the easier things to mess with.

The old ways are the best ways if you need to and are in a position to operate your systems as if you're going to face deliberate attack trying to take you down and you need to survive that. If that's your situation, then you do want private buildings with your servers and a private line connecting them on private land you own and patrol with private security guards and anti-aircraft weapons. Since almost nobody non-military has all those things, you would make do with the subset you can have. For many others, they don't expect that a government's going to aim to take them out and, if they are, they won't withstand it. If my country's government wants my employer's network to go down, they don't need to try to break the internet or strong-arm the cloud providers when they could march some police into the office and order us to turn things off, so hardening our infrastructure to withstand something we don't expect and can't survive is not a logical argument for doing so.

Researchers didn’t want to glamorize cybercrims. So they roasted them

Re: took long enough.....

I hope it works, but it's much easier for something to be successful when the same guy gets to write the proposal and the results. I'm not sure many of the criminals they're talking about are motivated by notoriety. I think they're mostly motivated by money. Those wanting their own brand will make one themselves rather than waiting for a security researcher to name them. So far, when organizations have given themselves a name, most researchers have used it in order to clearly communicate, but even if they insisted on referring to Dark Side as Pathetic Slugs #39, I don't think that would worry members as long as they were making money and still using the term Dark Side in their messages and advertisements.

It's always possible and it has happened, but there's a reason those we've identified mostly haven't been. If you're already employed as a security professional in a big company then you have a relatively stable and well-paid job. The theoretical rewards of crime may be higher, but those experienced in security will also be those who have had the most opportunity to see that a few people in the ransomware game make millions, many make less than a normal job would, and many make nothing at all, so the chance of getting that criminal payday aren't as good as they appear. Also, if you're ever caught, you lose your ability to keep having the security job you started with, so the risk is higher for someone who has one than for someone who thinks they won't get one.

If an AI agent screws up while running your business, there's nobody to sue

Re: Strategically placed scapegoats

So did the individual software engineers pay fines? No? Fine then, did blaming them mean VW didn't have to pay any fines? Again, no. Whether you think the external consequences VW got were sufficient or not, they didn't manage to get out of them by blaming the engineers. Of course they gave it a try, but like most things when it gets that far, it did no good. The problem is getting things to go that far in the first place and, when they do, making the penalties strong enough that the perpetrators don't repeat them.

Re: Bias will come from the training

Even if everything was perfect, that wouldn't be enough since most of the information about whether to hire someone is from the interviews, not the original application or CV. An application can get past the initial CV screen because the candidate lied, stated what they did in a self-promoting way that misled the screeners into thinking they'd be more experienced than they were, or accurately stated all their qualifications but demonstrated themselves to be a nightmare to work with in the interview. There's no way to put all the information into a model, which makes it virtually impossible to confirm whether it's done the right thing.

If AI CV sorting is going to happen, the only stage I'd suggest is testing whether a human would have approved it for an interview, not whether the candidate would have been hired at the end of the process. That removes many of the parts that can't be tested. It still has all the problems you listed and several more, for example that human reviewers have plenty of opinions on how to sort applications which aren't always correct or that lots of experience in Java applets and Silverlight is not the asset it would have been decades back.

Re: Bias will come from the training

It's slightly different from the above summary, but one study demonstrating this is from the University of Chicago article in Nature, summary. In some ways, this study is even more demonstrative of the problem since the text had been removed, but the model was successfully linking by institution anyway and using that to predict results. Machine learning is hard and can look easy to people who don't want to check whether their model is accurate on anything except the data they trained it on.

The normal system starts as you describe it. The customer goes to the provider for redress, then the provider can, if the manufacturer of their tool made a defective one, go to them in turn, up the chain until the original source of the problem pays or someone decides they don't care. That's why LLMs break down, since the people making the tool know liability will be extreme and don't want it, but fortunately for them, they have the ability to specify their product in a way that it becomes difficult or impossible to prove it defective. It would make sense for companies to be cautious about deploying it in that situation but many of them don't seem to know that, hence the warnings in this article.

Re: Strategically placed scapegoats

All of that is true. In the situation of a widespread problem with severe consequences, having a scapegoat on AI checking duty won't save a company which will remain liable for civil and criminal penalties. Scapegoats are tools to protect existing leadership from internal consequences, not the company from external ones, though those scapegoats might get some external consequences as a result of this process anyway. For example, a person who approved an incorrect AI diagnosis leading to medical problems might exist because the management want to escape blame for having put an unreliable piece of software in place, but they might also suffer career limitations when that's the official reason they've been fired.

However, from the perspective of the public, there is a change going on, because things have to get really bad for those external consequences to start. Companies don't get public scandals when they've substantially harmed one person. It takes a pattern of harm to get the attention of law enforcement, politicians, or legal support organizations, the places most likely to be able to do something against a large company. AI decisions make that pattern harder to detect since they're unpredictable and will fail people in random directions. Those who make LLM products are fully aware of that, so they will certainly avoid taking liability themselves. Companies buying them often don't know that and don't do the proper testing to find out, so they make a nice shell insulating AI companies from consequences. I think we're going to have some negative results out of this.

Forking frenzy ensues after Euro-Office launch sparks OnlyOffice backlash

Re: Pretty sure that

Ah, I understand now. They say almost exactly the same thing, but it's okay when the person you like says it. They are both saying it for the same reason: they have implemented compatibility with the other format and think they've done it correctly, though that's a much newer experience for Microsoft as they clearly didn't want to as much as LibreOffice wanted to implement docx support, but they don't want to guarantee that support. It makes sense in both cases.

Re: Pretty sure that

"MS Word complains when you save, saying that "some features aren't supported", which is a truly dark pattern when you opened an ODT in the first place."

How does that compare to what LibreOffice says if you open a Docx and try to save it as one:

Non-standard file format

This document may contain formatting or content that cannot be saved in the currently selected file format “Word 2010–365 Document (*.docx)”.

Use the default ODF file format (*.odt) to be sure that the document is saved correctly.

That seems very similar to me, and it's equally global, with no scanning to detect whether there is any known compatibility problem. In my experience, there very likely isn't a problem as I've been sending people docx files from LibreOffice for years without a single complaint, but I don't use anything very complex.

Re: Spreadsheets????

"I've often wondered whether MEGABYTE sized Excel spreadsheets using VBA could

have produced PLAUSIBLE BUT INCORRECT results which brought down huge

corporations!"

Of course they could, as could bad database contents or mistaken programs. Usually, a huge corporation has so much inertia from their size and markets that damage caused by such things can't kill them very quickly, and in the intervening time, either someone finds the problem and fixes it or the reason for that bad program ceases to exist and the problem goes away. I think we can say with some certainty that computer problems have caused several large failures, some of which exacerbated a weak company and took it closer to the edge.

But also, how would someone prevent that? You can have more review to try to catch it and better tools to make it less likely, but in the end, it comes down to someone making an important mistake in an organization with tons of people doing lots of things where mistakes are possible. There's a pressure system between the desire for more redundancy to detect and prevent mistakes and the desire for efficiency of not doing everything multiple times and checking and cross-checking everything, which will undoubtedly continue bouncing back and forth for the remainder of human organization.

Re: OnlyOffice quite clearly is intended to be proprietary software

I think the most likely result is a decision that 7B's "author attributions" doesn't extend to logos, but there's a chance this leads to yet another Neo4J-style lawsuit about what "further restrictions" are. I can imagine the next three years of judges having to argue about whether a logo counts as an attribution or not. Unfortunately, the more terms a license tries to add, the more opportunities for vagueness and contradiction they introduce since it's not obvious what kinds of attributions can be mandated and, if someone decides it does count, it would be difficult to define something done specifically approved by one term of the license as a further restriction not included in that license.

Mandates can cover only specific areas. Some EU proposals and many online comments encouraging them have suggested having single software standards that would apply to EU central government and member state government, likely also including private companies that work for them. The EU would certainly not mandate that everyone there use it, but deciding that all those governments will use or strive to migrate to a specific, supported variant is also a mandate and one that would have large effects. The more important part is the proposed funding to projects critical to the sovereign computing initiatives which, if they actually do it, is a lot of incentive for whichever projects happen to be chosen and is almost certainly a one-takes-all thing per sector because separate grants to every competing office program, even the ones that are forks of each other, is so inefficient even politicians would notice it.

"Having things under your jurisdiction is a key metric of sovereignty, so yeah, it demonstrably would result in 'more sovereignty' if the developers and all development activity took place within the EU."

That depends what they want sovereignty to give them. If we use your definition, then tautologically they'd get it, but that doesn't prove that it did anything they wanted. There can be plenty of sovereignty of the important parts, where your data is and who controls your systems, when you run software with local administrators on local machines even if that software wasn't written entirely locally, and more if you have the ability to continue maintenance and improvement of that software locally if you need to.

"if it also happens to increase competition in the office productivity software space, that's not a bad thing"

Fragmentation tends not to help with software quality. There are lots of projects which started as forks of something else and ended as forks of something else because the people forking it wanted their name and their donation button but didn't do all that much to change the software. If EuroOffice turns out to have a bunch of developers who didn't work on office software beforehand but now are busy building new features, you might be correct. Do you see evidence that they have? From the perspective of the EU governments, quality and speed of a patch might be more important than the dev who wrote it being Canadian.

I can only speculate, but a lot of this looks like people expecting that somebody is going to pass a law mandating a single project, likely involving a bunch of users and funding going to that project, so they're trying to position themselves to be the recipient. If that involves getting rid of people outside the area so you can claim that you're the most European, then why not; politicians probably don't understand the harms that causes, and some of them think that's a positive because surely, not having any developers outside the countries in the plan means there's more sovereignty...somehow.

'Uncle Larry’s biggest fan' cut by email in early morning Oracle layoff spree

Re: gliding to to a long lonely retirement

One possibility is that they had kids, but now that they're nearing retirement, they no longer have all the expensive parts involved in raising them. They could still see their adult kids in their retirement.

If they don't have them, then they decided that a while ago, given the close to retirement aspect. It might make people more lonely later, but I would want a better reason to have children than for the purposes of entertainment during retirement, and some people who don't have them have strong friend networks or large extended families which can also provide it. Everyone has a different reason to or not to have children, so I'm not surprised that one person's decision won't work for someone else or that some people will eventually regret the decision they made.

Re: Yep

The admin costs may be small, even then only in theory, but the absolute costs are enormous and there are enough corner cases that you'd need extra programs, almost as many as we already have, to cover them. That's without considering any possible economic reactions to such a program. If you think the benefits are worth those costs, you can make that point. Either way, don't pretend the system is easy just because you like the idea.

The company's biggest security hole lived in the breakroom

Re: I do doubt the story

Why is a cat5 jack in the room so unbelievable? They'd be used for WiFi APs or, if the room was also a group space, a computer running a big screen for presentations, so are frequently put in a lot of rooms. They're also easy to add when installing plumbing for the machine, and if the company has bought an expensive internet-connected machine and thinks they should connect it, they could easily provide a port. Now that we have one, putting it on the secure network is a mistake in the switches or configs, and it sounds like they had plenty of mistakes in that area.

Re: I dont doubt the story...

Usually, because a dev building the thing wanted a shell to see some logs or manipulate things to test that something worked, so they enabled an SSH server, and of course that's only for this prototype in pieces on the bench. Then someone, possibly the same dev but it could be someone else, took the dev's image and deployed it in production. You also have the occasional person who is much dumber and thinks that Linux doesn't get hacked, and if the customer ever breaks this then we'll need a way to go in and fix it and they deliberately put it in, but in my experience, leaving that in by accident is the much more common reason.

Re: I dont doubt the story...

Option 1: The machines are connected to a manufacturer's network and that has been breached, allowing the attacker to hop from there to networks of users of that machine.

Option 2: Malware on another computer found this machine and the attackers established persistence on it so, even after their initial vector went offline, that machine could still be used.

Option 3: The machine uses UPNP or some other method to obtain a public address or port, and the secure network wasn't very secure, which wouldn't be much of a surprise since someone's putting coffee machines on it.

There are plenty of plausible possibilities other than an internal threat and more plausible ones than that someone managed to guess an IPV6 address to find the SSH connection which is hard to do with the size of IPV6 address blocks.

AI server farms heat up the neighborhood for miles around, paper finds

Correct. I was trying to explain what this argument about heat islands was about. Of course, those who think all people studying climate have colluded or are deluded probably have explanations for always-rural stations too, but I am not one of them so I'll let the next one try to make them. Sadly, so many stations with the longest history of reliable measurements are now near something kind of urban so many studies of climate can't just ignore them without losing detail that provides a lot more information.

The problem is that, when people compare temperatures from the past and the present from the same location, there are various reasons for the collected temperatures to have changed. If you had a temperature measurement station in some farmland in 1920 and kept it there all the way through to today, but today that is a suburb rather than a field surrounded by forest, then it would have increased for local reasons rather than for climatic ones. This is known, and data from these stations is decreased by an amount calculated based on factors believed to correlate closely with those local changes. Of course, with any adjustment of data like this, it has lots of room for people to argue it was done wrong, either because they think it actually was done wrong or because they don't like the conclusions and want to throw doubt on the research.

We know what day it is but these Raspberry Pi price hikes are no joke

Re: 3Gb ram ?

It doesn't sound like the interactive exhibits are products that need to be frequently reconstructed. In that case, the only requirement is a computer of some kind capable of running the software, and there's no need to match a certain model. When building a bespoke system that there's only going to be one of, then there's little need to produce a rigid set of required hardware and advantages in not doing so like being able to replace or upgrade components when they break rather than requiring exactly the same thing.

There are plenty of valid reasons to use a Raspberry Pi. Continually narrowing things down to a requirement that says "it must be a Raspberry Pi, therefore your alternative is unacceptable" doesn't respond to comparisons considering the many cases where that is not required at all or it isn't required yet, comparisons those opposing the Pi are making since they're not yet committed to it. If someone is looking for a thing to run a DNS filter on, pointing out that someone else has designed an entire product range around a Pi and therefore can't just replace it does nothing to prove that they too should use one or that their argument about whether it's competitive is wrong. If you're trying to argue that it is wrong, you have to consider the situation they're arguing from, and one of the axioms there is that there are multiple options that could work and they're trying to optimize among them.

Re: Love

"I don't use them for general use so I don't know why people compare them for that."

I'd have thought it was obvious: some people do use them for that, likely the people doing the comparisons. There's a reason why "general use" is called that. That can be a reason why the comparisons aren't useful to you, but it doesn't make them invalid or really argue against them. I'd note that the keyboard 500s and 400 are especially likely to be used for general use and there are plenty of people who have those.

Many of the things I do with my Pis could be done with equal ease on a laptop with nothing connected to the GPIOs at all. I use the Pis there because I don't need as much computing as a bigger computer could provide, I like the silent and low-power operation when running all the time, and I have them already. Those are reasons I chose those even though I could have gotten more computing for a similar price (though my experience with secondhand machines isn't as good as those making these comparisons). It is a valid comparison I made when choosing and I chose the Raspberry Pi after doing it.

Re: Hopes vs reality?

The long-term example with computers is the same thing that most prices did in the initial stage of technological development because of extreme competition and fast improvement. Using the price trends from the 1990s isn't a good basis for predicting things today. I assume that RAM prices will eventually go back to normal and that Raspberry Pi will bring theirs down to match, and it helps that there are mostly only those two steps required since there are fewer parts in a Pi than there are in other computers. That's not a guarantee of either, and it's certainly not a guarantee that others will follow suit.

For example, I don't expect Apple to decrease their prices even though they claimed RAM prices is what made them increase them. Apple has been raising the price of the base-level iPhone for some time, and I expect they'll keep doing so until they think they're losing market share from doing it which might be why they decided to do a cheap Mac this year. Perhaps they could have, and if you compared the last three decades of Apple machines' prices then they did at various points there, but neither convinces me that they will again soon.

Re: 3Gb ram ?

Why would they know or care? I'm not sure what the interactive exhibit is, but I'm imagining a thing in a museum. So as long as the computer it's running on fits into the space for it, what does it matter whether the computer's new or old? The professional approach is making sure it's not going to fail, not buying new for the sake of saying so. If you bought a damaged system whose SSD is going to fail any minute, that's a problem, but in the exact same way that it's a problem if you build a Pi-based solution which is going to destroy the cheap SD card you put in it. Use components you don't expect to fail and either can be as professional.

Raspberry Pi leans into semiconductors as sales climb – especially in US and China

Re: Not too worried here.

Not quite. The available subsets I can buy are:

A phone: computing power is great, power consumption is great, software is somewhere between bad (unrestricted Android) and very bad (Android if I can't unlock the bootloader).

Raspberry Pi (also includes everyone else's SBCs): Computing power is pretty good, software is great, power consumption is bad.

Microcontroller-based devices: Power consumption is great, computing power is very limited, software support is adequate (E.G. it took Raspberry Pi years to get Bluetooth working on the RP2040, whereas on a normal Pi, I can use plenty of Bluetooth stacks if I need that).

I have things I'd like to do where a phone-sized device with phone-style battery life would be useful, but the only devices I can get that meet both of those are phones whose software is nothing like the convenience or quality of that of the Raspberry Pi. So far, I don't have a solution to this want, but if someone did build it, I'd like to buy it.

Re: RPi netbook when?

"Of course it’s measured. How else would I know?" Estimation, which is what both your comments suggest with the "seems to be". I get about ten hours off my laptop battery, judged from when I generally have to plug it in again and assuming I've correctly estimated how long each use session was and added them correctly, but if I wanted to actually measure it, it would take recording those start and stop times precisely and likely documenting what I was running on them, the way that the reviewer did in order to precisely report a 5H 59M runtime which I rounded to 6. There's no chance they miscalculated that, whereas if I guess I used the laptop for about two hours but it was actually 83 minutes because I overestimated, my anecdotal value could be off. That's the difference.

"Of course Pi isn’t designed for portable applications, but it really isn’t relevant."

That depends what you want to do with it. It's not relevant for what Raspberry Pi intended people to do with their first SBCs and they still mostly build new ones for the same general types of uses, which is why they built it that way. So from their perspective, it is indeed not relevant. If you choose to use it off a battery, it remains relevant to you because you're trying to accomplish something else and will have a harder challenge doing so. Pointing out that the Raspberry Pi was intended for another purpose so wasn't built with that in mind is one thing. Arguing that enthusiasts can't spend much time away from wall power is weird and pointless.

Re: Not too worried here.

It is not totally correct. James accuses Pete of "just look[ing] at the flagship products and think that's the only product Raspberry Pi make", when Pete's comment makes it clear that they know what other products Raspberry Pi make. Therefore, James is incorrect about what Pete thinks the products are, whether or not we agree with Pete's other opinions (my initial thinking is no, but there's not much clarification of why Pete believes what he does so I don't even know that much).

Re: Not too worried here.

Good ideas, and I'll clarify what I'd like. You can currently power the Pi from a battery, but its power consumption is much higher in all possible states than anything of the same capacity for computation. A phone tends to have a much faster processor, on which it runs a lot more software, and connected to more active peripherals and also runs on less current. That's partially due to optimization in software, partially the use of chips intended for mobile devices rather than chips for mains-powered boxes, and a few smaller factors. That makes sense for SBCs where they intended them to be used as desktops or servers, so a cheap chip from a TV box would do that fine.

I would try to just run the software on a phone except the software limitations make it harder to do many things with one; even a rooted Android device takes a lot of hacking to get anywhere close to the openness or configurability of an SBC. I already know that a device I can program with similar power to a Pi which can fit, battery and all, in my pocket is possible, and I can envision plenty of ways to use that power and portability. Now I'd like to buy that, but am only offered subsets of that in the hardware people sell. Since I couldn't buy it, I tried to build it, but all the SBCs around which I could try were Pi-like in power consumption. That meant either a much larger battery or much more frequent runs to recharge it.

This is possibly a somewhat minority interest. I've seen a few attempts and tried a couple myself, almost all built around the Pi itself or the compute modules, and all of them alluring but nonetheless limited by the same lack of power management. You can obtain something with long battery life if it uses a microcontroller, but of course that comes with limitations on what can run in that lower level of resources. I'm sure others have different revolutionary concepts, some of which would open other opportunities if they could be achieved. It strikes me that there was almost no revolution in the hardware of a Raspberry Pi SBC - the chips were common and used all over the place and the software wasn't new - except that they actually made, sold, and supported it to great effect when others didn't bother. That's kind of what I want here in a slightly different area.

Re: RPi netbook when?

That's cool. I don't have one, so I had to use the numbers from the review, but just to check, is that 7-8 figure your guess or measurement, because the review's number was measurement so I consider it far more useful than an anecdote.

Even if it is correct, presumably from better power management in newer software, it is still not the kind of impressive low-power operation the original comment was hoping for. A normal laptop can run for longer on the same size of battery, and modern ARM laptops with processors optimized for it can significantly exceed that. The Raspberry Pi's hardware and software were not designed for battery operation and it shows. I've used and made battery-powered devices out of Raspberry Pis from the first model B to the CM4, and one thing they all have in common is that you need a lot more battery for them to work than a comparison to other computing devices would suggest.

Re: Not too worried here.

It's also amazing how many people seeking to defend the Raspberry Pi don't notice that the person you're accusing of just looking at the flagship knew and mentioned the existence of the other models in their own post: "For those who want to build simple household devices the ESP32 covers the requirements for $5 instead of $100. As does a pi zero 2 or its many clones."

That doesn't make them right. I don't entirely know what they think Raspberry Pi should do or have done to avoid the "lost impetus and direction" they're complaining about. Your response to them, though, argues they're not looking at things they did and is therefore incorrect.

Re: RPi netbook when?

I don't think you're going to like it, but it exists. It gets 6 hours of battery life under low load with a fresh battery because the Raspberry Pi is not actually a "low power mobile device". Sure, I kind of want one, but not for any valid reason.

Re: Not too worried here.

I would like people to try for a revolution if it could be achieved, and so should you because it doesn't mean the old thing ceases to exist. If they invented a new form of computer the way they can get a significant chunk of the credit for the low-cost SBC, but that new form wasn't useful to you, you could still use the SBC.

To be more specific, something I want might count as a revolution or just evolution in another direction: I'd like an SBC that can run off a battery productively. Pis have been evolving in the direction of a lot more power consumption, and even the first ones consumed lots of power. Back when people were trying to put a Pi in a laptop, it would get two hours of battery life on a battery that could run a low-end Intel processor for eight, and a laptop is far from the most constraining environment for batteries. I think portability would be a useful addition to this area, but the point is that any innovation, whether in that direction or any other, has the chance to be very useful to a lot of people and reproducing Raspberry Pi's success when they invented something quite new.

Apple's last tower topples… and the others will follow

Re: Bus slots

Someone needs all of that, but so few need most of it that it's not very likely. A bigger machine with HDMI, ethernet, separate audio in and out, yes, people need that and enough of them that you can actually buy it. Parallel ports on a laptop? So rare that someone wants to buy that that nobody makes it anymore. Same with PCMCIA because nobody makes anything even slightly new that uses it so, unless you absolutely need something old which used that method, it's a completely useless slot which takes lots of space and internal components. You can hope for that all you want, but it's not surprising that you'll never get it.

Re: So, Apple Silicon

I argue that, if the internal disk is broken or inadequate, then an upgraded internal disk is, unless it has to cost a lot more for some reason, better for every case. There are some people for whom an external one is acceptable and some for whom it's not, but I don't think you can find anyone for whom it's better. The cost example does come up, for example someone who uses a cheaper spinning external drive and doesn't need all that space inside their laptop, but that's as close as you can get.

For example, let's take that person who bought a new machine because her soldered disk was too small. She needs more storage to store large scientific software packages. A network disk isn't going to be good there; pulling 6 GB of code and resources across a campus network or occasionally an ocean whenever she wants to open some research data isn't going to be enjoyable. Fine, an external SSD then. And loading that example would work for that, but not so well for another piece of software which isn't a self-contained application bundle. That one wants to put some stuff in a hard-coded path on the root volume, some things in /Applications, many things in /Library, and a few things in ~/Library. External disks don't hold any of those. I could try to set up something with symbolic links for this application, but that would be tricky, it would be fragile, and neither I nor she wants to have her run to me when something breaks one of those and important work can't be done and she doesn't know a lot about filesystems and symbolic links. That's why people benefit when disks can be upgraded.

Re: will someone explain in simple words!

So this is a "Macs don't get viruses" again? I assume that's what you're trying to say, although technically, "I have yet to see an Apple product unless with a Ransomware Message on the screen!" means the exact opposite. And in fact Mac OS is as possible to install ransomware on as any other operating system. Both LockBit (because they were big and targeted lots of things) and North Korea (because they focus on cryptocurrency researchers and operators with whom Mac OS is popular) are well-known for building Mac OS versions of their ransomware.

Don't open that WhatsApp message, Microsoft warns

It won't work unless you've first set up one of the mobile versions, but you can use it from the desktop clients as well. It's surprisingly common for business communication internationally, and I'm guessing that's why the desktop version is used so much.

Re: Employee training tops the list ????

What "flaw of directly running programs arriving by email" are you on about? There isn't one. The sender has to trick the user into running them manually.

And options to prevent people executing them manually? Yes, the IT department can do that and have been able to for years. Not Microsoft, the local admins who have the control over what users are allowed to do. Either they haven't because users sometimes need to execute unknown things, in which case you have to train users, or they haven't but have no reason not to, in which case you have to train IT employees.

Surprise! Big Tech has been a bit rubbish at enforcing Australia’s kids social media ban

Re: Who’s your Daddy?

One thing that doesn't help with that is no information about what counts. It's all down to whether a regulator decides at a later point that whatever they did was reasonable or not. If companies are supposed to guess at how to do this and aren't given acceptable methods, then you give ones with good intentions many opportunities to mess up and ones with bad intentions lots of opportunities to hide, especially as any claim of unreasonable is subject to extra appeal in the courts if they want.

The article provides a few examples of things the regulator didn't like which are, at least in my opinion, completely reasonable. For example, when someone claims an account from a third party is of a child, I think it's reasonable that you ask the claimant to prove their qualifications to say so. The alternative is that anyone can take down any account by making an accusation without being asked to demonstrate either that what they claim is true or that they have a reason to make the request. That is far too easy to abuse. If Australia wants this locked down, they need to decide how, and they likely haven't yet because they know the truth, identifying every user of any covered account, will be unpopular.

The first thing vibe coding builds is confidence it will help you succeed

I like Python for many things I write, but for something I was going to use an LLM to do, I'd want as much compile-time checking as I could get. The reason I like Python for some things is that it saves me implementation time for many problems at the cost of removing safeguards against me doing them wrong, and since I expect the LLM to do it wrong and it can write stuff much faster, those are both disadvantages with it.

So you chose to vibe-read this article? Read the first couple paragraphs then predict using your mental model what it was going to say, thereby missing sentences like "Is it disruptive in the sense of spelling the end of developers? Not at all." and "It worked … until it didn't."? The result being about the same that vibe-coding tends to give: a comment that looks like it responds to what the article says, but in fact does not. If you want to argue the author is a halfwit, read more than half the article and then you might be able to do so. I see plenty of actual arguments in here that I partially or totally disagree with and predict you might as well. They're all after you couldn't be bothered, and they're all more important than the initial version of the Norwegian example which gets more details later on in the same direction.

BOFH: Are you ready to raise our expense account limits now?

Re: Tip yourself - on a Friday

They don't want incrimination except as a backup, since if this guy goes down, it would be because his actions were detected and that would mean their extra access would be reset. No, the little bribe is useful both to make sure he won't tell anyone and that, if they need anything else, he will have more of a reason to provide it so they won't have to go through the whole blackmail thing.

Senators want datacenters to come clean on power consumption

Re: slightly off-topic...

Essentially, you seem to be describing a nonprofit organization setup, possibly with even more restrictions. That could work, but at that point, why do you expect it to be much different than a government doing it? Either way, there are people employed at a fixed salary told to make this thing work. It can also still go wrong in a combination of the ways government-run and privatized systems go wrong. The leaders can't increase their pay or sell something off and pocket the proceeds, but they can still be treated to nice things from potential suppliers or find reasons that something they want is valid expenses. The only benefit I can see in this structure so far is that, if the water service is profitable, the profit remains earmarked for the water system and can't be easily taken for other purposes.

Re: Don't they have smart meters?

It's an easy number for the DC operators to measure, but the suggestion being countered was collecting some of it from the power providers instead or while waiting, which is where total consumption becomes impossible to measure and rough consumption becomes harder.

I'm not sure the number helps a lot because, if it grows for some reason, the report will come after the additional load starts increasing prices. It would make guessing where that might happen more reliable, but unless there's a plan to do something with that guess, that's not of much help. Currently, not only is the pledge voluntary and unenforced, but it also only covers theoretical use and spending which will get changed whenever profitable.

US foreign router ban criticized for being ‘industrial policy disguised as cybersecurity’

Re: Nobody believed for a second this is security...

Yes, at least in the Huawei case, they had a specific government and law they were targeting, and the hardware was in much more significant areas. It still wasn't convincing, but the surface argument was much closer to plausible. A blanket ban on all companies from all countries doesn't let you tell that surface story.