Posts by doublelayer

SpaceX accused of firing employees critical of free speech fan Elon Musk

Re: Don't get this confused with free speech.

I didn't disagree with that. In fact, I was using that as part of my statement that there are always consequences, and that the free speech legislation in the US does not prohibit companies from using employees' speech as a reason to fire them. What I asked about was why the original poster thought their opinions should be excluded on a free speech basis, but that the people in this article don't qualify for the same. I think the law doesn't give an exception to either, which is internally consistent, but their post did not have a similar consistency, nor even an argument for why their exception should be specially excluded from consequences.

Re: Don't get this confused with free speech.

So let me see if I understand your statement correctly:

If someone says something that you agree with then it's a free speech issue and you should be allowed to say it without fear of losing your job. If it's something different, which you probably disagree with, then it's not a free speech issue and bring on the consequences.

Sorry, if your boss can fire you for one thing you said, they can fire you for other things as well. Consequences always apply. The only exceptions are when labor law restricts them, which could be in everything or not. In the U.S., those laws are not strong, meaning that you can be fired for lots of things you may choose to post. This means that I don't expect these workers will get anything in this case, as they can be fired for the statements their managers did not like. It does not mean the views you prefer would or should get to bypass this reality unless the law is changed. I'm having trouble seeing your distinction between what counts and what doesn't other than you agree with one and don't with another.

Google to start third-party cookie cull for 30 million Chrome users

Re: What's the betting

I'm sure you can purge the advertising cache files every time Chrome creates them, but that will just cause Chrome to recreate them every time. With a 10 MB limit, it's likely those will be hideously compressed on disk to fit as much as they can in there, so expect to be downloading 10 MB per advertiser when it restarts. The good news: it will probably be different advertisers each time.

This is a good enough reason not to use Chrome if you can't disable this, and I already had a good enough reason not to use it.

What comes after open source? Bruce Perens is working on it

Re: Better idea: Go nuclear

That wouldn't happen. If everyone got the source code to every application, the freedom to share being absent would not stop people from sharing it anyway, nor would it prevent other groups from making "competing" versions that are pretty much just copies. If a country tried it, they would simply not get any direct sales. People would still be using Windows, but they'd be using Windows imported from a country next door or using pirated license keys. The same is true of most other software. In practice, we all have the ability to reverse engineer software we have, and we have some rights to do so, but the practical difficulties doing that make it useless to most people. After all, what reverse engineering tasks are you capable of completing but currently blocked from doing by legal obstacles? I imagine that the only significant legal obstacle is with distributing the result, but there are some practical obstacles as well.

Re: MIT/BSD anybody?

It's true that Linus didn't have the ability to just switch everything to GPL 3, but he could have done something to make the terms apply to some important part and apply the restrictions to most users. Based on his comments at the time, it seems he didn't have a major objection to the tactics that the FSF wanted to prevent with the GPL 3, so he didn't try to do that.

Re: Eclipse Versus Borland Delphi

"Companies don’t need Open Source if they want to cooperate."

True, but it helps. They could form a consortium and build something closed. That consortium could even sell licenses and that might help make some money. But it makes it really hard for new companies or individuals to join and start working on it as well, and it reduces the number of users, which can be important if you want contributors. They appreciate that it's free as well, but part of the reason that it can be is that there are so many people interested in maintaining it.

Re: Let's say he creates this post-open "contract"

It's not about willingness to pay so much as knowing whether or not I have to. I don't object to proprietary software; if I want it, I will and have paid for it. The companies that employ me do so all the time. The charities I used in my example do so less frequently because, as I said, they don't have a lot of money so asking them to spend five figures on some enterprise software is a hard sell, but they still do it if it's important enough. However, I know that when I suggest the option. I don't say "Let's use Office365" and get surprised that Microsoft wants money for it. Using something sold as open source and then finding out that the license is not compliant is a much larger surprise, especially if I've previously contributed code or donated money to that project.

One of the reasons I might adopt open source software instead of proprietary is the freedom it offers me, and I don't just mean the freedom to read the source and modify it in my environment. Chances are that I won't be doing a lot of that for the infrastructure software anyway. One other freedom that I value highly is the knowledge that, no matter what the authors did, I can keep using this. If the authors gave up and stopped developing it, I still have the software and the right to run it here. If the authors started demanding a higher price than I can pay for some feature, I still have the old software and can work around this change while using what I already had. In either case, I can take the code, make my own version, and work with others to keep the software alive despite it. In an environment where certain uses require permission or payment, I no longer have that guarantee. If the authors have the choice to tell me that my use is now forbidden, then I have lost something that traditional open source provided.

Re: Let's say he creates this post-open "contract"

I doubt it. The ambiguities in all the licenses that implement something like this have already caused problems for projects that switched to them. While I have argued and still believe that the switching was instrumental in some of the chaos, the licenses themselves haven't helped. For example, there's usually some restriction on commercial use, but the license isn't clear what it is. I sometimes volunteer for charities to write software, but that is not necessarily commercial or not. In order to find out whether it counts, I probably have to negotiate with whichever company has written the license to see what they think, and since I'm not likely to pay them much if at all because it's a charity and they're not very wealthy, those people are usually not very helpful if they respond. If I come across a project like that, am I really motivated to contribute to it directly when I'm not sure I can use my own work later? Usually, the answer is no. It's hard enough to find developers who want to work on open source projects, and for reasons discussed in other comments, I have no faith that I'll actually get rewarded for doing the work, so the least I expect is the right to use my own contributions later. If I don't have that ironclad certainty, it might as well be proprietary to me.

I stand corrected. The license does implement that properly. I still don't like it for a variety of reasons, but at least it's not failing in the way I supposed.

Re: It ain't going to work

Any scheme to direct funds to people based on some statistic is going to be gamed if it actually generates money. This is not an easy problem even if someone does think they've written software to manage it. I haven't found a good example of what Merico's software actually generates, but I'm really worried that it will end up being like the managers who try to make something out of lines of code written. You don't need me to tell you why that doesn't work. None of the other statistics work either. I also had a manager who tried to do something with PR frequency, if you didn't open one a week, it was not good for you, and one a day was much better. Obviously, that doesn't help much either. It takes quite a good manager to actually judge the relative usefulness of a developer, and when the manager isn't promoting or firing people, it doesn't make sense to have one. As such, I fear any automatic allocation process is doomed to uselessness. Attempting to use it anyway will motivate people to figure out what brings in money and do that; if writing docs counts, they might just have an LLM generate a lot of long-form documentation and collect the points from that, with a few people making corrections collecting a few more while doing all the real work.

Re: Stop working for them for free.

I agree with you there. The question here is whether Microsoft's purchase of GitHub was required for them to do that. I contend that it was not, and that complaints that Microsoft should ... I don't know, just not be allowed to have bought it because years later they might do something are misplacing the blame. Had they not bought the platform but still trained on its contents, as they have for books, newspapers, and likely lots of other things you can find online, we would have the same problem and we'd still have to resolve it somehow.

Re: American influence over matters IP, and software licensing

Or it might be that the organizations that played a central role in the development of the FOSS concept and the licenses that implement it happen to be based in the US and wrote their licenses specifically to work with US laws as well as international ones. When The Register interviews an American who worked with American IP law, you wouldn't be too surprised that he sounds a bit, you know, American.

Re: Stop working for them for free.

They're running their training efforts on everything public, which already includes all the PRs, approved and not, from all the open source stuff on GitHub. Their ability to do that does not require them to have bought GitHub, since I can get that information just as well by building a basic scraper bot. They have built plenty of scrapers for sites they haven't bought, so nothing prevents them from having it.

You're right that their ownership of GitHub allows them to find some reason to cancel your account, and by the way it also allows them to not bother and just cancel your account because some admin doesn't like you, but that's kind of obvious from the starting premise that they own the platform. In short, you're concerned about their ownership of GitHub, not their exploitation of copyright, but the latter is the damaging thing in this situation and does not rely on the former to exist.

"Used correctly, identifying code collisions could be used to break copyright, trade secrets and patents."

Or you might find that, for a complex system, code collisions are not at the level you're thinking they should be. Oh, it looks like that when you ask a programming-trained model to answer a simple, introductory question. There are only a few ways to write a function that are sensible, that is to say not playing code bowling, when that function does one basic thing. Here's a parallel.

There are only a few ways to express the sentence "She put the box on the table". You can use synonyms, for example that she placed or deposited the box, or you can order the words a bit differently (the box was placed on the table by her), but if you do anything too extreme, you end up with a sentence like "The furniture item known as a table was the recipient of a placement action involving the box, and she was the initiator" and you sound ridiculous. When the sentence is longer, though, you quickly expand the number of valid options to express it. No matter how complex the sentence, if it sounds natural, you can break that sentence into components that have been seen before, but the sentence as a whole is new, and the organization of multiple such sentences expands that. The same is true of programs; as they get longer and more sophisticated, the chances of finding a meaningful collision decrease and you'll be reduced to arguing over idioms or common constructs.

You're hoping that collisions will be common enough to invalidate protections you don't like, but it is unlikely you'll find them, and if you did, it likely wouldn't get the result you want. If it's really easy to recreate code, then why not recreate it yourself. If you can demonstrate that you reproduced the code on your own, then copyright is already limited. People just won't believe that you did if you show up with identical copies because they know that really doing that from scratch would be hard, so you probably didn't do that. LLMs aren't doing that either, no matter how many times their authors claim that the training data is not stored in them. As text-based ones have demonstrated repeatedly, a lot of it is in fact still in there verbatim, even if it has been reorganized. Ones that must output valid* code make it harder to get it to regurgitate large chunks, but it will and has happened. Copyright exists because it is not trivial to generate the work, even if every individual component looks simple afterward.

* Sort of, no guarantees.

It sort of would, but it would break basically everything. A Red Hat that charged for the service they provided and provided everything to the community without any roadblocks would also be disallowed by this license, since they just say that paying for an operating system is what you're not supposed to have done. So it doesn't really achieve the goals of freedom, just of not paying.

Consider how weird it can get. I have had a free Windows license. No, not what you're thinking, a real license code issued directly by Microsoft to my name in exchange for no money. They have programs for educational and charitable use that have such things. Technically, the one I used did have some money involved, but it wasn't mine. Some of them can be entirely free to the person and the organization. Theoretically, I am free to run that code on that Windows installation, am I not? Similarly, did I pay for Mac OS? I did a while ago when Mac OS upgrades cost money, but since 10.9, those are always free. I just bought a computer from Apple and the OS happens to install on it if I boot to media*. So doesn't that count as a non-paid-for OS? In practical terms, we all know that I did pay for it, but I don't have a purchase agreement for the software or a license code for it, so I can argue that I didn't. Loopholes like this are why a license needs to be clear about what they're going for and specifically write to that effect. Oh, and no matter how they write this intention, it will still be a non-open-source license and I won't be very pleased about it, but that's a separate issue.

* Mac OS installing from media: This claim is only valid for Intel-based Macs. ARM-based Macs can do that only if the hard drive is not damaged and has firmware on it. Be careful.

A tale of 2 casino ransomware attacks: One paid out, one did not

Re: Forensic accounting is a thing as is KYC.

"10M without someone getting their fingers dirty, from a legal perspective, would be really hard to shift quickly."

Let's be clear that the people who would be doing this have already started out by breaking into someone's system and installing ransomware on it. I don't think they're too worried about staying legal. They're only worried about the pragmatism of whether they get the money without being identified. Therefore, several of your objections do not apply. For example, tracking the person who bought the gold before it got to you: they don't matter. Paying ransoms is still legal, so that person can go out in public and say "I'm buying gold for these evil guys" and they are fine. So are the criminals. Nobody cares if that link is tracked down.

Similarly with the arguments about the slowness of cash processing. This is the same problem that criminals receiving cryptocurrency have. They don't really want cryptocurrency, as there are only a few things you can buy with it directly. They need to turn it into something else, and there are problems doing so quickly in a way that evades local authorities. It has no advantage over any other commodity that isn't immediately convertible into high-value purchases. While doing it in gold or physical cash has that annoying feature for the criminals, cryptocurrency has it as well.

Re: All because of crypto

"I never claimed (and if you disagree: show me were I did) that this is a silver bullet that will magically solve all cybercrime."

I agree, you did not. The person to whom I first replied did, or rather that it would magically solve all ransomware. I've seen the argument before, and similar to the discussion elsewhere in these threads about whether you could start a global trade war to stop ransomware, it could theoretically help but not as thoroughly nor as cheaply as people would like to believe.

I also agree that disabling cryptocurrency, if we could do so unilaterally, would cause some serious problems to ransomware operators. So would several alternative methods, such as making payment of ransoms illegal or having a larger dedicated police force for identifying operators and tracking them until they come to a country in which they could be arrested. We could try all of these things and more. Each would probably have some effect. None would have the ultimate effect we both want, and I fear that cryptocurrency might be among the weaker of them given the scale involved. There are a lot of criminal organizations that have spent time and effort figuring out how to move large quantities of money before cryptocurrency existed, and ransomware has become large enough that they could start to do the same. I think that there is an appetite to shut down ransomware that is strong enough that people are abandoning the step of considering the costs and likely results of possible measures. This has led, in these comments alone, to suggestions to send assassins to kill the ransomware operators and to commit acts of war against the countries in which they are located. Banning cryptocurrency is much less outlandish than either of these, but that doesn't make the description realistic.

Re: All because of crypto

Did I say that you need to keep cryptocurrency because it's so useful? I did not. What I was arguing against was this: "Get rid of cryptocurrency and the problem goes away." Sorry, that would be great, but the problem won't go away with a flip of a switch. If you pretend it will, it will only result in disappointment when you spend a long time convincing people to do the difficult work required to shut down or make effectively worthless all cryptocurrencies and ransomware operators are still around. A counterpoint is that, if we could retroactively disable cryptocurrency a decade ago, I think we might have prevented ransomware because they started with small attacks and small ransoms where finding a transfer method was not worth it, but you rarely hear about a ransomware attack on a personal laptop anymore. Nowadays, it's large groups going after large companies or governments for millions in ransom, and that scale is where finding alternatives is worth the effort and much easier to try.

I don't much care whether cryptocurrency exists. I have none, I don't want any, and the benefits it was supposed to bring it hasn't and won't. Let's still be honest about the realities involved in both the ransomware industry and the cryptocurrency industry before we claim easy answers.

Re: All because of crypto

Challenge: find a way to get £10M from someone who has it and is willing to pay it, into your hands without the police identifying you and you win £10M. Do you really think that, with an incentive like that, people will really just give up if the first one they used isn't working anymore? Can you really not conceive of a method that might work?

What if I give you an extra asset, one the largest ransomware organizations tend to have: if you have some people in Russia, the Russian police won't try to arrest you anyway. That means you can do things in Russia that wouldn't normally work in a different country, if any physical interaction is needed. I'm guessing you have some ideas that don't involve cryptocurrency. And they already have the software, so trying some new ones would be close to free. If it fails and their transaction is reversed, they can always try another method with the next company, or even call back the first company and try again.

NAT, ATM, decentralized search – and other outrageous opinions from the 1990s

Re: Year End Reminiscing

I'm not sure they're trying for high quality sound. I agree that it's not good enough for a lot of uses, but they are probably willing to forego the customers who need studio-quality sound. Those people, and I have at times been one of them, are really particular about their needs and may not be great at understanding when the network really is the limiting factor preventing it from happening. Meanwhile, there are a lot of people who, if they weren't using a videoconference, would be using a conference call over traditional phone lines. The sound on those is generally even worse, but people put up with it. These are the majority of users who will be easy to please with the basic audio and who will be using whatever cheap laptop microphone they have, which doesn't help improve the quality. If I choose to use more of my bandwidth on the audio, I can produce significant improvements without making a real dent in my available bandwidth, but most of the people I'm talking to wouldn't care that I had.

CEO arranged his own cybersecurity, with predictable results

Re: Unannounced security tests

This kind of response is exactly why tests at my company do not go to everyone at once. Someone refraining from clicking a link because someone else specifically told them not to is not good enough. If an attacker sends them, and only them, a message, then they still need to catch it and at least ask someone who knows what they're doing. Warnings on something you know to be a test are harmful.

"This implies that the best way to do phishing is by posing as a phishing test service."

There is a reason why I said that, to be reasonably sure, you would not just check that it looks like a phishing test service, but the one that previous tests have used. And why I said that to be completely certain, you would get explicit confirmation from the team that handles reports or sends out the messages. No, it does not imply that impersonating a test service is the best way to do phishing. Unless it can successfully impersonate your test provider, it won't pass my stated requirements, and if you know enough to check those requirements, you're not going to be filling it out anyway.

Re: Unannounced security tests

If the link they're clicking on is to a phishing test provider, the one your company contracts with, you can probably take a pretty good guess that it's a phishing test. If you're not sure, you can ask whoever sends them out, and if they tell you that it is one, then you know for certain. There are ways to know that other than being the sender. If you're trying to test vigilance, it can be useful to know who will click on something when there isn't someone shouting for them not to; a real attack will not necessarily go to the users who know what they're doing. Clicking on a link even when I'm looking over your shoulder and telling you it's a bad idea is a bad sign, but someone who avoids doing that is not necessarily good enough to avoid the real risks.

"our process for checking suspicious links that look plausible involves clicking them"

And almost certainly telling the attacker that this address exists and has someone who clicks on links, so send some more over there. Sure, it's a lot better than if they actually got what they wanted, and I'm assuming you clicked them on a machine that didn't have any credentials to try to steal, but does the risk of sending information to the attacker about the link clicked cause any concern?

Re: Unannounced security tests

If you know for certain that it is a test, then you also know that there isn't a risk to security if someone clicks on it. Not telling them means that the test is better, because it tells you what things would look like if they were the only or first person to get this message. They can't always count on you having received any phishing attempt before it is sent to them, so they need their own vigilance as well as listening to you. This is why phishing tests at my company are intermittent, so if one person gets a suspicious email and asks me whether I got one as well, I will usually say no completely honestly because I did not get this one. This requires them to either test the email on their own or enlist someone like me to help them do so, which is what we want them to be doing with suspicious mails anyway.

If you don't know that it is a test, tell everyone.

Re: Customers are the security liability

It wouldn't have happened automatically, but it would have been easy to open. I'm not sure about others' workflows, but I read the message before opening the attachments, mostly because there's a chance that the message will tell me that I don't need to bother with that file so opening it is a waste of time, but also to detect risks with the file. If others weren't doing that, suggesting that they might want to read the message first is not a bad plan.

Irony alert: Lawsuit alleging Chrome’s Incognito Mode isn’t will settle on unknown terms

Re: a new computer and trashing it

Why is it the default? Google wants data and will use the smallest of excuses to begin collecting it. The difficulty turning off location tracking makes this obvious; it's probably illegal, and that hasn't stopped them, so it's not surprising that they use any data the users enter. It also is necessary to use some of the features that Google provides and people tend to want if you ask them about it. For instance, you can't use Google's find a lost device or wipe it if it turns out to be stolen without having logged in. That doesn't bother me, but it is probably something users have come to expect. In general, though, they tend to expect that if the phone asks you to do something when you start it up, you have to do that thing, so they create an account and log in. Google doesn't see any reason to change this.

Re: I'm shocked!

I'm curious why people disagree with that? It's not too different from what Firefox says:

Private window: Firefox clears your search and browsing history when you close all private windows. This doesn’t make you anonymous.

Learn more

Of course, Google does a lot more nefarious stuff than Mozilla does, but their warning doesn't hide that incognito is a very basic thing that is limited to the local device. I don't mind that someone sued Google over collecting a bunch of data and tried to link it to their Chrome privacy record, which is pretty terrible, but the message they show does indicate that they haven't opted out of collecting data on their end and nor has anyone else.

Is it time for 6G already? Traffic analysis says yep

Re: Who's really driving the bandwidth demand?

Not only is there significant increase in demand from more people watching more video on their phones, but there are also more people using the networks for other devices as speeds increase. When my network went down, I still had to work and didn't want to go into the office, so I used my phone to tether my work computer. It worked pretty well. With a lot of people doing things like that, some of them will wonder about whether they need wired internet when they're already paying for this, so some people may shift even more traffic to those networks as they cancel expensive ISP contracts. I wouldn't want to do that, but with my usage, it would probably at least somewhat work. Bandwidth demand has been increasing for a long time over both wired and wireless connections. I'm not surprised to see projections that it will continue to do so.

Re: Pretty sure you can't change the laws of Physics...

The laws of physics don't change. That's why your radio receives one channel, just Morse code. After all, the first radio experiments worked that way, and the physics isn't any different today. The advantages are in our more efficient use of the laws of physics. Instead of using all frequencies, we limit to a smaller band. Instead of using analog data, we use digital data which can be more tightly packed. Some of the changes are redesigns like that, whereas others are simply improvements in manufacturing technology allowing transmitters to use less bandwidth while not getting prohibitively expensive. The laws of physics say such things as how much power your signal will have based on the distance and items between the receiver and transmitter, that is unless you use other physics hacks like finding a way to reflect it, but those laws say nothing about how low the power level can get before your receiver can no longer use it, let alone how many bits per second you can manage to shove in there. That's down to your tools, and those do change.

As for WiFi, you can have that. Lots of countries have some ISP who came up with this idea and has a public access method. Find which ones do that in your country, they probably exist, and sign up for an account. Congratulations, you have WiFi access from the house of anyone who has the same ISP and didn't change the settings on the supplied equipment. It doesn't work for people like me who buy my own equipment and configure it myself, but my neighbors have these, so you'll have coverage. That is you'll have coverage when you're near someone else's house, but I spend no effort making sure my WiFi coverage extends outside those walls very far. If you're in the street near a house, you'll probably have something, but if you're in a more open space, you won't. The same applies if you're moving between these access points. Most importantly, there are a lot of places without houses where you won't have any APs to connect to at all, and if you want the internet there, you'll need something that has better coverage than home WiFi.

Re: 5G speed not just affected by signal

I would guess that it's for tracking and billing because your provider will be responsible for charging you for billed usage and paying their partner network. Alternatively, it could be related to protection of your data so they can use networks you otherwise wouldn't trust, but I don't know whether their tunneling includes encryption or not. Both of these are guesses, though. I don't know if they are correct.

War of the workstations: How the lowest bidders shaped today's tech landscape

Re: Survival characteristics

"dictating correspondence to experienced secretaries and having it produced by a dedicated typing pool is almost always going to be quite a lot faster than writing it in something like Outlook or Word, for example."

This might be just because I'm young enough that typing pools were gone by the time I started working and dictation is done with a speech recognition program, but this doesn't seem true to me. If I dictate into a basic audio device, I can't edit anything without either verbal backspacing "Let's review this ... discuss this ... just use discuss this" or figuring out what I want to say and then reading out in one go. Either way is slower than just using the backspace key to erase the words until the document only contains the words I want. Part of it is probably that I can type quite quickly and the previous managers could not, but even so I'd imagine that they could figure it out without having to wait for their dictation to be sent to someone to type it up, a paper copy produced, them to review that, and it to be given to someone else to drive it further along. Which brings up the other speed advantages of computers which could, once a document was produced, transmit it to places faster than previous methods. I'm thinking of sending a ten-page document to ten people over one fax machine, for instance, versus sending an email to ten addresses and going back to some other work while the early network transmitted it.

Re: Correctness and Simplicity

If your comment was meant to explain what I missed, I still don't get it. Yes, one of the reasons people may have used the option the writer doesn't like could be that that one came first. Or that that one is cheaper. Or that that one was faster. Or that that one is better. Or something else entirely. Either way, it still boils down to the author complaining that people don't use the thing they wanted.

Here's an example. I don't much like Javascript because writing it is painful and it lacks a lot of stuff that a proper programming language would have, it's inefficient and calling into anything else is a mess of incompatible standards into which companies like Google want to cram everything. Unfortunately, it is the only feasible option for client-side web scripting, so we're stuck with it. The above two sentences are a better argument than that essay was because I at least told you what the system I don't like is and why, albeit in very little detail. The famous essay doesn't do either. Neither of the approaches refer to a real system with real complaints, but to theoretical systems with complaints that have been deliberately exaggerated. My two sentences are also a bad essay, because if I'm actually going to complain about web scripting, then I need to acknowledge the also ran, which would involve me explaining why Javascript is certainly a lot better than client-side Java was, and it beats Flash, and Silverlight is a word that nobody wanted to see in this sentence. In short, to admit why Javascript is where it is today. I think it is possible to create a better scripting language than JS, but that doing so is not justified because it is too popular for an alternative to be adopted, but I could specify what characteristics a replacement should have. An essay like that would make a point that others could debate. If I instead chose to make a fake language that could be JS or maybe not, then explained that everyone who was involved in it was sloppy and undisciplined, but that the good people couldn't succeed because the sloppy people got to market first, then I'm doing my own argument a disservice by ignoring reality in two different ways.

Re: Multiple languages were and are needed

And in most cases, the bloat doesn't much hurt you. In most devices, the fact that the operating system uses a few more gigabytes than it would need to isn't a problem because your disks are so massive and RAM is so cheap. If it is a problem, of course, you can start to strip out pieces to make it run more smoothly, but most people don't have to. The hardware is faster, so the software doesn't have to be really efficient to run at the same speed anymore. If you need a program to run really fast, chances are that you don't have to try running it close to bare metal to get what you need.

It's similar to our bodies. There are various things we eat which are not digested, and even those things we can digest are not completely turned to useful energy. That's not a problem; we have no need to get a digestive system that can usefully consume paper and turn it into energy, and most of us live in an environment where consuming more than we need is the bigger problem. The inefficiencies of our digestive system have, so far, not been a problem justifying extra effort be spent to improve them. Similarly, although developers could probably go to every program and strip down its memory usage, that would take more developer time than the efficiency justifies. This is especially true when we compare this software to software from decades fast, where one resource would usually be spent to avoid having to use another. For example, compressing data in memory so you didn't need much RAM to run it, but now the CPU had to do more work to get at the data, so the program itself ran more slowly. That is much less common now that we can keep all of the uncompressed data there.

We also take fewer shortcuts. For example, the versioned filesystems that Liam bemoaned the loss of in the article. We didn't have those on early personal computers because the disks were tiny and more files needed to be stored on them. Running a versioned filesystem would have used up the disk way too fast. A server had an admin who could probably make a dent in that, which is why those made some sense, and they were probably also not storing a lot of large files which personal computers eventually did. Because they had to be cheap enough to use, the feature was too expensive for users to have. While they may not come out of the box today, we can have a versioned filesystem if we want it just by configuring one. So what that it uses more storage, we can afford a bigger disk, right? This is what we mean by "fast": you get what you want out of the computer with speed that's a lot faster than you need it to be.

Re: Well written

"Much like the article I read about learned helplessness in software writing these days. I can no longer find the article"

Could it possibly be Learned Helplessness in Software Engineering? It seems to fit your description, and it wasn't too hard to find.

I think it makes some relevant points, in that the problems we're complaining about are more our fault due to our expectations. If we really think a certain system is better, we can try to get others to use it. If they don't, there may be more than we considered that goes into assessing its quality. Sometimes, the things we thought about which could make a system bad or good are not the only problems with it, and something we didn't consider will be the reason people use something we think inferior.

Re: Correctness and Simplicity

I've always had a problem with that essay. It basically doesn't explain any details, and not only sets up strawman arguments (as admitted), but doesn't even clarify what they're supposed to represent. I think my objections can be summed up in the following quote:

I have intentionally caricatured the worse-is-better philosophy to convince you that it is obviously a bad philosophy and that the New Jersey approach is a bad approach.

That's not a valid argument. It's as if I used the following setup in an argument about political systems:

The Welsh approach

* People should have choices about how their government operates and what it does.

* Torturing people is just wrong.

The English approach:

* People should have choices about how their government operates and what it does unless they're causing problems.

* Torturing people is wrong, unless those people are causing problems.

It doesn't actually tell you what the English system is, but it is pretty clearly saying that, if you agree with any other point I might have thrown into the English system's list, then you like torturing people and that's terrible. It misconstrues the options available, simplifies things into two not necessarily opposed alternatives, and it doesn't even have the courtesy to make this fallacious argument well by telling people what the right option is. I've generally summarized this essay as "There's some software I don't like and unfortunately others have used it more than the alternative I like better, so we're stuck with it now", which is a style of essay I've seen lots of times and written myself, but this one pretends to be making a wider point that I don't think it does.

Re: Survival characteristics

"They died out because their environment suddenly and dramatically changed so that they didn't fit it any longer. They were superbly adapted and thrived for 5x longer than the mammals have done to a variety of different environments... but they could only adapt so fast."

Dinosaurs as a group did, but that's a pretty broad group. We might as well say that dinosaurs are still surviving because birds are basically the same, right? In reality, there were lots of types of dinosaurs that didn't require a catastrophic change to die. Many went extinct anyway, slowly. Most of the dinosaur species that existed at some time were extinct by the time an asteroid caused some trouble for the rest.

The analogy is not exact, but it fits the computers. Lisp machines died out, and Lisp as a common language has been dwindling, but a lot of concepts that made Lisp what it is were included in other languages. I can argue that Lisp is still around in lots of places, or rather that it isn't because languages from which Lisp took many of its structure were the more influential. Your article effectively draws a lot of lines between two computers and says "these are different. This one was good and died. This one is worse and survived". As demonstrated by your examples, those lines aren't as clear as you've painted them, and when the comparisons get added on top of another, they stop making sense.

Cyber sleuths reveal how they infiltrate the biggest ransomware gangs

Re: Fascinating insights

Yes, absolutely. As a powerful ransomware operator, I have drawn the following conclusions from reading this which I will be conveying to all my staff:

1. They try to trick you.

2. They have some people who speak languages well so they can pretend to be from a place. Those people are actually from that place.

3. They try to know what they're talking about so they don't look incompetent.

As a result, I will be instructing those who look for new people to only accept incompetent people who sound like they're not from the place they say they are. We'll be victorious. All your data will be ours.

Unfortunately, the article had to leave so much out to avoid what you're worried about that it basically said nothing at all other than that the company named does this kind of thing.

Apple's timepiece turmoil taken to appeals court

Re: Stolen tech

I had assumed that they would just say "Fine, the hardware is disabled until this case ends, so people can buy them again if the hardware is completely unused", but I haven't managed to get interested enough to actually figure out what the patent is on so that assumption could easily be wrong.

Here's who thinks AI chatbots will eventually be smart enough to be your coworker

Re: I have a question:

Of course you can exchange calendar data with other organizations, but I don't and never will. My work calendar has data they can't have, and it's not like I can easily grant or revoke access to certain parts of it. This means that, unless we need to meet so frequently and irregularly that it becomes a necessity. Otherwise, I don't want people outside the organization to know who I'm meeting, the titles of the meetings, whether I'm rejecting their invites because of a real clash or not. Similarly, I wouldn't want my employers to be able to tell if I'm having an external non-work related event by seeing who has accessed my work availability. That's why less invasive methods are used.

Re: I have a question:

Also, you don't need any AI to do this. The easy cases are already fine to handle with basic software. I need a meeting with four people in my organization, so I go to whatever calendar setup we already have, add all recipients to an invite, and pull up the shared calendar feature and look for a time where everyone says they're free, then try that and see what happens. Of course, this mostly doesn't work, but not because they weren't free when they said they were; it's sometimes the problem, but usually the problem comes earlier.

When this doesn't work, it's usually for one of a few reasons that AI won't fix. It doesn't work because everyone's in some meeting and there is no time when all five of us are free that is feasible for our use case. Or there is a time zone conflict, and trying to keep this in normal work hours for everyone is proving infeasible. If I'm the only one on this continent, it's usually fine for me to take a meeting at a weird time so that it works for everyone else, but if I have a colleague in the same position, I would probably not do that without apologizing to them and getting their permission to do something like that, which an email probably won't do. Or it doesn't work because one of the four doesn't want random meetings scheduled, so they've blocked out multi-hour chunks as busy time when they are working but could be available, but I won't know that without asking them. How hard I have to ask them depends on the importance of this meeting and their presence at it, and their reaction will depend on the same thing. If it's me organizing it, it's probably pretty important because I hate meetings, which also means my colleagues are usually happier to attend something I've organized, but the AI won't know that, nor would their responding AIs.

This all gets more broken when I drop the "in my organization" part of the premise. The AI doesn't have access to the calendars of anyone outside the company, and if I have to meet with them, then we're dealing with a lot of possible ways to ask for and get scheduling information. The AI could probably deal with some of this, although I'm worried that it will get confused about someone's time zone data and make a complete mess when I'm not looking. There's a reason there are at least two companies I know of who offer a calendar reservation application as a product. An AI that lacks information will do it badly and giving the AI sufficient information to manage it is a sisyphean challenge, both in the "a lot more work than is justified" perspective and the fact that it will never end.

New York Times sues OpenAI, Microsoft over 'millions of articles' used to train ChatGPT

Re: If it's free on the Internet

Well guess what. She would be forbidden from using that power to make copies of books by reading them once. It's not how you do it, but what you do. Copying stuff that you're not permitted to copy, not allowed. Copying substantial portions of what you're not allowed to copy, not allowed. The courts will need to decide if that's what the LLMs are doing, but they have done exactly that in the past, so they'll have to find a cool new argument for why they're technically doing something different. Your simple incorrect analogies aren't going to cut it.

Re: It's all about profit

Or to limit the actions others may take with the work. For example, to have a restriction on how something can be distributed or used. I can require people using copyrighted code to release changes as open source, or someone using copyrighted text or artwork to only use it in noncommercial situations, and I have those rights because of copyright. It can also limit where the work can be displayed. For example, if I write something on my website and I want people who read it to look at other things on that site, whether because it could earn me money or not (it's not), I can restrict others' right to put it on their website instead. Those things are not necessarily about profit, though they often have an option to have a commercial benefit as well.

Programmable or 'purpose-bound' money is coming, probably as a feature in central bank digital currencies

Re: Whilst it has a lot of drawbacks...

"Yes, the shops do that, but then there isn't enough money on the card so they have to choose what to delete, from both piles."

That's an issue no matter what restrictions exist. If the quantity has any limit at all, then you can hit that limit. The only way to get around that is to increase the limit. As such, it doesn't argue against their recommendation that the specific items you can purchase be restricted.

CLIs are simply wizard at character building. Let’s not keep them to ourselves

Re: GUI-Based Admin

"Imagine you have 1,000 servers, and have to double-click on the server's icon to open up its admin GUI."

Not the right comparison here. Imagine you have 1,000 servers, and have to activate your SSH client, decrypt your public key, and enter the root password to elevate to run your script on each one. That would be painful too. That is why nobody who has to do something on a thousand servers does that. People who operate on that scale have a lot of tools to take an action on a large set of hosts. That can at times be a CLI program (take this script and run it on all hosts), but frequently is not. There are GUI-based tools that allow you to change the configuration of a set of servers too, and many of these are popular because they allow the admin to visualize the effects of a change: how many machines have completed it, how many errors have happened, are error rates changing, etc. However, there are even more interfaces for such things, including ones that deploy all those changes as code in version control, which often have multiple frontends with different capabilities, but those frontends are mostly for retrieving information, not setting it. You seem to be alleging that GUI administration involves handling one computer one by one and CLI administration is not, but neither generalization is correct.

Re: -h or --help

So your complaint is about the lack of --? That's really not the important part in all of this. Lots of commands have different syntax for their parameters. For example, do you do --param value or --param=value? It doesn't really matter, because the man page tells you that. Similarly, --install or install makes no difference. IP is different, but it doesn't really change the mess you're going to have regardless. IP could be split into different programs, and each program's arguments could be simpler, but then you'd just have thirty commands for doing something to a network interface and the complexity would be figuring out which of the thirty you needed. It would be no easier than figuring which of the many functions of IP you want and how to invoke that.

Study uncovers presence of CSAM in popular AI training dataset

Re: Problematic associations

They probably didn't, since they were referring to geographic differences. There is a lot more traffic going through countries like the UK, US, and Australia than there is in others. That traffic will represent the inhabitants of those countries more than those of others, regardless of the ethnic background of the users concerned. There is, for example, more likely to be data from people of African ancestry now living in those countries than those of African ancestry living in various countries in Africa where internet access is limited to a small subset of the population, even though the size of the latter group may be higher than the former. Similarly, the traffic generated on the African internet is likely to be biased towards countries like Nigeria and Kenya with a lot of internet infrastructure rather than countries like Chad or Eritrea which are quite lacking. This is a pattern that an AI trained on the internet will be repeating, along with many other patterns. Depending on what you want the model to do, these patterns, either that one or something else, may be desirable or undesirable, but ignoring them and expecting the AI to bypass them is a fool's errand.

"May as well just randomly scrape images off random sources."

Yes, that's the plan. Then, if you can be bothered, hire cheap labor to filter out some of the worst stuff. Then just train on the remaining mass. That's the models we have now. They contain stuff nobody wants in there, they contain illegal versions of works that the AI companies don't want to pay for, they contain complete gibberish, they contain personal information, and the AI companies are fine with it because they still look sort of authoritative when they make up something.

FTC bans Rite Aid from using AI facial recognition in stores for 5 years

Re: Skull measuring

"It is surprising how many people fail to recognise the official political philosophy of this party."

And one more of them, since what you've put in there is inaccurate. No, they did not plan or try to destroy capitalism. Having the word "socialist" in the name does not mean that, and you can look at speeches or actions to demonstrate otherwise.

Millions of Xfinity customers' info, hashed passwords feared stolen in cyberattack

Re: LOL! Just another day in the cloud

I'm not sure about the other times you said it, but you do know that A) this could happen equally well whether the server with the database is running on the cloud or in one of their buildings, B) this was initially caused by a bug in networking equipment, so if they were running a cloud-only network it wouldn't have happened (this time, but probably eventually something else would have done it) and C) this is a company with a lot of servers in their own buildings, so it probably wasn't running on the cloud? By attempting to blame the cloud for what was actually a system configuration problem, you risk proving that you don't actually understand what the cloud's security implications are.

Musk floats idea of boat mod for Cybertruck

Re: "You know, can't all vehicles briefly act as a boat?"

I don't know, Tesla was pretty good at taking something that someone else invented, making a modification to it, and getting the credit while the actual inventor who achieved it first was relegated to a "who is that". Then again, although almost all of his accomplishments were derivative rather than revolutionary, he was making them through his own knowledge and effort, so Musk still doesn't get that far.