Posts by doublelayer

Spy school dropout: GCHQ intern jailed for swiping classified data

Re: Bethan David, head of the counter terrorism division at the CPS, said

The point of the signing is right there in the quote:

"Hasaan Arshad knew his actions were prohibited"

The actions would have been prohibited whether he signed that or not. The signing was just a demonstration that he was aware of that, so ignorance of the prohibition is not an option. Had he not signed it, that ignorance could be argued as an extenuating circumstance which could possibly reduce his sentence, depending on how the judge interpreted it, but signing that bit of paper means that the information has been delivered and he has agreed to it.

Re: "Signing" The Official Secrets Act

I'd check the legalities before jumping to conclusions. A lot of security clearances are not secrets in themselves. If someone worked in a certain area of government or military which wasn't a secret, then they needed a clearance to do it. If the clearance was an official secret, so would their employment there and they'd have to come up with some other explanation for what they were doing at the time. Except in particular cases, this is usually handled by not making their clearance or employment a secret. The stuff they did there is one, but that they were doing something is not. It will be different in different countries and for different jobs, but I think you may be incorrect about the untrustworthiness of the applicants involved.

LibreOffice adds voice to 'ditch Windows for Linux' campaign

Re: Is Windows 7 THAT different in the APIs???

No. Nobody ever has or ever will. It's that OS that everyone, hopefully including Microsoft, is content to not think about. If something is supported under 7 and 10, then 8 should probably run it, and if it doesn't, too bad. If something is supported on only one of those, then 8 might run it, you find out. It has the benefit of being out of support too, meaning we have even more reason to not think about it.

UI changes meant that almost all businesses decided to skip Windows 8, and it only stuck around for a couple years before 10 replaced it, so it's also one of the versions that had the least market share. Another reason for that is that Vista was unsuccessful, so many people who were buying new hardware had bought it when Windows 7 was new, meaning they didn't need to replace their hardware for something with 8 included. Unlike with Windows 11, there wasn't as much time between 7's release and 8's, so there was less replacement due to attrition.

The launch of ChatGPT polluted the world forever, like the first atomic weapons tests

Re: Huh what a load of twaddle

I wonder if that's true. A lot of people have learned guitar because of the famous music they like so much. Even more have practiced singing, especially as we have things like karaoke that encourage them to do it even if they can't play the music to go along with it. Do you have any evidence that recorded music has stopped or reduced the number of people from learning to play it? I admit I am biased as a musician* who was born after that period, but I know a lot of people who have learned to play an instrument, and their failure to play professionally has more to do with downsides of the musical life** which are no different now than they were before.

* In the sense that I have played music in public for payment, but I don't do it now because writing software pays better.

** Playing music professionally is not the easiest, so most people I know who were good enough to do it looked at how much work they'd have to do, how much disruption doing it would have on their lives, and how low their chance of achieving success would be and decided to do something else for a job and play music as a hobby. Many others were motivated to start learning to play an instrument but realized that playing well takes a lot of time and passion and they didn't have quite so much of those, so they are not capable of playing professionally.

Danish department determined to dump Microsoft

Then don't worry about the terminology. You've complained about people calling themselves "power users", but the article only quoted someone referring to other people as such. Thus, your complaints are irrelevant. Instead, use whatever other term you like and talk about the kind of people who have the most challenging situation when asked to switch from something. I used Excel as my example as it's a pervasive one, but the same thing happens with any complex bit of software. If you dislike the term "power user" so much, call them whatever you want, but acknowledge that they're employed there for a reason and that, if you can't find an administratively acceptable way to maintain what they were doing, your plan is likely to be rejected.

The rest of your comment does seem to boil down to "my change is good, other changes are bad". Many changes might be good ideas, but there's always resistance. Replacing cobol might be helpful because it's easier to find people who know other languages and redesigning the application will make it possible to better structure and document it, but that will be resisted by people who don't want to change a language that was functioning just fine, they're comfortable writing, and taking the risk that rewriting the thing introduces bugs they'll have to spend a long time hunting down and squashing. Replacing Microsoft products has its advantages, and in all my comments, I have said so. What I'm telling you is that it is not as easy to do it as you've suggested if they've been heavily used, especially as your solutions have completely ignored how large those challenges are. If someone wants me to change things, I make them justify the change in many ways and I challenge them to explain how they will deal with problems we'll encounter during the change. If they ever say something like "if you're not smart enough to be able to make this change, then we don't need you", I resist their change because they clearly haven't thought it through and will blame its failure on me. I therefore expect that any change I suggest to them will and should face the same level of justification and planning to handle problems before it is accepted.

No, it is what you've guessed the article means. Instead, what you have is it being imposed from the top by people who have to. Do you think the Danish minister of technology, or since they don't appear to have one, the minister whose role includes this is going to personally make sure this works out? Do you think they know all the tech tools and have a plan for their replacement? I know this one. They don't. They've written a piece of paper to an employee telling them to do this and that's the last they thought of it.

The important question is, when the IT people to whom the task has been delegated come to the minister and tell them that there is a problem with some users of a program, do you think the minister is always going to side with the IT department over the others? If the IT department says the solution is to replace the workers involved and rebuild the workflows they were using, the minister will respond "Absolutely, no problem, I'll get right on that"? That's what "motivated" means, and the article makes no statement relevant to knowing how motivated any of the people involved are. What I suggest is ways to make it work without having everyone in authority motivated to accomplish this change over any other goal, since you will almost never get that. In this particular case, it being run from the top is an asset and makes some things easier, but if you wanted to accomplish that somewhere else, you might not even get that, so I also pointed out some ways to help it happen if you're in another situation such as a company where you might want to do the same thing.

Is this your way of saying that yes, you would complain loudly about any of the changes I mentioned, but you don't think others should because your reasons are better? In which case, you appear to have missed all of my points.

"Power user" is not a well-defined term and it can mean almost anything, including people who have absolutely no clue what they're doing. Let's take a more specific example. I know one of those people who is a big Excel user. He probably should have been taught real programming, but nobody did it. Instead, he self-taught a lot of programming concepts in Excel. I know the image you have, because I tend to have the same one: someone who thinks they're an expert programmer because they've managed to have ten formulas in the same sheet without any compilation loops. That's not what this guy is like. If you want complex mathematical models performed, he can probably find a way to make Excel do it faster than a programmer will, mostly because the programmer will need to spend more time on data import than he will. He has been using Excel for decades, and although the things he makes are less efficient and less maintainable than an actual program, they do work well.

If you tell him he has to use LibreOffice Calc instead, he will react much like if you went to a Cobol programmer and told them that they will now be using Java for everything and to start converting. He doesn't know how to use LibreOffice, and some things won't work, and because these are spreadsheets, there's far less automated testing, so he's going to have to manually check everything to verify that they didn't break. Let's see how to handle this and what happens if you, the IT team in charge of the transition, try them:

1. Fire that guy, replace him with someone who knows LibreOffice: his team freaks out because there aren't too many people listing LibreOffice as skills and they're the people who would have to find the replacement. They join ranks to fight you.

2. Fire that guy, replace him with a programmer: Again, the team freaks out because although they aren't as good at Excel as that guy, they can use it, but they don't know how to program so making changes will be challenging. Also, they're going to have to explain lots of details about their processes to someone who has a completely different background and they're worried it will go badly.

3. Provide training to help him and them learn how to use LibreOffice. They grumble, but if the training works, they will be able to convert.

4. Buy an Excel license and stick it in a Windows VM (if you're also switching to Linux) and let them keep using Excel. They're mostly happy.

5. Hire a programming team which you can divert among the various teams writing programs, have the spreadsheet users work with them to reimplement everything as a program, then either move the Excel guy onto the programming team if he learns how to code, have him focus on more business-focused things when his spreadsheets are obsoleted, or make him redundant when nothing he does is necessary because it's been replaced. The outcomes are very positive, but doing it properly takes a lot of time and money.

When comparing these options, the best outcomes tend to come with choices 3 to 5, but choice 5 is so hard that it's usually not an easily chosen option. Choice 3 means you have the least disruption, and choice 4 means you have the least financial cost because intensive retraining in LibreOffice, the kind that shows you how to accomplish those things you did in VBA, is more expensive than just giving that guy an Excel license. You have instead chosen the first two options whose primary advantage is that there's less work for the IT department because all of the disruption is the responsibility of the team with the spreadsheets who has to find new people and get them trained. This is why, unless you are an all-powerful leader who can do whatever you want on command, you will have more trouble choosing those options. If you can't even recognize why this is the case, then the people who make the decisions are all too likely to choose option 6: ignore the change request from the IT guys and buy some more licenses from Microsoft. If you want them to choose something other than option 6, you should start realizing why that "power user" is still employed there and why you can't ignore them the way you want to.

In this case, maybe, because you're assuming this will be imposed from the top by motivated people. Two problems prevent that from happening in many other scenarios, one of which might affect this one:

1. When trying to implement this in a company, for example, it doesn't come from the top. If you think your employer can move to Linux, you're probably right, but you don't get to tell people who disagree with you to shut up the way you could if it was your company.

2. Even when you do have people at the top mandating this, they're often not the most motivated. They tell you to accomplish it, and they expect you to achieve that and deal with the negatives. If they're bad at the job, that means broken Excel workflows are your fault. If they're good at this, they'll give you resources to help retrain people. Either way, they won't make it simple to fire those who complain too much, no matter how incompetent you say they are, especially when those Excel workflows do important things.

Conversations such as the one you've written overestimate the authority and competence of the minister. A lot of things are easy if you have complete authority. Few people have that power. Many who do don't know how to use it. The authority of the governing minister is not high enough to do many of the things commenters advise, for example summarily firing people who claim LibreOffice can't do something that they needed from Excel. Even if they did, most people in that position would not understand what any of this was about and would not care either, having delegated this switch to Linux initiative to IT employees who have even less authority. So what do you do in the real world when you can't just accomplish this by proclamation? This is what I think we need to think more about, both because it will help an initiative like this that has already started come to a successful goal but also because it will make it possible to do it elsewhere. Dismissing problems means you get ignored. Solving problems means they'll let you do what you want because they trust it won't blow up in their faces.

Re: Well good luck, seriously ...

Making any large change is difficult, but their position is stronger than many in that they have a larger scale, meaning more economies of scale with system management, and probably relatively few proprietary applications.

Also, the article isn't clear on how far they intend to go. Dumping Office is a lot easier than dumping Office and Exchange, which is easier than dumping Office, Exchange, and Windows. Lots of companies already use other providers for email, so although there's some pain in removing Exchange if that is what they were using, it's been done many times and can be managed. There are plenty of word processors, meaning that a lot of users can be switched from Word to LibreOffice Writer quite quickly, and if they're starting there, then they can afford to keep heavy Excel users in Excel until they have a transition plan. Once many of those things have been done, replacing the OS is somewhat easier. A gradual replacement is more likely to work, and from the article, it sounds like that is their plan. Of course, history gives us many examples of people giving up, and that could happen here. If they don't choose not to, I'd say they have a better chance than many of accomplishing some or all of it.

Re: No Discussion Here About Communication.....

Structured text has many benefits over unstructured text. Instead of making a table by manually typing in _, -, and |, I can have a program draw actual lines and move them automatically as required. If I want to emphasize some text, I can change the font attributes on that text, something that people have been doing with handwriting for centuries. Having headings as a deliberate element makes navigation much easier, as does hypertext. You may not need any of the features that a word processor has, but a lot of people do use them for real reasons, and most of the more complex things that nobody needs are also the ones nobody uses unless they're doing something like typesetting, in which case you can safely ignore their presence.

Do you apply that policy to when Windows messes with the UI, or if I gave you a different desktop environment under Linux than the one you like, or if I mandate you start using a different programming language? We in IT and programming are often quite averse to changes, often for what I see as very good reasons. The least we can do is not be hypocritical when others are also resistant to change, and it has another benefit, which is that we'll have a better understanding of the things we have to overcome when we want to make a change of our own.

That doesn't mean the user is always right. Change can be a good thing, and there can be benefits large enough that we require users to deal with it. Training Excel users to use LibreOffice and helping them rewrite anything LibreOffice can't run would have benefits, and you could do it if you recognized how large a change it is for people spending a lot of their working day in a spreadsheet. People who assume that it's really easy, or if people balk, they should just be replaced are as ignorant of the realities involved as the users are of how Linux works. The failure to recognize this is one reason why attempts to convert users to Linux fails. Sometimes, that happens right at the start:

IT: We're going to change the operating system and almost all the applications. Then everyone needs to learn to use the new set.

Manager: Why?

IT: [Some improvised complaints about Microsoft.]

Manager: No.

And sometimes it happens in the middle:

Manager: That thing the finance department built using a tower of Excel spreadsheets isn't running properly.

IT: Have them rewrite it.

Manager: They don't know how.

IT: Have a programmer write it in something better than a spreadsheet.

Manager: We don't have programmers on that team, we can't budget for a permanent position, and they need to change that frequently enough.

IT: Fire the Excel users and hire a programmer.

Manager: The Excel users know a lot of stuff the programmer doesn't about what we're using the spreadsheet for.

IT: The programmer can learn that.

Manager: How much are ten Windows licenses and ten Excel licenses?

In both cases, the IT department isn't wrong about the things they suggest, but they didn't pay attention to any of the related systems, making their points unconvincing. It often makes sense to replace long-running spreadsheets with a real program, and it probably is practical to convert something that remains a spreadsheet into a LibreOffice compatible version. The problem is that, if the need to do this has come along because of something IT changed or wants to change, it will be considered at least partially IT's responsibility, and if IT doesn't want to have the responsibility for it, their change request is more likely to be refused.

Re: I've heard this too many times to believe it.

China, or rather Chinese institutions, have developed a lot more things than Harmony, which is a proprietary Huawei product that isn't suitable to replace most important systems, unlike their many Linux-based alternatives, most famously Kylin. However, that is irrelevant to arguing against the point made by the original comment, which is whether they have actually started using them. That commenter did not say that nothing existed to be switched to, but that people have done a lot more talking about switching than actually switching.

Suitable arguments against this include pointing out that, while China's government does still use Windows, there are more things running Linux than most (nothing as far as I know running Harmony OS though), arguing that the reasons to switch are more noticeable now, or suggesting that Denmark will have some reason to succeed where previous efforts have failed. Mentioning that a Chinese company wrote a mobile OS doesn't do any of those things.

Do you trust Xi with your 'private' browsing data? Apple, Google stores still offer China-based VPNs, report says

Re: Does China do this?

Actually, yes, they do scan devices at the border or wherever else they may choose, and they've been doing it for longer, which is a problem. And they do deport you if you say something about their leader, although that's not the worst thing they do to such people. You might have had a point if you were trying to compare them and negatively describe the US's unjust and possibly illegal actions by pointing out the pages they copied out of China's book, but by asking whether they do the same while implying the answer is no, you're just showing that you don't know what you're talking about.

AI coding tools are like that helpful but untrustworthy friend, devs say

Re: 76% ... won't ship AI suggested code without human review

Not really, depending on what the "junior" is doing. If they're putting in prompts until the thing compiles, then sending it to the reviewer, then the reviewer is picking out lots of bugs and not getting any real work done because they're cleaning up after the LLM. Of course, the junior could be doing a better job, actually writing some code that works and having someone reviewing it, and if they can manage that by using the LLM then more power to them. Unfortunately, in my experience, most usage of the LLM ends up being counterproductive because it introduces bugs that no human would write, meaning whoever ends up having to remove those bugs is going to take a long time and the reviews have to be much longer because we're searching for things that would never be an issue in the real world. Some of the time, that's the original user of the LLM whose productivity suffers as they try to use something that should save them time but doesn't, and sometimes, it's the reviewer who receives things that obviously haven't been tested and either needs to review it and report the problems, review it and fix the problems, or throw it away and write it from scratch so the patch can go out now (the most dangerous option because they may not have another reviewer to send it to).

In traditional code reviews, there were various pieces of code that didn't have to be reviewed too strongly because you could reasonably assume that they had been written correctly enough and tested enough before it got to review as long as the person who wrote it was someone you knew and trusted. Of course, that's not the most professional approach, and ideally, every line would be ruthlessly scrutinized, but in many environments, that didn't happen because speed was an important factor. LLMs have introduced the ability to fail in places where professionals would not. That means the really detailed code review is being done, which would be an improvement if the code quality was the same or better, but it isn't because we're now dealing with the possibility of ridiculous errors having been written in by the LLM which even an incompetent human wouldn't have done. There are probably some people who have found ways to use an LLM to write functioning code, but unfortunately not as many as people who think they have but are wrong.

Re: IP Protection?

It depends whether you trust whatever system you're using to not collect it. Most of them do claim not to store or train on the data you send to them, but I don't trust that. Neither do my employers, who have decided that AI coding tools are useful* but that they need a contract to ensure that their code is only processed by one where there is a legal agreement not to take the code fed into it. I have some more trust that that agreement is being followed. You do have the option of local models if you're worried about that and still want to run one.

* I'm still not sold on them. They do not work well when making changes to anything large, and that's pretty much all of the things I do. The only situation where they're capable is writing the code for small functions you don't want to write, except that I've had to clean up from times where someone did that and it still went off the rails, for example when a colleague used some LLM to write a function to escape characters in a URL which happily mixed HTTP escaping and HTML escaping, and this was Python, so the right thing to do would be to just call the standard library function that someone already wrote to do HTTP escaping.

Cyber weapons in the Israel-Iran conflict may hit the US

Re: "fully reversible DDoS attacks"

If I'm reading that correctly, they're using "fully reversible" as an opposite of "destructive", meaning that a DDOS attack doesn't cause permanent damage because it only blocks access until the attack is terminated. They're using that to contrast with the ability to damage stuff permanently, where the DDOS is something you do to point out you're annoyed without getting anyone too angry at you. I would think "non-destructive" might be a better opposite, but I suppose it's not entirely inaccurate.

BOFH: Rerouting responsibility via firewall configs

Re: The BOFH is back to his true form

That doesn't have to be deliberate. HVAC systems are perfectly capable of doing that all on their own. Two adjacent labs in a building far too new to get any slack for this were consistently about 10 degrees apart in temperature, up and down five from normal room temperature. Our theory was that the temperature sensors were recorded as in the opposite rooms, but it could have been an even simpler problem. Either way, it never changed no matter how many people commented about it.

Re: Reichenbach...

Doyle wasn't trying to do that. He wanted to kill Holmes off, but the readers who didn't want it to end eventually made him cancel it. He probably would have had to cancel it even if a very identifiable corpse turned up just to stop the mail.

Apple-Intel divorce to be final next year

Re: Royalty

I wrote about royalties, the kind you would actually have to pay in real life, which is what we were talking about. You introduced the useless "imaginary property" term, which would have just been annoying if I had prompted you by using the actual "intellectual property" term. I didn't use that term. Therefore, your comment was not only annoying, it was also completely unrelated to anything involved. If I didn't say it was property, and I didn't, nor have I yet, then pointing out that it's not property is not a useful response. If these pedantic little points are so important to you and they probably shouldn't be, you could at least look to see if the person said what you want to argue against before you jump right in.

What I said is that you do have to pay the people who made RISC-V cores when you want to use them, and you do, and that's important if you think RISC-V means you get the practical designs without payment because you'll be surprised when you have to pay the core designers for that. Part of the reason you, personally, will be surprised is your mistaken assumption:

"As far as I can tell, with hardware fabbed to make a RISC-V chip, most businesses that sell such hardware, appear to almost always use the RISC-V reference designs and don't pay royalties for modified designs."

Look harder. SiFive, Alibaba (Xuantie), and SpacemiT are all producing their own designs. When you find RISC-V hardware for sale, what chips are they running? There are a lot of variants, but three of the most common sources are...SiFive, SpacemiT, and Alibaba. Not that those are the only places making their own designs. Those designs are not available for free, either the common definition of not having to pay or any license with some or all of the spirit of free software or open source. Nothing stops someone from making ones that do, and there are people who are doing it, but the fastest implementations that get used when people want performance aren't those.

Re: good performance

I appreciate having some things running simultaneously. I can have a navigation app with directions, send an email, and go back to the navigation app to find that it still knows where I am and where I'm going. If I had the voice instructions enabled, it would have read them out while I had the email program open. Without multitasking, it would stop working until I reloaded it, most likely parsing the map data from disk again and, in my experience, placing me back on the main screen. I can use multiple messaging apps and receive messages on any of them, meaning I can use encrypted chat with people who want while keeping an unencrypted version for everyone else. Apps that are downloading something don't have to stay in the foreground, and ones that are set to do something on a schedule don't need me to awaken them to do it*.

My response was about more than multitasking, though. I also agree that data speed was an important factor, because it made many applications viable which were never written for earlier operating systems. So did many other increases in hardware enable software which people wanted but could only really run on the more heavily resourced smartphones that came after the iPhone. Without the iPhone, that would almost certainly have happened anyway with other companies building the thing. Apple can take no credit for causing that, but if they want to, I think they can fairly take credit for being one of the most influential manufacturers around when technological development made it happen, hence why the iPhone's design** has had a lot more effect on modern phone design than Nokia's or HTC's 2007-era smartphones.

* Yes, I effectively run cron jobs from my phone. Weirdly enough, there do seem to be a few gaps in both IOS and Android's ability for apps to run in the background all the time. All the important things I've done with it have worked flawlessly, but a few unimportant things seem to have hiccups occasionally. For example, I listen to podcasts and I want my podcast app to refresh and download them overnight, which it usually does, but occasionally, it hasn't for some reason and I don't know why. So it's not perfect.

** I refer to UI design, OS design, and hardware design. It's remarkable how similar modern Android phones are to the first iPhone.

Re: Royalty

I don't think you're correct. Apple and ARM announced that they had signed another license only a few years ago. According to their own announcement:

1. Apple had to renew a license, meaning they did not have a perpetual, everything for free license before or they could have kept it.

2. The license they signed ends in 2040, suggesting that they will need to relicense then if they haven't switched to something else, so they don't have one now.

3. It doesn't say what was paid or how regularly, meaning they may have a large recurring payment requirement.

In addition, ARM makes new things. Does your theory account for what happened when ARM released ARMV9 and Apple started using it? Qualcomm doesn't seem to have that either, or some of the cases regarding their use of Nuvia's licenses wouldn't have gone the way they did.

Re: Is it just me?

In defense of the article, many of the phones that came before the IOS and Android wave needed low power but didn't need or, in many cases, have good performance. Larger mobile OSes, running applications in an environment that allowed them to do more things, meant that there was a substantial increase in the need for CPU power without raising the power consumption very much. Many of the operating systems that came before had limited multitasking, either because they hadn't bothered to write it or because they wanted to have good performance out of relatively restricted hardware. Windows Mobile was to some extent an exception there, but IOS and Android did see an acceleration in resource usage. I'm not entirely sure that's what the article intended to say, but it is the most accurate way I can interpret their statement.

Re: Royalty

Given that I was not talking about IP, but what actually happens out here in the real world when you have some hardware with a RISC-V chip in it, your post is irrelevant. If I was talking about IP law, your post would still be irrelevant, because it's an opinion on what they dislike. Here's when that's relevant: when we're having a discussion about what we want the law to be, what we would write if we had that power, or what we're campaigning for. We are not having that discussion.

Re: Take it or leave it...

I have a feeling OCLP is not going to help Intel machines when Mac OS 27 goes ARM-only. There are only so many things the skilled people there can do. It works for now, but don't count on it lasting.

Re: Royalty

"How much does the *A* in *ARM* pay in IP royalties for its cores?"

Apple isn't the A in ARM. That was "advanced" and before that, it was "Acorn". Apple was there when the corporate version of ARM was created, but they didn't stay with it all the way through. They are a licensee like others, and while they probably got a discount on their license by planning to build so many of them, they got the full-strength license which lets them design their own cores, so the answer to your question is probably a lot plus the cost of doing the design themselves.

"That is the big draw of RISC-V; royalty free designs for all."

You are mistaken. The ISA is royalty-free. The designs are covered by whatever terms the designers put on them. If you want to use a core today, not by designing your own, then you'll be paying royalties to whatever design company you select for that chip design. The difference is that the design company didn't have to pay the RISC-V people to make it.

Canva to job candidates: Thou shalt use AI during interviews

I would guess that the "defense" had more to do with it than the "American". Lots of sectors have dramatically different cultures. I have worked for tech companies, and I know people who work for finance companies in the same part of the same city, and what I would do naturally, unambiguously disagreeing with my boss in front of others for example, is something they wouldn't do. From their reactions, I have the better side of that deal. I imagine that defense tech has a lot of variation from those with a less hierarchical tech culture to ones that have the military chain of command setup. The latter acquiring the former might have culture shock even though it's the same industry.

AI's the end of the Shell as we know it and I feel fine … but insecure

Re: Too many options

I'm not sure I agree. Ffmpeg has a lot of options but that's because it has a lot of capabilities. Simple operations are quite simple and come with sensible defaults. I can easily take a file in one container format and convert it to another one, and they'll keep as much of the choices they can in the conversion. It is really simple to do that.

The complex part is when you need to change these, for example when I needed to match the compression algorithm with the one supported by a little embedded device. The two challenges involved were identifying what format that device could actually use and figuring out how to tell ffmpeg this information. The former is not ffmpeg's fault, and the latter is probably not something they can easily fix when there are twenty different options and most people don't have to know too much about them.

In bash's case, I think the problem is too few options. I tend to write any script longer than about thirty lines in Python because I can deal with actual data structures. I can do almost any of the same things in bash, but it's more annoying. I can, for example, loop through each word in a string in bash, either using bash itself or by calling out to a different binary to split them, but not as simply or as readably as writing for word in myString.split(" "):. If I want to pass around structured data, bash doesn't have as many of those, nor can I easily create my own types to store it. Bash and many other shells also come up shorter when comparing error handling options. Like ffmpeg, simple things are really easy to do in bash, but unlike it, the problem is scaling to hard things without making a spaghetti script. For historical and compatibility reasons, bash won't change their feature set too much, but I don't think an overabundance of options is my concern with it.

I disagree. No script you write is going to teach you what all the options are for ffmpeg or how to use them to accomplish the conversion you want. You have only two options: learning it and searching for it. The LLM is doing the searching part, and it might find something useful as it only really has to quote someone else who found the right options. Of course, like the manual version of searching, there are plenty of cases where nobody posted exactly the right commands for the niche conversion you want to perform, so you have to dive in and learn all the annoying details of container formats and video formats and compression possibilities and scaling and what happened to the audio there and why do there have to be so many options. Searching, whether a roll of the LLM dice or the less hallucinatory web search only work some of the time.

Tape, glass, and molecules – the future of archival storage

Re: 'Write only'?

To the extent that they do use Latin, they only use it for specific legal terms. Habeas corpus is Latin, and it's used in legal texts, at least in English common law countries, all the time. Does it mean exactly what it would have meant to Romans? Literally, it translates to something like "that you have the body" and they didn't use it as a legal term, so no. But it also doesn't mean the same thing legally that it did when first used as a legal term. In its first usage in the 10th century, it refers to punishments, and it's one of the things you're not supposed to do on a whim, but only when you have a reason. Now, it generally refers to the right to trial, which is similar in spirit, but quite different in effect. For example, in its initial usage, passing a law saying you're a criminal, then arresting you without trial, would not violate your right to habeas corpus because the law was passed; it would only count if I just arrested you because I felt like it that morning. Nowadays, that would violate your right to habeas corpus because I'd at least have to let you defend yourself, even if I was allowed to pass such a law which in many countries I am not.

That applies to those little bits that still use Latin words, but most laws use the local language in all its vagueness, and Latin is not immune to that either. Many laws leave things to the decision of a court. Sometimes, this is intentional. Any time the word "reasonable" appears anywhere, it means "the judge or jury will figure this out when it comes to it". Sometimes, it's because the people who write laws are not always as smart at encoding their desires into legal language as they think they are. That was just as true in Latin as it is in English or any other language you name. It would even be true if we all adopted Lojban. The problem is not solvable unless you are willing to reverse the cryptocurrency fans' credo, making "law is code", which is both almost impossible and undesirable.

CIO wants to grow tech team by cloning staff as digital twins and AI agents

Lack of understanding

This guy doesn't seem to have much familiarity with the kind of science fiction where most of the things he's talking about are found. Cloning people by brain extraction tends not to be a good thing. He could really use a coach to help phrase his idea in a way that doesn't sound like the villain's plan right before the machine starts cranking out zombies.

Oh, and having any understanding whatsoever about the technology he wants to use so he can know why you can't just feed a thought process into AI, but I figure that ship has sailed or he wouldn't be making such ill-informed statements in public. At this point, I don't think we can do anything to explain to him why that's not going to work. There's still a chance he could learn not to sound so stupid when he's describing the goal, though, because really what he's asking for, but probably lacks the technical experience to phrase properly, is more automation of typical network or security operations. He's hoping that he can use AI to write that automation rather than having a human do it, and he has little chance of achieving that, but it would sound more obtainable if he said what he meant.

Old but gold: Paper tape and punched cards still getting the job done – just about

Re: Many years ago

"Like most of us here probably did when creating our first web page in Notepad or similar text editor. Or adding some of the limited HTLM markup allowed on these here forums"

No, not like that. In both of those cases, you could write the thing and then take a look at it before publishing. On these forums, there's that preview button which you can use to test that the HTML is working well. I used it quite recently when I wanted to use the <del> tag. As you can tell from the tag appearing in this comment, it isn't allowed. You can get a similar effect using the s tag (no brackets allowed or it will insert the tag). Similarly, my attempt to use <ins> was not permitted either, though I didn't find a replacement. If you were writing your first HTML in notepad, then you were on a machine capable of running a GUI program that could read and display your file, and I have a feeling you used it to see what it looked like.

From their description, they didn't have the luxury of running out one copy to check their work, nor could the computer give them enough information to detect all the possible errors.

Re: digital vs physical

Which is a problem, but the other situation was often not different. The file once was stored in twenty places, but there was no list of twenty URLs. Instead, there were many pages with only one or two of them. And they didn't work, and they didn't work because people who had once stored it somewhere said "there are twenty places where this is stored, nineteen will be just fine" before they deleted it. I have a few pages specifically intended to archive things, and I have designed a few safety mechanisms to try to keep them available for as long as possible. For example, I have built a link shortener and I always give out those addresses, which means that if I ever have to move these to a different URI, for example if the domain name I used to use is no longer available to me*, then I can change the place the shortened link goes without breaking anywhere it is posted (the benefits of running the shortener yourself). Still, if that bus with my name on it comes along, I wouldn't count on the longevity of this system, which is why I manually point archive sites at such pages when I've made them.

* Some of the archives I've made are hosted on someone else's domain because they don't relate to me personally but might relate to them, but that makes it more fragile than if they're on my domain, because I'm a little obsessive about not breaking links even if it means my site's directories are not the cleanest.

Trump administration's whole-government AI plans leaked on GitHub

Re: GSA to operate like a software startup

In the startup dictionary, pivot means that your original idea is about to crash and burn, so before that happens, you lose all your money, and your company is shutting down, you quickly find some new thing to do and start doing that. Some times, it is reasonable, as your new thing will at least use technology you built for the old thing. Sometimes, the new thing is completely unrelated but, by chance, it ends up being a good idea and works out. Most of the time, you are no better at doing the new thing and you crash and burn as expected.

Forked-off Xlibre tells Wayland display protocol to DEI in a fire

"I don't get why there is a division"

Because in most cases, the alternatives work completely differently. There isn't a single standard that anything can implement and you just drop in the one you prefer. If you mix and match, you'll get things that break in places you don't expect because they were written for or integrated with only one of those. Therefore, almost everything chooses one of the options and makes everything work with that. In principle, you could integrate everything with everything and then let people choose what they wanted to run, but that's an inefficient process that most people don't want to bother with.

It all gets settled eventually; a bunch of projects start to try implementing some part of the system, many crash and burn, the remaining ones fight among themselves until the less popular ones start to lose users and be cannibalized for their useful code, and we find something that pretty much everyone uses. Then someone comes up with something the existing one handles badly and they start another version and this process starts over.

Chap claims Atari 2600 'absolutely wrecked' ChatGPT at chess

Re: eight-bit processor

What does your algorithm look like for that, because I had a program that played it which I could beat if I started first, but only because it had what I assume was a hard-coded algorithm rather than matching mine. Thus, if I started with a corner cell, it would take the middle one instead of blocking me with the opposite corner, and it would always obligingly pick one of the remaining corners for its second move, allowing me to guarantee a win. It had two ways to guarantee a stalemate. The only way to win that game is the other side isn't thinking right and you are. Once both players are older than about five, I think that covers most of the games.

Re: Why?

I would guess that many took this sentence:

"Try training your AI model on chess game records and you would have, I suspect, a very different outcome."

And interpreted it as this sentence:

"Try training your AI model large language model on chess game records and you would have, I suspect, a very different outcome."

If you read it like that, then they're probably wrong, because there were chess records in this LLM's training data. Of course, you could train an LLM only on chess records, which might make it a little better, but probably just shoot its language abilities in the foot without helping too much with the chess because words are far less reliable to encode chess position than a chess board data structure. But if you read it the way they actually wrote it, then they're right, because people have used non-LLM machine learning with the goal of having it be good at playing chess and it worked very well. Of course, if they were still thinking of an LLM there, then the interpretation could be correct, but I didn't think that was what they were saying.

Re: Why?

A lot of those chess-playing programs were trained in a very similar way that LLMs were trained. Throw a bunch of chess situations at it, have it make moves, reward it for winning, punish it for losing, go home for a while, see how good the thing is eventually. LLMs were trained with a completely different goal (how likely are these words to appear next to each other). In both cases, a lot of thought went into the mathematical aspects of this process, but if you're suggesting that the chess-playing bots are following algorithms written by chess masters that they're merely evaluating, that's not generally what happened. You can make a chess engine that is good at playing chess without being able to play chess well yourself.

Playing chess does not require intelligence. Therefore, I have a feeling we will agree about the intelligence of current software. LLMs are not intelligent. Chess engines are also not intelligent. Neither of these will ever become intelligent as long as they keep training it for the set of goals they have. But a chess engine is capable of playing chess so well that one way to see if people at competition level are cheating is to run some of them and see if the human is getting too close to their moves, and it isn't doing that by having some intelligent human personally tell it how to make the right moves. It is doing it by being able to calculate much faster than the human can.

Ex-NASA Admin pick blames Musk ties for pulled nomination

I have searched for this, and so far, I haven't found it. The closest I came was an article describing a failure to detect life using instruments the rovers have on them, but that means either sending more complicated equipment that can detect it (probably not Mars-certified yet) or bringing the samples back so it can be tested here (exactly what they're talking about).

That doesn't prove that the study you're talking about is incorrect, but to justify the added risk of sending humans and the added complexity of getting them there and back without contaminating Mars or killing them, we should have a clear understanding of what limitations they would face using robots and what limitations they would face while present. They would have the ability to look around more freely, which should help, but most other actions they do would be limited by the need to survive and preserve as much of the original area as possible so, if they failed, someone or something else could try again. If we concluded that the range of vision was the problem, we should ask ourselves whether that problem could be fixed or improved by changing the way the cameras worked before jumping straight to sending humans.

If you want to start paying for the human travel capabilities, go ahead. However, for a government-funded operation whose goal is to get samples, they should decide whether human flight is cheaper (no), better (no), less prone to disaster (no), more likely to succeed (no), or less likely to cause damage to other things (no) than a robotic mission. If several of those things were yes, it would justify the expense, even assuming that the astronauts were just fine with the high chance of death for the spirit of adventure. Which they probably are not, by the way.

This doesn't mean that humans will never leave, but that in our scientific exploration, we're not going to prioritize having humans dig up some rocks at the cost of probably killing the humans who do it and being able to do it once for every ten times we could have done it with robots. Sending robots will also teach us more which will be useful when we want to send humans. Returning rock samples will give us scientific information, whereas landing a human there will just show off that we could or more likely demonstrate that we actually couldn't.

Musk and Trump take slap fight public as bromance ends

Re: No room for two

Weirdly complex. For example, some of the territories have the concept of legislative citizenship, meaning that they are not considered normal citizens, but a law was passed saying they are. This means, theoretically, that if that law is repealed, they would lose their citizenship. If that happened, most likely some courts would have to look at whether that violated the constitution, which it might but it's also unclear. If the territory concerned obtained independence, it could probably work in the same way that most colonies did post independence, but if the territory was still a territory, depriving its inhabitants of citizenship could be considered a violation of the Universal Declaration on Human Rights. For the moment, it's a lot like an old server that is running, so nobody wants to try poking at it in case it doesn't come back up.

Citizenship is really complicated. Determining who is a citizen and how can be a much harder process than it appears. That is why my explanations on 18th century US citizenship appear to be quite unpopular. Either people think I'm wrong, but haven't tried to explain why, or they don't like the fact that citizenship was already defined by colonies* and the replacement government attached its definition to those definitions, creating an ugly mess of contradictory definitions that were only cleared away with time and a complete rewrite of the citizenship definition eighty years later, doing plenty of damage to people in the meantime. These things are not limited to that time either. There were confusing aspects to what happened to citizenships when countries divided, for example Pakistan, Czechoslovakia, or Sudan. It was a bit easier when countries like Germany or Yemen united, but not one sentence definition easy.

* And by the way, it wasn't just the US. The same colonial citizenship regulations and need to redefine them for the new country, often inheriting the data from the one thing they had, was true of all the British colonies. And the French colonies. And everyone else's colonies.

Re: No room for two

I get what you're saying, but it is not matched by the way citizenship law worked at the time. I know that the US did not exist at the time of the birth of these people. Theoretically, that means that there were no "natural born citizens" until the first birth post-independence. That was not how it worked. The relevant portion of the Constitution reads:

The Citizens of each State shall be entitled to all Privileges and Immunities of Citizens in the several States.

Could that be clearer? Yes, but although it later needed to be expanded, it meant that being born in the territories that became the states was enough to make you a citizen, and the laws upheld this. For example, when naturalization laws were passed quite quickly, those laws imposed race and religion restrictions, but those did not apply to people who had been citizens in British America before it became the US, who by law remained citizens. The presidents you are referring to were born in the colonies that became the US.

As I said in another comment, the distinction is mostly irrelevant; they didn't care at the time, and nobody alive has any effect from it now. Citizenship law is a bit more complex to "no country at the time, therefore no citizenship". You can see many more recent examples with countries emerging from colonialism, splitting, or coming under other governments that reset the laws. There are a lot of more complicated parts.

Re: No room for two

Because when the people were writing the rules, they were nervous about foreigners, and nobody's bothered to change them yet. Nothing says that rule is the best one, it's just what's written there and will be the rule until enough people decide to change it. A lot of the rules those guys wrote have aged badly and either been replaced or should be, although some others have worked out well enough, so perhaps this is one of them.

And what "natural born citizen" means is that you have to have been a citizen at the time of your birth. That means that being born in the country or to citizens counts, everything else doesn't. The complexities of 18th century US citizenship law which decided who was a citizen when the country was founded are not very important nowadays since they no longer affect anyone alive.

Re: No room for two

"It's how George Washington became president - he was born outside of the US at the time."

Not quite. Washington was born inside Virginia, which meant he was considered a native-born American as Virginia was part of the US. The same is true of all the presidents who were elected. There were some people who were possible candidates under that rule, though, most famously Alexander Hamilton, who didn't run for election but considered doing so. He was actually born outside of the land that became the US but qualified under that rule.

Your ransomware nightmare just came true – now what?

Re: Stop paying. Stop making excuses for piss-poor IT.

I definitely expect some will do that. I think it's more likely that will happen if you just have to use a method other than cryptocurrency than if it's always illegal, but either way, there will be those who ignore the law and pay anyway. The hope is that there will be enough who don't pay because it's illegal that ransomware operators find a different thing to do, hopefully not crime, though I wouldn't hold my breath. I think that's the only chance we have to accomplish that goal, and although it won't be fast or easy, nothing else we have will work at all.

Re: Stop paying. Stop making excuses for piss-poor IT.

I'm not sure I agree, but even if I did, it doesn't matter. There are lots of dishonest ways of slipping out of responsibility and people in power are great at using them.

Responsibility is a complicated thing and dangerous when used improperly. Even if you decide that it can always climb the corporate org chart without dissipating at all, it would still extend back down. So we'll fire and fine the CEO, but why not also do so to the CSO, the middle manager, the line manager, and the team that were supposed to have backups but did not? It was the lack of backups that was the problem, no? Well, we could do the same to those who designed the security systems between the machines which allowed it to spread. And to that user who clicked on the phishing email and entered their credentials. And to that guy who used the USB drive which spread it onto a different network. And to that IT person who could have blocked USB drives but didn't. Do any of these people have no responsibility?

In my experience, a lot of people have some responsibility, but no single one deserves all of it. I prefer to focus on solutions rather than punishments, and one of the reasons I prefer that is that the punishments always seem to go to someone who probably does have a little responsibility, but certainly not close enough to all of it to deserve the consequences they get. Hard-coding that to the CEO isn't going to fix that, because it gives everyone who isn't the CEO carte blanche to act as incompetently as they like on the theory that the CEO is supposed to identify that they have and prevent it from causing any problem. In my company, the CEO has no clue what I'm doing. Expecting enough information about that to rise through the management chain and go to them is not logical. If they specifically decide to cut programs we need and that causes the problem, then by all means punish them. If I am the one who broke everything, then I should face the consequences instead.

Re: Stop paying. Stop making excuses for piss-poor IT.

I prefer banning payments to banning cryptocurrency for three independent reasons:

1. Banning payments is much more practical. Banning cryptocurrency is going to run you up against a lot of people who want to keep making money from normal cryptocurrency trading, assuming you're doing it globally, and if you only do it in your country, it makes little difference; the company can pay someone in another country to do the transaction for them and that would be legal.

2. I'm not convinced that a company willing to pay an illegal ransom under the table will somehow balk at making an illegal crypto exchange under the table, so it doesn't sound like your alternative fixes any problem you were referring to.

3. If cryptocurrency was eliminated entirely as an option, but it was still legal to pay ransoms, then as long as the ransomware people can think of an alternative, they continue. They're getting payments measured in millions. That's a lot of incentive to find some other way of transferring money around, and this was possible before Bitcoin came into existence. If paying ransoms was illegal, then no matter how the companies did it, they'd still be breaking that law. I fully expect that some would continue to pay, violating the law, but I think there would be a significant decline and I'm not convinced anything else would make such a similar dent. Banning cryptocurrency might be the second largest thing we could do, but because of option 1, I don't think you can.

X's new 'encrypted' XChat feature seems no more secure than the failure that came before it

Re: E2E not a means to this end

While any company can try that in court, they are not guaranteed to win. That's why Musk isn't claiming E2E on his thing; he isn't offering it, and if he lied about having it, he might get a big lawsuit like Zoom did when they lied about offering it. A vague "Bitcoin-style encryption" which means nothing is much easier to defend and hopefully sounds good to the kind of people he wants to start trusting it.

Cops want Apple, Google to kill stolen phones remotely – so why won't they?

Re: Just legislate

Of course, they could do that. Pointing out that we already have that, the replacement would be less convenient all round*, and that there's no need to legislate, would perhaps indicate that you could skip that part. Unless you disagree with that, in which case you could explain why that legislation makes any sense.

* Current way to disable the lock: you authenticate with the account you used to lock it. Your proposed methods:

"getting a GPS reading from it": Spoofable.

"knowing other local device MAC addresses": You could try to harvest those, but the bigger problem is that most users wouldn't know them and they change, so the phone's accepted set would be small and possibly out of date by the time you got it back.

"showing up at a DMV or other government facility with sufficient ID": And this is all you're left with. So instead of authenticating normally, we're going to waste the time of multiple people to do a less reliable version of the same thing.

They can be reset, and therefore erased, but you still won't be able to use them because they won't finish setup without being connected to the internet, authenticated to the old account, and disassociated from that account. You can erase an iPhone too by using the firmware reset option that exists for cases where the phone is otherwise bricked, but again, that just lets you erase the original user's content, not take over the device.

Datacenters have a public image problem, industry confesses to The Reg

Re: Cry me a river.

While there are politicians who don't get it, a lot of people will try to see if they can get someone else to pay for their new building with bad promises. We should reject those applications if we get a chance. Keep in mind that many of these arguments aren't about government-funding, or that would be the first, most convincing, and likely only argument. They're about not wanting the buildings there at all, even when they're privately funded. There are some valid reasons not to want them there, but as long as the people making those arguments stick in the "I don't benefit from datacenters" argument, they're hurting their own point because they benefit from many, and this might be one of them. It's not only the CDN nodes that send static content, either. Plenty of DCs that have more processing capacity are important for many people, and if it's not you, it is your neighbors. I may not use TikTok, but a lot of people do and they want their videos chopped up and filtered and distributed, and that takes servers. If my neighbors switched the script, they could name things I use servers for that they don't. The argument of benefit is never going to work well. Those who would prefer for the DCs not to be built would do better to focus on the harms they cause rather than the benefits they think they're not getting.