Posts by doublelayer 9408 publicly visible posts • joined 22 Feb 2018
People don't object to it because it does look a lot like Chrome if you only use the browser, and most users will not be using the rest of the things Chrome OS supports because they either don't really work or they only make sense for technical users. It does look familiar to users who already use browsers most of the time anyway and can quickly grasp the idea that this is only a browser.
Yes, which makes a lot of people resist updating Windows until that causes problems. Which also means they will sometimes resist Linux similarly strenuously. I'm not sure whether a UI that looks like their version of Windows will trick them into not complaining about Linux, but I can see why someone thinks it's worth a try.
And, in that case, they can tell you about restrictions. It's not the restrictions themselves that this complaint is about as much as not informing people of the existence of the restrictions while charging them for a service that would normally remove those restrictions.
I suppose if they don't have retail near them, online shopping may be the cheapest way to buy things. If that was the case, Prime might be more cost-effective* for people who are placing a large number of low-value orders compared, for instance, to me where I place on average two orders per year above the free shipping limit and don't care about waiting longer for them to arrive. It depends how easy it is for people in this area to get to an alternative shop with comparable items and prices.
* If you actually get the fast shipping they advertise, that is.
"As I understand it, the agreement between Musk and Tesla depended on Musk meeting some performance and valuation goals that were, at the time of the agreement, thought to be somewhere between wildly aggressive and impossible."
The problem seems to be that there's not an agreement about whether those goals really seemed impossible at the time because the people who set the goals were Musk's friends doing things that Musk asked for. It would have been easy to find something that looked more impressive than it was and then set that as the criterion for reward. Figuring out whether that happened is more difficult now that we have hindsight and can see that and how the objectives were achieved, which is why the conflicts of interest among board members are an issue here. From all that's happened so far, I think Musk will be getting lots of stock now and through the future, so this may be a mostly moot argument about how many months he has to wait before he can get it.
I don't think "subnets should be sized in powers of two" was trying to argue against /25s, since the smallest IPV4 subnet you could create if it was would be a /16. I think it was making the equally useless point that subnets should contain 2^X addresses, for example your /25's 128 addresses. Otherwise phrased as "you shouldn't create subnets that are impossible to create" since a /25.356 for 100 addresses isn't generally supported.
Probably to make simple tasks, the only ones their bot does well most of the time, something a person is willing to write for a study. Python has several downsides, but one advantage of it is that there are various simple tasks that don't require lots of basic plumbing code to get working right. Reading in a json file can be three lines because there is a json parser in the standard library and you don't have to be too specific about types (which can be annoying for larger tasks but speeds up smaller ones a lot).
"However, the OP asked what would happen if laymen started using those tools - which IMHO implies a scenario in which they actually work as intended by the user, without hallucinations."
I don't think that's correct. People use things that are not perfect all the time, and sometimes it makes sense to do so. For example, let's use their own example of accountancy. I am not an accountant. I have not taken an accounting course. Yet I have learned a lot of the basic terms and can avoid doing the most idiotic things. Someone who employed me to manage their finances would not be making a good decision, but while their needs remained simple, it would probably work. The problem would come when they needed something more advanced and I either failed to do it or tried and demonstrated my lack of skills. Still, for someone who didn't have the funds to employ an actual accountant, someone with my level of knowledge might be good enough for something simple.
I think the same will apply to LLM-generated code, with the slight mitigating factor that the quality is more random. If you choose the right model/specific layer above a model, one like GitHub Copilot that was specifically designed to generate code rather than one like ChatGPT which can generate stuff that looks kind of like code, you can get valid code for some simple programming problems. As long as that is all you need, you may get what you want. When you start needing things that are more complex, the models break down, and by relying on them instead of having someone who knows what they're doing, that invalid code is going to run and possibly cause damage. One major difference between this and the me as an accountant scenario is that, if I was somehow employed as a cut-rate pseudoaccountant, I am likely to identify when something is well beyond my area of knowledge and refuse the task. That means that the business needs to be most concerned with tasks that are slightly outside my area but I might be stupid enough not to recognize this, but they don't need to worry as much about big things because I will point out I don't know what I'm doing. An LLM doesn't do that, so no matter how ridiculous the assignment they're given, they'll produce some kind of code. As soon as people start asking for things that connect to systems, chaos should be expected. That's just the most immediate and obvious version of chaos. A lot of other options are worse.
In my experience, yes, they can get close to the throughput depending on whether you've put extra load on the CPU. However, I only use OpenWRT equipment for home networks and I use relatively capable hardware. That means my point of comparison is often to the other cheap boxes supplied by ISPs or easily purchased on Amazon, not to managed hardware that offices use. I also have never had anything faster than gigabit for my internet connection, and I have less than that now, so at best I can say that my OpenWRT hardware from five years ago was indeed able to push packets at gigabit speeds when the rest of the path could take them. There are devices that people run OpenWRT on to extend their security update lifespan which will be limited by still having a CPU from 2010 in them. Those will almost certainly not be able to do the same thing.
It's not only WiFi, as there is one LAN port in addition to the uplink and you can always connect a switch to it. Most of the OpenWRT devices are intended primarily as WiFi APs. I do find that a little disappointing, but I'm not very surprised they've chosen to do it that way. It's less convenient for my home use, but it wouldn't have as much of an effect on someone using them as individual APs on an existing network, and I'm guessing that was one of the groups they thought a lot about.
Note: Although they've labeled the 2.5 port as WAN and the 1 GB as LAN, nothing makes you do that. You can swap those in the case where you have faster internal hardware but don't have an upstream connection that makes use of the 2.5 Gb speeds.
I generally find that those are good as long as there is a native build for it. I don't like their interface even though it has a number of conveniences built in that take manual configuration from a stock version, but if you can flash that stock version, it also means you're likely to get updates for a long time.
It can sometimes work. For example, Raspberry Pi, the charity, owned Raspberry Pi, the for-profit company that made the computers. That company made its profits, and the charity got those profits and chose what to do with them in a way that didn't involve paying it to shareholders. Now, they've decided to sell off some of that company so they get less of the money every time but they still get some of it, and the investments from others in the now public company means they can probably make more things which means more revenue for the foundation.
However, I don't think OpenAI is anywhere close to this. They started themselves as a nonprofit in order to attract people who had concerns about ethical problems related to AI research, and they probably never really intended to live with that if they could get lots of money, the same way that they keep firing and ignoring anyone who suggests an ethical problem might exist. I don't think Musk is doing any of this for any ethical reason, nor am I certain that he has the standing necessary to make anything happen, but some of his complaints are factual.
Of course there are many incompetent politicians. My point was not that they are competent after all. It was that their incompetence is not why the countries aren't dictatorships. Similarly incompetent people have become dictators. Much more competent people have at times partly run democracies without taking them over.
Putin probably does consider himself a success with a couple decades and counting of massive dominance. Not every dictator has even gotten that. Some of them were not only incompetent at running their country for its citizens, some of them were so incompetent they couldn't even manage to sustain their power very long. I consider that threshold to apply even if they lasted several years if, at the end of it, they ended up killed or imprisoned.
Of course every dictator has a lot of help on the way up. However, I can give Putin some credit for building his new dictatorship even though the pieces to do so were all there around him. Someone without his background would probably be unable to manage it, but a lot of people with his background couldn't have succeeded either because it took an eye for detail and an ability to manage loyalty and betrayal with a fine touch. As dictators go, I have to give him more credit than most, if only for correctly employing his various assets in a way that lots of others didn't do.
That's not usually caused by better managers. It's caused by missing or broken controls on the rise to power. Dictators who manage to self-build their dictatorship, E.G. Putin, tend to have some level of competence in order to accomplish that, but dictators who simply get one, E.G. Kim Jong Il, can be as incompetent as they like.
The success of democracies is not that the people in power were too incompetent to take over. It is a combination of those people not wanting to take over, institutions stopping them if they tried, and voters being smart enough to detect the most dangerous and not let them in. Many dictators, even those who managed to get power in the first place, were completely worthless at running their countries afterward. Sometimes they managed to last a long time nonetheless by killing people who threatened to do a better job.
I tend to disagree because most software certifications I see do not provide any useful skills checks. They are often tailored to some specific skill, test something that's easily scored but not very related to the difficult to score work actually involved in writing software, and mostly act as gatekeepers from some organization that will use its ability to decide whether you can get a job to extract as much as they can from the people doing work. If I were considering paper requirement, I would consider a degree from a programming curriculum that I can verify more highly than certification, because that takes years and teaches a lot of different skills. However, there are people who have all the same skills without the degree, and they should also be considered. A certification doesn't demonstrate that to my confidence, so without any better alternative, I would end up testing them myself.
It depends what you want students to gain from their studies. If you only have a little time to teach someone to program, you want to give them enough information that they can decide whether they're interested in learning more and ideally enough information that they can do some useful stuff. If you get to have a programming course every year, you can cover lots of useful things, but not everything is or is considered important enough to get that much of a student's limited time. If it ends up being a half-year or even a single-year course, some things have to be postponed to the next ones. Various bases generally get covered briefly in mathematics. Covering it again in introductory programming isn't very useful if you spend a lot of time on it, because almost all introductory programs and a large number of more complex ones don't require the programmer to directly use either of those. It will be covered later when they're using it directly and you won't get a computer science degree without knowing it, but that doesn't mean it needs to be done first.
I think this is a good summary. I've seen people who try to fix this by insisting on covering something complex. They start teaching programming by going to the individual instructions at the CPU. This doesn't work for two independent reasons.
First, people want to actually accomplish things, and there are a lot of instructions. Pretty much every language abstracts some of this, like finding where the variable you referred to is and bringing it into registers so you can work on it, but why not make people who have never seen this before do that manually. In other courses, we can teach them chemistry by manufacturing all of their reagents from precursors they mined themselves. Going too deep doesn't help, even if you started at that level.
Second, this makes it a lot harder to find someone who knows how to teach this properly. An introductory course that tries to teach that many levels needs a teacher skilled in all of them and enough time to handle them all separately. That's possible, but it generally doesn't happen often. Trying to do it when you can't is like teaching someone to drive too quickly; they might look like they know what they're doing, but it's likely to go badly later unless they train on their own.
I maintain that a good teacher and a student who is trying can do a lot to make programming make sense even if it didn't come instinctively. There are three interrelated reasons it doesn't happen when they try to expand it too far:
1. They get teachers who don't program well and teachers who don't program at all but took a four-week intensive course over the summer to teach it. Many of those teachers are trying, but they don't understand their subject, so they can't teach it well. Programming takes a lot of background knowledge to get things back onto the rails. For example, a good teacher needs to understand compiler error messages because students will do something weird and the compiler is going to do it. You may not bother teaching introductory students what l-value and r-value mean, especially as they're both sort of types but not types you create deliberately, but they'll need to understand that because compilers complain about them. The teacher needs to understand why and explain it in simpler forms so the student understands what is wrong and some basic level of why.
2. Not all students are going to try. Of course, not all students are going to try at all the other subjects either. The problem is that programming properly requires learning a lot of little things and putting them all together, and if you miss a couple of them, nothing works. Students who don't want to learn might learn something else by accident, but they will not learn any quality of programming that way.
3. People who want to teach programming to everyone sometimes also want to teach programming far too early without doing anything else. Without knowing how to drive the computer a little, programming isn't going to get you very far. You can probably teach a mostly unfamiliar student how to calculate some numbers, but they'll be missing all the rest of the reason that programming is powerful. That won't help them learn, it won't keep their interest, and it won't train them on using a computer by proxy.
That works for nearly everything that isn't considered important enough. A good teacher who understands more things can come up with a quick answer for why they're not studying cryptocurrencies:
"So far, and despite predictions, cryptocurrency is not used very frequently by average consumers or businesses as either a store of value or a medium of exchange. If you end up spending or buying cryptocurrency, you will need some additional information. If you end up building the software that implements cryptocurrencies, you will need a lot of information and practice."
Or ancient Greek:
"Most of the stuff that you could read by learning ancient Greek has been translated and is not necessary to read for most people. If you want to be a historian of any of the civilizations of the Mediterranean, you'll likely encounter it at least sometimes if not nearly always and should either learn it or work with someone who has, who you will assist with the ancient language you learned instead. If you're not planning to be a historian or in a related field, modern languages will be more useful than ancient ones."
Or cliff diving:
"We don't live near cliffs and cliff diving could kill you and I don't care if you like the sound of it."
You don't need to study any of those things to learn why not. There are many skills that we need to leave off the curriculum, sometimes because they're not very useful, sometimes because they would be very useful but there is only so much time available and students need a general education, so taking the time to study those properly would not allow for teaching them all the rest of the things that could be useful. We mostly get to try to even out that to try to teach people the set of things that would be most useful that they can actually learn in the time and resources available.
The explosions of the pagers required deliberately-installed explosive materials. You cannot cause that kind of damage or injury with a battery. At most, you can create a nasty fire and hope the holder of the device can't throw it away immediately. Most of the time, you can't even do that as the battery safety mechanisms will disable things when they get sparky and can't be deactivated via normal hacking. Normal hacking could possibly have broken some pagers, but they needed to build their own custom pagers and get them into the hands of people they wanted to hurt in order for the explosions to happen.
That might be the case for Arch, since its live installations tend to boot to a shell directly. That problem has cropped up for other CLI installations that don't start up a shell first. Other TTYs might not work if the environment isn't processing the key commands that normally access alternates. Even if they do, if the environment I'm in doesn't have a shell running or doesn't have utilities I'd use to find information, then I'm also stuck. That's why a lot of installers provide information about the thing they're asking you to select.
Exactly. From the position where you know everything about what you're doing, an installer isn't very important, but not that many people are in that position and those who are are already scripting it. For those who remain, a GUI installer is more convenient for some users and, if it is as powerful, it's often no less convenient for those who would be comfortable with the CLI. Thus, the people saying that a GUI installer is advisable have a point and should not be dismissed, especially with incorrect claims of "dumbing down". At most, the argument could be, and in my case is, that the lack of a GUI installer isn't a big deal for me personally. That isn't a very convincing argument for any broader point, though.
It depends a little on who is running, but not as much as your comment suggests. Let's consider the part where you'll possibly erase and partition one or more disks. Let's also say that you have something else running, for the sake of simplicity on another drive. When you get to the point in the installer where you specify where it is supposed to install to, which partitions it should use, and what it should create, how do you select that? For a lot of users, the GUI is the way that makes sense and anything else doesn't. CLI tools would just confuse them and mean they can't install the OS, so they won't use the OS, so developers who are targeting them won't build for the OS, and the OS is less successful than it could be.
But who cares about those lusers anyway. We don't need anyone who doesn't have at least two terminal windows as soon as they log in. We only need an installer that works for us. Great. I live in the CLI a lot of the time and I'm quite familiar with the tools to get information about and modify disks and partitions. So I run the CLI installer, it gets to the part where I specify that information and I ... well wait a minute. I would know what to do if I dropped to a shell and could start executing some commands, starting with lsblk. That shell isn't an available option right now because I'm in the installer. I don't have multiple windows, and if I exit this, I have to start from scratch again. So what happens is that I have to start an Arch environment, get a shell, find all the information I'll need during installation, write that down somewhere, then enter it during installation. At that point, why shouldn't I just script this; it will mean less risk of typos at any rate.
So who needs an installer. Anyone who uses a CLI installer is just dumbing it down from the script they should be writing. Amateur idiots, all of you. Is that approach helpful? A GUI installer is not dumbing anything down from a CLI installer, assuming they both let you do the same things. It's just presenting exactly the same options in a different form, and a form that may reduce complications, allow users to configure their installation more quickly, and attract others to using and therefore benefiting the software.
This site is mostly read by people in countries without licenses. A lot of the readers are in UK, Canada, or the US, none of which have an ID requirement. Many others are in European countries, several of which (E.G. Ireland, Netherlands, Denmark) don't have a requirement. Even Australia's closest neighbor, New Zealand, doesn't require it. So while you are right that a lot of other countries do require it, it's not that odd that most of the readers here have lived somewhere where getting a connection was as simple as paying the money.
I prefer it that way and don't see sufficient benefit to adding an extra verification and reporting step. If a country I was in wanted to institute one, I would oppose it and try to stop it. If a country I was in was considering removing their requirement, I would be happy. The fact that many countries have decided to do it doesn't change my mind.
... are you surprised by something they've done, and explained, forever? Also, why is this such an issue for you, given that they're not expecting you to change your money to dollars to buy one, and the exchange rate charged by pound-using UK sellers doesn't tend to change daily, so once you find out what your local source of compute modules is charging, it stays that much all the way through, at least it did for CM4s.
The satellites don't cover enough of the surface to interfere with UK signals since the LTE parts would be disabled when not in view of countries where they are licensed to use them. Canada and Mexico, at least those parts near the border would be less fortunate.
Yes, that is why I was careful to get the units from their request rather than risking mistaking one unit with another. The dBW to dBm conversion I can do: -110.6 dBW/m2/MHz = -80.6 dBm/m2/MHz. Working out how much of that signal you would receive on average on the surface is harder.
Starlink is trying to increase that limit to -110.6 dBW/m2/MHz. It appears it's mostly because their satellites weren't originally intended for phone use and they bolted it onto ones they originally designed for home internet, as opposed to competitors who see the phone market as a core business and designed specifically for that.
I would advise against that idea. None of your assumptions are true by my knowledge or experience. It does cost some amount to print a letter, put it in an envelope, and mail it. It isn't very much. Lots of scammers have sent mail through the history, and it isn't much more expensive if inflation is considered. Those doing it in bulk can often find ways to decrease the price. As for penalties, I don't see anyone going to lots of effort to track down and punish people who misuse the mail, not that it would be that easy to do. I can write any return address or none on mail and send it from lots of places and it will be sent. Tracking me down later wouldn't be easy even if someone was doing it.
Paper mail would be expensive for spray attacks, but it wouldn't be the first time. If they're picking specific victims, the cost of mail would be tiny in comparison to the potential reward. Mail is no more trustworthy than email, and in fact it is worse because you can theoretically verify DKIM on a message, but nobody does that on paper.
That defense effectively doesn't exist. A true insanity defense requires demonstration, or at least a convincing argument, that the person committing the crimes was so out of reality that they didn't know what they were doing or understand what the effects would be. Someone having a mental illness making it more likely to do stupid things and then doing a stupid thing, fully understanding what they were doing, that they had signed an agreement against it, and what the effects would be won't count. Those illnesses can have an effect on the punishment, but not always, but they have almost no chance of convincing anyone to acquit on that basis. I'm not sure he has other options worth trying, but nearly any alternative would be better than this.
There are two problems with that. The first is that users have a reason to want feature updates in addition to security updates. As soon as a computer can't update the operating system, it also can't update any Apple apps. Third-party ones also tend to drop older Mac OS versions quickly, but with apps like XCode, iMovie, or Pages, those updates require the new version basically the same time that version becomes available. Even though security updates are still being made for that OS, someone who has lost support is going to have that repeatedly pointed out as more and more things refuse to update but reliably tell them that an update is available.
The second problem is that Apple used to release security updates for older versions quickly. They don't do that so much anymore. Some fixes eventually get there, but it is no longer rare to see that a patch released for the latest Mac OS a month ago fixes something that is still a vulnerability in the last one. Maybe they'll patch the older one in a bit, or maybe they'll just ignore it even though that version is still getting security updates at other times. Their record isn't as good as it once was.
Meanwhile, similar to Windows, there is no good technical explanation for this. There is OpenCore Legacy Patcher for installing modern Mac OS versions on machines that Apple doesn't allow to do so which demonstrates how little there is to justify those machines being left off of support. The same complaint that applies to Windows 11's fake hardware requirements that can be bypassed safely applies to Macs. I've made the comparison before between a Mac that can run the latest version of Windows 10 without any need to bypass something but lost its Mac OS updates three years ago according to Apple. I can run modern Mac OS and Windows 11 on it by bypassing both requirements, and my need to do so started much faster for Mac OS than it will for Windows.
I have experience with the 2010 MacBook which included, in the little manual sent with the laptop, instructions on upgrading the RAM and hard drive. Battery and screen replacements were not in the manual but easily accomplished with basic tools. Apple hasn't always hated repair.
It isn't a requirement. Lots of thin machines can manage to be repairable with M.2 drives and slim batteries secured only by screws. These aren't design decisions required to make the laptops really thin. There are a few of those, like not having large or standard batteries, which are virtually required to have a thin machine. Soldered storage, parts pairing, or tons of adhesive are not required. Apple is doing those because either it costs less, it generates more revenue from replacements, or they prefer it on a philosophical level (Jobs did, but I doubt many are particularly worried about philosophy nowadays).
Probably not. There are several that haven't, but they are mostly ones that haven't had much of an opportunity and might change their plan if they got one. However, there are differences in how often they have used those dubious arguments, how much extra power they got, and how much they abused that power when they got it. China is high on all three metrics, whereas there are some groups that used the arguments, got very little, and didn't accelerate their attempts, and some others that used the arguments, got some powers, and then seemed not to use them to oppress anyone and surely that's the point of having powers that have no legitimate uses. China is far from unique, but in addition to not putting in regulations they need or putting them in in words but never enforcing them, they have lots of other abuses of the powers they do use.
Anyone who thinks I said that didn't get to sentence number three in my comment. My entire point, which I made repeatedly, is that China and the Soviet Union made a habit of saying things that sounded nice and never doing any of those things. Whether it's China with more specific claims that aren't true or Soviets with more general humanitarian claims that weren't true, the common part is that I'd like to actually have those things but neither of them provided them.
There are a lot of things in China's statements that I quite like to hear. The same way that there were a lot of things in Soviet propaganda that I liked the sound of, for example all the peace and prosperity stuff. My biggest problem is that China's implementing their ideas about the same amount the Soviets did. They frequently describe some abuse that I dislike and announce that it is now forbidden in China, and that's the last you ever hear of that, even though those abuses continue to happen there like clockwork and nobody's doing anything about it. Occasionally, they enforce one of those regulations in a drastic way against someone who just happens to have done something else they didn't like. Theoretically, China's got one of the strongest privacy protection laws in the world. In practice, it's enforced even less than GDPR is.
China has the ability to regulate as they wish, and sometimes they actually do create and enforce a rule. The problem is that, whenever they actually try that, they enforce a rule that's either repressive and authoritarian or paternalistic and useless. When they suggest and claim to implement something I'd actually like to live under, they tend not to bother really doing it. I can only conclude that those are intended as scenery.
I'm not sure any of those create an echo chamber by design, they just end up making it really easy for people to build their own. You follow your friends. If someone disagrees with you and you don't like that, you block them or get so annoying that they go away. The echo chamber isn't necessarily made for you, but if you make it yourself, the results are the same. The problem is that, if the user makes their own, it's hard to blame the operator of the service for it.
By that definition, literally everything is user friendly. Systemd is very user friendly; if you know how to use it, then you're a user, if you don't know yet, then you're not a user, and if you don't know how to do it as quickly, then you're not a full user either, the category in which you would seem to be given your choice to avoid it.
That definition only works in tautologies. A lot of people are users of something without being expert administrators of it. There are a lot of systemd users who use it because that's what they're using to manage services, and on that basis it can be less user friendly than alternatives if it is harder for those users to make it do what they want. Whether it actually is is not required for this discussion, so I'm going to opt out of that particular fight/lecture, but at least with that definition, the adjective means something and can be used in comparisons.
A lot of things can be relevant depending on the context. Yelling at others tends not to help. However, in my opinion and experience, yelling at people as a first step because that's what you like doing is pretty different from yelling at someone after they've caused massive problems several times in a row and insists on continuing the behavior that breaks them. The response to both situations might be to tell the yelling guy that he should not react like that, but the latter situation will not be entirely fixed by doing that because the problematic situation resulting in the argument would still be around, so you would have to take an additional action. Finding out which of those or where on the spectrum between them we are can help to deal with it appropriately.
The problem I see here is that some are trying to identify a right and wrong side here. Either Overstreet is correct about the technology, so shouting was completely appropriate, or shouting like that is never appropriate, so whether he was correct or not is irrelevant. Both those approaches are guaranteed to get at least one thing wrong which will cause trouble in kernel development, what both of the people involved theoretically are doing this for. In the situation where Overstreet's technical complaints, in short that proposed changes will cause existing functionality to break and possible security vulnerabilities in the kernel, are correct, then he may still need to be told not to use the language he has, but the person suggesting the changes needs to be told not to implement those changes, and attempts to do so need to be inhibited. If Overstreet is wrong, then he still needs to be told not to use the language, and his mistaken understanding should be countered by official guidance about what other developers who are watching this and implementing something connected to this should do. Or in other words, both branches allow for the CoC group to take action about the language, but I also want to see the technology problem resolved because failing to do so will have significant problems for future development in this area. Ignoring that because of angry sentences in an email would be dangerous. Deciding about it because of angry sentences in an email would be even worse.
That all depends on whether that technical detail is, in fact, irrelevant. I wasn't on the original discussion, so if his description of the consequences of different memory failure modes is just lies, then maybe I wouldn't know that. If what he writes is true, then there may be more issues than unprofessional language, not all of them his fault. Of course, since you say they are irrelevant, maybe you can explain why he was incorrect about the technical detail. Angry words, while they can cause some problems, are less of my concern than the technical disagreement, and they are by no means new to the Linux kernel devs.
That helps when you're just trying to imagine what that number of gallons means. It does not help estimate the danger that comes from that much waste, because that volume of something really dangerous would still be a big problem. The important detail would be how much damage could be expected if that much low-level nuclear waste was released, no matter how large a volume it might be now. Of course, accurately answering that would depend on a lot of details about what it would escape into; while neither is good, running into the ocean and getting diffused by currents would create a very different* situation to pooling on one specific spot on land. Of course, finding more reliable numbers for how much waste was generated and whether it's all still there would be useful inputs to figuring this out.
* Specifically, causing smaller effects on more things versus stronger effects on fewer things.
Well, duh. It would also only work for a database where the table was called "Students". If they called it "ActiveStudents", they'd be fine, or at least their problems would only be limited to Robert's missing record, ignored because the table didn't exist. Once Robert reaches adulthood, the problem mostly goes away when he's no longer listed as a student anywhere, unless he's planning to work in education where his employee record could clobber the student table. I'm not sure if you expected that to be surprising.
Also, not all of your critiques are correct. For example, the single quotes versus double quotes thing. That's an SQL command. Strings in most implementations of SQL don't get to use double quotes. For example, if they were using Postgres and they put their strings in double quotes, that's already a syntax error. So yes, if they have a system that doesn't enforce that limitation and they used double quotes, they would not have a problem. Well no, that's not correct. They would have a problem: their inputs are not sanitized and a one-byte change breaks everything, but they would not experience the results of that problem with this specific string.
Why would you want more examples? Clearly, I had them, because I also mentioned Adobe Reader which you didn't bother to rebut and mentioned two other pieces of software: audio editor (no response) and unspecified one that preinstalled Chrome with no option. That one was a media player that, although I didn't use it, I knew people who wanted it because they asked me to remove Chrome after installing it. So those are three more examples. Since you focused on Flash and decided, without reason, that Chrome installation was justified, I'm sure you can find a reason why an audio editor that didn't use a browser was clearly only used by people who truly wanted Chrome but just hadn't bothered to install it themselves.
"Likewise the people that add the additional features and/or support alternative platforms - how is this not a part of the main programming job?"
Most of them are programmers, though not all, but they're not the main programmers because the main programmers have their hands full with more recognizably browser tasks. They need to write a JavaScript engine that both supports new things added to standards and has a faster compiler. You need someone else to test that it hasn't broken other JS stuff, because the people building the JIT compiler are kind of busy to do all the amount of testing that requires. Fixing compatibility so the Linux versions work with different distros is going to be done by different programmers. Testing to see when that's required will probably be done by some other people entirely, because you can have one testing team that will test compatibility of lots of stuff even if different programmers might work on fixing problems that team finds. New features are also being developed by programmers, but again, the people working on the core browser are still doing that, so the new features will have extra people working on it specifically.
There is a pattern where people try to estimate how many people are needed to do a job by simplifying the task into its most basic form, only considering the job they know, and often making an optimistic assumption at that point. For instance, I remember an argument with someone who described Paypal's business needs as the programmers to develop some web forms, completely ignoring all the financial services licensing and operations they would also need. I don't mean that every penny Google spends on Chrome is necessary or well-spent, but it is easy to assume that a modern browser is a lot smaller than it is. Part of that is due to our previous experience with browsers that were just browsers. Chrome, Firefox, and everything else no longer are and it's likely not going back. For example, we may feel that access to cameras from a browser is not necessary; browsers managed without that for most of the time when there were browsers. However, it does let you have a videoconference system from almost any machine without having to install a local binary to do it. Maybe we don't care or would actively prefer that. Too bad for us then, because many others have come to like that and won't accept the degradation to their workflow just so we can have a browser that better matches our impression of what a browser should be. To give them what they expect, not only does our browser have to include camera access on every supported platform with different APIs, but we need a WASM-style fast computing interface because JS interpretation is too slow, enough JS libraries that it can call video functions and outsource to WASM those things that need it, and to support some other features that users want, some graphics layer so the GPU can be used on that video. You wouldn't need as many people if you didn't need those features, but the way we've gone so far, you do.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
