Posts by doublelayer

Drone firm DJI promises 'local data mode' to fend off US government's mooted ban

Re: Flying sharks

Drones which don't have additional controller hardware but operate from a phone are almost certainly using WiFi. They might also have a Bluetooth connection, but it is shorter range so it's not likely (and that can be logged as well). For that, you can in fact tap the connection on the phone's end or with something in the middle. For drones with a separate controller, it's harder but you can still figure out some things. For example, you can figure out what the frequency is and see if you have to worry about someone listening to it. If it's a band used by cellular providers, then it could theoretically be using that to exfiltrate (and if it's not using the mobile networks but still using those frequencies you shouldn't be operating it), but most likely it's a higher frequency and the only way to listen to your commands, even for someone who does have the decryption keys, is to be near you.

You could test it by connecting both the drone itself and the phone controlling it to a network which logs all the packets. While a company which wants to collect could think up a sneaky way to hide data, it's a lot harder to communicate without an interested party seeing that it's happening. Unless they've decided to include a ridiculously expensive and pointless cellular connection circuit, they will either have to use your network or keep things local. It should be provable whether they've lied here.

Go Huawei, Android: Chinese telco biz claims it will spread Harmony OS for smartphone to devs come December

Re: “Others have tried and failed"

It's already a Chinese ecosystem. Android, minus Google's spyware, plus the popular Chinese apps. All the stuff they use can be maintained and has successfully been maintained for years. They don't need this in order to have a home-grown system.

Chinese nationalism is one thing. It's the wrong thing for this argument. It's the thing that convinces people that Huawei might really be the Chinese government's sigint system, because of course all companies were founded for that purpose entirely. It makes a false equivalence between Huawei and the Chinese state, and it ignores that there are other manufacturers of phones based in China. There are several. Huawei has the largest section of the market share, but if you combine numbers 2 and 3, they have more. Include all the smaller ones, and they dwarf it by hundreds of millions of devices. You can be a Chinese nationalist, want to avoid any foreign-made tech, and still buy an entirely Chinese phone (hardware, OS, apps, network) without Huawei being involved. Or you could buy a device where Huawei is a little involved but it still doesn't run Harmony. Unless Huawei manages to convince the Chinese that other companies based in Shenzhen and Shanghai are somehow foreign, they can't just ride the wave of nationalism for their OS to be a success.

Re: The biggest problem

That's exactly the point. All five of those examples are China-based. All compete with Huawei. This isn't a market where all the domestic players drift along in unity; they're constantly looking to improve upon their devices and increase their market share. And all five of them wouldn't want to adopt an operating system that gives a major market advantage to Huawei because Huawei has seen the code and they haven't. Chinese nationalism goes only so far. It will likely get the most popular Chinese apps ported to Harmony. It may convince people to buy Harmony phones even if the OS is worse than AOSP. It will not make all the other companies fall in line behind what is just another phone company. Huawei isn't the government, and the other phone manufacturers there are not slaves to what one company does.

Re: “Others have tried and failed"

That's far too generous to it. China is a more captive market, but there are lots of smartphone manufacturers there and they're not all Huawei. Huawei isn't making Harmony OS completely open for all of them to adopt; the parts that will be released won't comprise the whole thing. Do you really expect that, when they release this next year, several other phone manufacturers will immediately start licensing it and competing against the people who wrote it using the same software? They won't. Instead, they'll keep using what they used before: Android based on AOSP. Given that's what most Chinese customers are familiar with, that will still be a formidable competitor. It isn't a foregone conclusion that Harmony will lose to Android, but it certainly might. Thinking that Huawei owns the China market is remarkably similar and similarly incorrect as assuming that the Chinese government's international spying section and Huawei are the same place.

Re: The biggest problem

"There will almost certainly be some form of ABI for running Android apps with minimal additional overhead."

Uh, I really doubt that. They might have a few things they can do which increase compatibility, and their compiler will take the same set of languages, but they're ditching all the Android system APIs. If they're going to emulate it, they'll have to include a lot of duplicate APIs and implement what's effectively the thing they already have. Minimal overhead is tricky because we're layering system API calls on top of one another, but it's possible with concerted developer effort. If they do successfully implement this, there's no reason to develop for their OS, because someone could just develop for Android and run that. It also won't help in other markets, because they're not creating an alternative implementation of Google's GMS APIs.

China won't mandate Harmony OS on local devices either. If they do, Huawei's happy. Xiaomi, Oppo, Realme, ZTE, TCL, and everyone else is unhappy. Unless Huawei really does have a very close relationship with the government, close enough that the government will voluntarily kill most of the domestic competition, it's not going to happen. Harmony will have to compete on its own merits. We'll have to wait and see what those will be.

Unexpected risks of using Apple ID: 'Sign in with Apple' will be blocked for Epic Games

Re: More secure? Only maybe

I'm afraid I don't agree on either point. A site can get hacked in a variety of ways, including in a way that allows someone to provide a password and impersonate a user. Just because they originally didn't do so doesn't make the site any more secure unless the site requires third-party sign-in. If the site does require it, it can still be hacked in such a way that the information available to a user is stolen.

On the subject of nothing important being connected to such accounts, you might be surprised. You're correct (I really hope) that no banks or email providers let people do this. However, places which do have this option include places which process payments, collect names, addresses, and phone numbers, enable access to potentially-sensitive documents, or can be used to impersonate someone. A lot can be done with access to lower-level accounts.

Re: How very petty

There is no technical issue. Since Epic violated the agreement, Apple is cutting off all the services they used to provide to Epic. Some of that is going to affect Epic's customers. Apple knows that and accepted the consequences. Epic knew this would happen to them and affect their customers. They chose to take that risk. Who you blame for the pain to the customers is your choice. I can't even bother to make up my mind anymore.

More secure? Only maybe

"It can be more secure to use one or two identity providers run by top technology companies, rather than using separate logins for every internet service, since the likes of Google and Apple are likely to run more secure systems."

That is true, which is why it's so important not to reuse passwords. However, while your password is likely more secure when the big four are the only people who have it, here are some other things that can happen:

Privacy nightmare: Any time you want to log in to something, your provider knows where, when, and how. Including things they have nothing to do with. Do you trust them to have that information? If they ever do get hacked, all your information is neatly stored in one place.

Companies can take it down for you: If they feel like it, the authorization providers can cut off your account or the ability for places you use to use that sign in. In the former case, you lose the ability to log into anything. In the latter, just the specific place (that's this article). Either way, your access could be disabled by someone who isn't the place you're interacting with.

Single point of failure: If the service you're using has a technical issue, or your ISP or theirs has an issue, you could lose access to all your SSO abilities even though you don't have any problem accessing the thing you'd like to log into.

Openness to breach: If your account isn't well-protected, for instance because someone offered you the option to log in with another party but served you a spoofed page which you didn't catch, they could be able to log into other services as you. While all four of these companies offer multi-factor authentication and it's usually well-implemented, that doesn't mean that everyone has that turned on and configured securely. If they don't, this could be a lot like reusing a password.

Compare this with a password manager, and in each case the password manager will win. Use one.

I won't be ignored: Google to banish caller roulette with Verified Calls

Re: I'm trying to think how many unsolicited calls I've actually wanted

That's why my suggestion would include the ability to reject blocked caller ID (or omit it entirely). One interesting option is to be able to announce to the caller that blocked caller ID is not accepted so they can choose to show the number. You are only slightly correct about the phone companies being able to identify spoofed numbers. Often, the originating connection knows who is really calling and bills that person, but that doesn't necessarily mean that, by the time it gets to you, your phone provider knows who it is specifically. This variable is meaningless, however, because you can't really get a phone company to do anything about a report. If you call them today, they won't have much infrastructure for identifying or pursuing the scam, partially because, for connecting the call, they'll get paid a small amount. My suggestions would require them to do this and remove any possibility that they'd conveniently fail to identify the source. Any unwanted calls would have to come in clearly identifying their source, which means most criminal attempts would be stopped quickly and commercial bulk-calling could be more easily targeted by data protection authorities.

Re: This is pointless

There is a basic phone dialer as part of AOSP, and I've seen a lot of Android phones use that as the default. I think this one is a different, Google-specific dialer, and I imagine it's shipped on Google's phones. Whether it's routinely provided by other manufacturers I'm not sure, but if this is valuable enough to Google, they'll make it one of the required preinstallations for manufacturers using Play Services. Android users can, however, change the default app which receives calls to something more trustworthy.

Re: GDPR

Yes, they would have to add Google as a data processor and specify in their updated privacy notice that they're going to do it. It will definitely be specifically opt-in, an easy switch that isn't connected to anything else, just like people have to do with tracking cookies now. Now once the data protection authorities realize that nobody's doing that with tracking cookies, maybe we'll actually make progress and stop this idea only five or six years after it's implemented.

Re: the customer's number and reason for the call are deleted from Google's servers

No, I think it's the truth. No footnote needed. That information is in fact deleted. The minutes between sending it to you and deleting it permanently are just spent running it through a parser which adds relevant chunks to your advertising model.

There's another truth in the article. Some of us might feel doubt when we hear that "The advertising behemoth also insisted that it wouldn't share sensitive information about users with its business partners", but that's also entirely honest. It will not share any of that information with the partners. It will sell the ability to use but not view that information to the partners, entirely different.

Re: its clients might stop you from ignoring calls from numbers you don't recognise

I believe the intended meaning would better be expressed as "Its clients might be able to assure you that you would like to accept their call even though you don't recognize their number". However, there's another meaning, and Google might like to implement that one. That meaning would be clearly expressed as "Its clients might be able to use this to bypass existing restrictions and get their notifications into your face more often, rewarding Google for the privilege while irritating you".

Re: I'm trying to think how many unsolicited calls I've actually wanted

I tend to answer any calls that come in (assuming I'm free), known or unknown. Scammers find themselves subjected to whatever comes to mind, although most of the time it's automatic calls who just get silence until they hang up (I'm hoping that this costs them more). However, I have the luxury to do this because somehow scammers already don't call me very often (about twice a month). Maybe it has something to do with my efforts to irritate them, but it's probably more dumb luck.

What would be useful in this circumstance is a policy change and a technical one. The technical change is to verify calling numbers and prohibit number spoofing*. The policy change is to require a mechanism to report scams to phone companies, which would be required to investigate and terminate those who get too many requests**. Both of these policies could be implemented without much consequence, and they would likely make a large dent in unwanted calls.

*My suggestion would entirely eliminate the ability to spoof a number. Two modifications are possible. First, we could allow people to provide an alternate number that will be recognized as long as their calling number is also provided and verified. Second, we could allow a blank number which clients could explicitly accept or reject. If spoofing is really that important, I think those suggestions will provide any benefits I'm willing to accept.

**In order to deal with the risk of using faked scam notifications to attack a number, the policy would only require action if a verified call from the number was made in a short period of time preceding the notification. Perhaps there would be a noise level wherein an investigation is only required after three reported scams.

Open access journals are vanishing from the web, Internet Archive stands ready to fill in the gaps

It's not that there is a problem downloading a single PDF. There's a problem downloading ten thousand PDFs. I could download them in bulk, if I were paid to click each link, get the URL, send it to wget, yawn, and find the next link. However, nobody's going to do that. Sites can take steps to make it hard for a program to find their files. They can arrange never to have a full list of things, avoid serial numbers, track and block things that look like bots, captcha when they don't need to, include circular links, change their format twice a week, only spawn links with JavaScript, require an account and log it out every ten minutes with required 2FA to get back in, [editor's note, removed the next nineteen suggestions as we presume any reader is both bored of this and gets the point by now]. Any or all of this could make it very difficult to get a bot to crawl a site successfully and actually retrieve all the content. Some of these sites aren't doing it for a public purpose; they're not Wikipedia. They want people on the site giving them the opportunity to advertise, so they'll do their best to hold their content back even when it's not their content and someone else wrote it.

Re: OA publishing

Journals could still do that. People could pay them for copies of a journal where the journal's employees have read a lot of papers and republished those they view as meritorious. The journal becomes a service which provides their reviewing expertise and a smaller number of total articles to the reader, and the reader pays for that service. At the moment, they're failing to really do that and simultaneously demanding fees from everyone involved who is actually doing the work. There's your alternate suggestion.

A lot of research in academia is done at a place which at least pretends to be working for the common good, rather than a search for maximum profits; if they had profit on their mind, they might not publish the most interesting research. Yet there is a layer between researchers and the public (or other researchers), which takes advantage of this and reverses it.

Apple to Epic: Sue me? No, sue you, pal!

Re: Microsoft Has anyone else noticed...

No, not about those. The "careful" referred to desktop operating systems, where they have a monopoly. They've been careful not to do too much to competing desktop operating systems so the courts don't go after them again, because the courts would likely hold that they still have a monopoly. They don't have a monopoly in mobile OSes, or in code repositories, or in social media, or in job placement websites. Different markets, different levels of control, different restrictions. What's illegal is to use dominance in one market to either try to destroy competition in that market or obtain dominance in another market. Know this so you can argue on the facts.

AI in the enterprise: Prepare to be disappointed – oversold but under appreciated, it can help... just not too much

Re: Performant in the Enterprise

So the question is "Have people rebranded old things as AI"? Yes. Question answered. A pity that doesn't seem to be what the arguments in the articles are talking about, since they both agree that this has happened.

Maybe it's "Everything called AI is something old which has been rebranded"? Depending on your definition of "old", that's either an obvious yes, because every program is going to be based on things that were known to us a while ago, or an obvious no, because I can point to at least a couple tools we didn't have before but we now do.

Either way, if we're just arguing about whether AI is old, we're going to come up with obvious answers. I interpreted the spirit of the question as involving some level of "Is the stuff called AI of use compared to what was previously available", which would make the debate more worthwhile, but I've now seen at least four interpretations of what the question really asks so I haven't a clue now.

Re: Performant in the Enterprise

I disagree. For one thing, what is a "dumb algorithm"? Does "dumb" just exist in the subject to contrast with "intelligence"? Does it actively mean "stupid", which means what is called AI would be less useful? For that matter, what about "yesterday's"? Does it mean that, if anything new was created and called AI, then I have to disagree because that algorithm wasn't here yesterday? Or perhaps it means that it was based on things known to us already, making pretty much everything a thing of yesterday.

I think both sides are agreeing that AI is a nebulous term that has come to be applied to many different things, but the question asks us to decide what AI is and we only have two choices. I've seen things called "AI" which are old code and most certainly stupid. I've seen "AI" which is old code but it's rather useful. I've seen "AI" which uses new techniques and has shown dramatic improvement recently. I've seen "AI" which uses new techniques and is either going to be abandoned as useless after eating through a large budget or cause active damage to its unfortunate users. How am I supposed to assign all of this to one of two buckets when I don't even know where the boundary is?

The Honor MagicBook Pro looks nice, runs like a dream, and isn't too expensive either. What more could you want?

Re: Or you could be in a queue

Depends on the location and style of queues. I'm in an urban place so I see the typical types, but I've been informed by people who live in more open places of queues of cars, sometimes very long ones. If you have to sit in a car for two hours (this happened to a friend of mine), you can type on a laptop. Especially if you're not the one driving.

Re: New requirements / brandname

Probably. It depends on repairability, compatibility with existing docks, and a reliability calculation that might not yet be available, but otherwise it's just a different model. Companies already routinely consider at least five brands, so adding another one shouldn't be that difficult.

Re: Gaming...

A cursory check says probably not. Thunderbolt is more an Intel thing. AMD processors can support it, but usually only with one of only a few boards, and all the ones I found are for desktops. This laptop's may be customized to support it, but more likely they haven't considered it as it doesn't appear in marketing for the device.

Re: Decimated

I'm afraid you may have shot your argument in the foot there. You said:

"Are you aware some words have more than one meaning?"

Yes, they seem to be. The original argument about decimate only applied to one of the meanings, and they proved that there are additional meanings which were used in this situation. You have only proved that the original meaning is still one of the options, not that it is the only option. By asking that question, I think you are also admitting the validity of their definitions, and thus the article's usage.

Re: Decimated

Deci does mean one tenth, but that doesn't necessarily mean that any word using the root means to reduce by one tenth. It could also mean to reduce to one tenth, to reduce by tenths until the desired outcome, to divide into ten even sections without reducing any of them, or any other mathematical operation you want to create where 0.1 is an important factor.

Re: How much does it weigh?

Businesses want that, but personal machines often are purchased with different requirements in mind. For one thing, you don't need a big screen or keyboard if you're going to be using it as a desktop with a battery backup, and in fact size is probably not helpful as it makes it harder to put the computer away while using the full-sized peripherals. Meanwhile, people still have a need for a laptop which is portable and runs for long enough because there are times when we have to wait elsewhere. Parents, for example, may need to wait for their children and may wish to be productive while doing so. Or you could be in a queue at somewhere which has a long wait because of pandemic restrictions. For use cases like this, battery life is quite important as are the quality and size of the internal display and keyboard.

The Wrath of Amazon: JEDI wars rage on after US Department of Defense affirms Microsoft contract

Re: Was this the contract that was originally awarded to AWS?

No, the article describes in more detail, but the course was basically this:

Contract created by department, bids requested -> Oracle's bid to run part rejected as the provider needs to run all of it -> Google decides to drop out -> Amazon and Microsoft submit bids -> Microsoft's bid accepted -> Amazon sues saying Microsoft's is invalid -> Oracle sues saying "we want money and we're unhappy we didn't get some" -> Amazon wins a preliminary case -> Oracle eventually loses theirs -> Amazon loses a case -> Amazon wins a case -> department says they still like Microsoft -> Amazon doesn't like to hear that, tries again -> you are here.

Expect that this will end soon, maybe 2027 or so.

Amiga Fast File System makes minor comeback in new Linux kernel

"Fast forward 30 *years* and Windows Explorer has finally got "native" support for reading/writing ZIP files. (But no other compressed formats.)"

I remember that already being there by the days of Windows XP, so we could just fast forward ten years. Then again, the way that that is implemented isn't particularly useful; it's just the same interface to view files but it doesn't make archive handling transparent to an application. Fortunately there are libraries like the ones designed for 7zip which can provide a common interface to various compressed files which programs can use to implement better support inside them.

Mate, it's the '90s. You don't need to be reachable every minute of every hour. Your operating system can't cope

Re: All Employee Emails

My problem isn't all-employee emails, but rather being added to too many groups. There's a group I'm in which is rather small, but we support a variety of things. The problem is that I support about two of them, so I'd not be able to help with most requests. We created multiple email lists so users could ask only those people most likely to know about the thing, but a lot of users don't appear to be sending their requests to these specific mails. Instead, they find the larger all-team address and send their request there. I can't ignore that group because some things that do apply to all of us get sent there but also because the requests for those few things I do support are going there. Still, about 90% of the mails that are sent to that group are useless.

Re: Perhaps

"Remember, a lot of people were already used to "instant" email - using a terminal login to a Unix server meant you were under a "push" model."

Is that true? I wasn't around for the older systems, but any time I've used modern email on a terminal it hasn't pushed notifications even though the mail was pushed to the server. The two ways I saw things were if I launched a terminal mail client or a message on login informing me of unread messages. If a message arrived while I was working, I didn't see it. Of course, I didn't bother trying to change this since I also connected the account concerned to a normal client, but did it work differently in the past?

Re: Perhaps

At my job, if I receive an email during work hours, it will have come in fast and I'll read it reasonably soon. I have to parse through emails that I don't really need to receive, so it's not instantaneous but I'll get to it soon enough. If you send me an email after work hours, we'll talk tomorrow. I don't sync my work email to my phone or personal computers and I don't use the work computer when not at work. If something urgently needs my attention, and it probably doesn't, there's a reason that some of my team members have my personal phone number. There's also a reason why nobody else at work has it, because there will always be someone who thinks their problem is urgent when it's not.

If people want constant contact with me, even outside work hours, those people can inform me that I need to be on call, provide me with the equipment, and pay me extra for the fact that I will limit my actions when on call so I'll be available to answer calls and respond quickly. Otherwise, they will get contact during work hours at an initial priority level that allows me to get normal work done between requests.

Apple commits to support human rights - 'We believe in the critical importance of an open society'*

Re: Engagement

"increasingly a thing now"? "not a good look"? "These are not the phrases of an expert."

Actually, they are. While that post may not have come from an expert, experts use those phrases with frequency. "Increasingly a thing now" is a quick and informal way of saying "is a policy which has been adopted with increasing frequency by a variety of participants". "Not a good look" is a quick and informal way of saying "even if the decision is in line with the ideals of the institution, it runs contrary to the ideals of an important section of the public. Continuing to pursue the current course may result in a negative reaction by the public which may carry with it additional consequences".

But also, who cares ["What relevance does the phrasing have to the discussion"]? This is an internet forum. We state our opinions here. This is not limited to experts and there are nonexperts here. We also write informally to get our point across. Do you have any comments about the opinion stated in that point, or do you simply want to point out the use of an informal expression on an informal forum because you aren't able to refute the original point on its merits?

Re: Engagement

This is Apple we're talking about. Why you brought up Google is beyond me, but since you have, Apple could take a page from Google's book. Granted, a page Google's been trying to rip out of the book, but they haven't yet. China told Google that they have to filter things for Chinese searches. Google decided to stand for freedom of information there and refused. China blocked them. That's standing up for what you believe in. Apple ... hasn't.

Now that's not fair to either of them; Google has been trying to build a Chinese search engine, only being stopped by small commercial things and a protest by their own workers. Therefore, I can't give them any credit now for standing up for stuff. The fact remains that at one point, they had principles and they stood up for them. That would be nice to see again. It's not exactly fair to Apple either, as they have stood up to some things. They stood up to the Americans when they wanted a backdoor; that was nice. So they're not terrible. They have not, however, stood up to China's requests for censorship.

Remember OpenAI's GPT model that was too dangerous for mere mortals? Well, it's now for sale on Azure

Re: facial recognition

"Also, why would 2 days in detention mean $12 Million payout? Was Mr Oliver paid $6M a day and he missed out?"

In the case of the teacher, incorrect testimony, while damaging to the one arrested, is likely not intentional and can't produce damages. In the case of the police, let them suffer. They used software that isn't capable of doing anything correctly to arrest an innocent person. Despite it being easy to disprove the incorrect software by comparing pictures, they got him anyway, failed to do the comparison when looking at him in person, put him in a lineup, failed to see the differences, put him in prison, failed to check that picture again, pressed charges, failed to see if he was the person they thought he was when they were constructing said charges, and it eventually fell to someone else to point out this rudimentary difference and end the farce. If I had kidnapped you for two and a half days, I'd spend years in prison (in that state, I could get a sentence of life imprisonment). I view several million as an acceptable alternative punishment for what is, in effect, a very similar offense.

Surprise! Voting app maker roasted by computer boffins for poor security now begs US courts to limit flaw finding

Yes, it seemed to me as if you were trying to be conciliatory. Statements that start like "I can't see anything wrong with Voatz position - providing" sound as if you think there's a possibility where their recommendations can be accepted. You clearly hedge against that with the recommendation you provided, and that limitation is useful, but in my opinion you've already given up too much by giving them anything. My reasons for that opinion are stated above.

I get that you're trying to be conciliatory, but that approach is extremely unacceptable. They shouldn't just be liable after their travesty causes massive problems. That way would help, but it's a lot like saying that I'm responsible after my non-IAEA approved nuclear reactor spreads radioactive waste over the local area. When things are this important, they need to be responsible before anything happens. That means that, when there are sufficient amounts of consumer information involved, when the government is going to use it, or when a malfunction could cause injury or death, the law should require that they do testing with independent testers and they should either have to implement the fixes to any bugs the testers find or appeal the decision not to. Having researchers who test systems simply helps this process and makes it cheaper. We require it of people making medicines or medical devices. We require it of people making cars and aircraft. We require it of people growing or manufacturing food. We can require it of people using the public's data, too.

Re: Voatz meet the Streisand Effect.

Ah, but it's not the discussion that they mind. If people investigate a system to find that it's a hideous mass of holes, they're violating the trust of the organization that put out the hideous mass of holes. It's important that we respect the rights of places that don't bother doing their own security testing and choose to use untrustworthy and unsafe code to store and process our information to make money. More than that, we must protect those who don't want to bother making good products from people who shamelessly figure out whether something will become a safety risk and, these people have no scruples, have the gall to tell the public about it after they tell the company who doesn't fix it. Consider how you would feel if someone researched the safety of cars and told people about the ones that blow up so you couldn't purchase one of those. Consider how you would feel if there was someone with the audacity to check if the claims of other product's advertisements were true and call out the selfless manufacturers when they were found to be lying through their teeth. These people must be stopped today.

As Amazon pulls union-buster job ads, workers describe a 'Mad Max' atmosphere – unsafe, bullying, abusive

Re: Not enough

Not really true. Some regulations vary between states, including some important ones, but a lot of others are federal. There is much more similarity between states than there is between European countries. In addition, all of Amazon's warehouse activities in the U.S. are directed from the same location, whereas their operations in other countries are often managed locally (those local managers still get commands from the American office, but they do things differently and may be in a position to operate differently). Therefore, the management decisions are much more likely to apply to all of the U.S. than they are to apply to other countries.

Why cloud costs get out of control: Too much lift and shift, and pricing that is 'screwy and broken'

Re: Cloud certainly has its place....

I don't think that's it. If they just kept increasing the prices for the same thing, that might apply, but they don't seem to do that very often. Instead, they have so many products, each of which has sixteen variants, and each variant of each product has a different price depending on where you host it, so nobody can really keep track of the price differences. Also, they don't bring people's attention to the things where they make the most revenue, meaning they can slip in a slightly higher than expected bandwidth price and people only notice if they do a lot of research. It's not quite extortion based on lock-in, but instead a morass of complexity.

For example, just look at bandwidth charges. I wanted to do a simple comparison to demonstrate how quickly those could run away on you. So I tried to find the prices. I found Azure's quickly. Google's took longer but was still in my first set of search results. Amazon's was tricky to find via DuckDuckGo. I found unrelated AWS pages, third-party blog posts purporting to explain it, outdated information of all kinds, and increasingly shrill requests by Amazon to let them do the math for me (only about 58 steps). If you scroll down a lot on this page, there are transfer prices for EC2 VMs, and for all I know there are different charges for every other product but I got bored. Now we can begin to actually compare the prices for bandwidth. Microsoft's seems cheaper, but Google's is a little unclear how a public-facing one will work, and each one offers different prices based on the selected region, includes exceptions for certain types of traffic, and doesn't show you the full set of numbers in one chart. I think this illustrates the problem. I could figure this out, do some calculations, and actually estimate the bill for each system. Except these prices are for one aspect of one product and there are a hundred more products and price lists to consider. The more chaotic the set of systems, the harder estimating a price will be even for routine usage and the easier to quickly find out that something was overlooked.

As promised, Apple will now entertain suggestions from the hoi polloi on how it should run its App Store

Re: A larger share ?

"It costs $99/yr to register as a developer and be able to submit apps to the App Store. Not free but not far from it."

Well, that does build up. Especially if I'm already making free apps and giving my time away, it's a little annoying to have to give Apple money so I can give my code away for free.

"There's no requirement that the developer charge for their app, and indeed many are free and rely on advertising to make money for the developer. Apple does not get a cut of in-app advertising, so all those developers are "free riding" and the 30% they make off those who do charge goes partly to defray those costs."

Rubbish. A free app does make Apple money; it makes users want IOS devices more because that app is there. They don't pay the developer for it. They do provide some bandwidth so users can download it, but they already have a bunch of servers and most apps are small. It does not take much money to provide that service, and as we've already established, they already get money from the developers of said free apps with their annual fee.

"If they were forced to significantly lower that 30% they'd probably have to institute a minimum price for apps."

They wouldn't have to and they wouldn't. The free apps help them, and if there was a minimum price, developers wouldn't publish many of those free apps for a fee. Some would not want to attach their donated code to a charge, and some companies wouldn't want their users to have an initial charge for an app that might put them off from trying it.

"Compare with Google where they are collecting both the 30% fee AND making money off the in-app advertising for that majority of apps which use Google for their in-app advertising."

That's also potentially problematic. Google has one argument which is that you can use other stores, but I don't think that's enough. So I'll compare and recommend investigation of them as well.

Re: A larger share ?

Google's security testing has been repeatedly shown to be harmfully insufficient. The level of malware in the Play Store is ridiculous, the length of time it remains there is extensive, and the ease with which developers of the crap make slight modifications and get it back there is alarming. I do not consider "security testing" which routinely allows clearly faked apps containing nonobfuscated malicious libraries to be published to be worthwhile. What's worse, when the headlines and articles tell those stories, the apps are usually still there, when you'd assume that as soon as someone found one, Google would have quickly verified and killed the listing. While I can't give Apple that much credit, the incidences of this in their repository are significantly less worrying. It will happen to everyone, but if one company makes a real effort to prevent and curtail it and the other one runs a small automatic checker and calls it good, one is a lot better than the other.

'We're not claiming to replace humans,' says Google, but we want to be 'close enough' that you can't tell it's a bot talking

Re: And . . .why ?

If they can get it good enough to convince someone with a big call center, the business will fire half the call center workers and put that money Google's way. Then the businesspeople who heard the demo won't ever call said call center, so they'll never hear how well or badly it does in real usage. Then Google can show the same convincing demo to other businesses and cite that one as a proven success story.

I can answer the article's question

"Google's criteria for ethical use and how it will prevent misuse was not stated."

However, I have access to some documents which should make them clear. Here is their ethical review form:

Your name:

Purpose of your use of the system:

Credit card number:

How much credit would you like applied to your account (note: this credit is nonrefundable):

Urgency of your request:

And this is the misuse form letter:

We have received a notice that your usage of the system is considered misuse. Your account has been closed to prevent you from continuing to misuse. If you wish to continue to use these features, please submit a request for appeal within your account dashboard. We ask you not to create a new account as we are currently unable to process appeals between accounts for privacy reasons; information connected to one account cannot be attached to another one. [Wink wink]

Salon told to change ad looking for 'happy' stylist because it 'discriminated against unhappy people'

Re: They have a point.

"Successfully suing someone would make you happy."

I haven't ever done so and I hope I never will because I honestly don't think it'd make me happy. If I'm suing someone, they probably have done something bad to me or to others, and I'm trying to get them to stop doing that. So I would be pleased if the successful suit meant that they would cut it out. However, it's not the kind of happy that lasts very long, being the cessation of a bad thing rather than a start of a good thing. The only other part that might result in happiness is if I get a large settlement, but I think even that would only be of minor benefit to my mood. Maybe I'm just imagining it wrong.

Re: I'm with Richard

"The old one cranked up their prices by 50% following the post(?)-covid reopening."

Most likely, they are having fewer customers because they have to disinfect everything every time one leaves and because they might have had customers closer together previously. If they can't do as much work in the same time, they might only be able to manage to stay afloat by charging more. Not that that puts any restrictions on you, but it's worth considering why they might have done that.

Anyone else noticed that the top countries for broadband speeds are well-known tax havens? No? Just us then?

I really don't think that's how it works. If 71/100 meant that 71% of the population had a device, it might be arguable that the age disparity explains it. However, given the figures for developed countries, I'm guessing the truth is very different. Germany, for example, has 55 fixed connections per 100 people and 132 cellular connections per 100. It's possible that this means that nearly all Germans have a device, 32% have a second one, and a little over half the people still use fixed lines. I think it's more likely, however, that most of the fixed lines are going to corporate or government offices, and a lot of the cellular devices are likewise. This especially applies as some of these devices might be IoT things using the network.

It seems we are unlikely to agree on what exactly these numbers mean for the availability of networks, so let's look at a different set. Wikipedia has a table estimating the number of internet users from subscriber information and surveys. Each row of this table could be disputed, but I think it's likely that most rows are trustworthy. It is estimated there that 38.66% of Pakistanis accessed the internet in the last year. The reason this number doesn't surprise me is that, even where mobile phones are available, there isn't necessarily data usage from them. A phone attempting to do much online over 2G isn't going to succeed very well. Modern feature phones may have web browsers, but modern web pages are going to take forever to load across the miserably slow connection and they're also going to render badly. Also, if people are having trouble affording things, they probably aren't using the most recent of feature phone designs. An old 2G phone still works fine if the area has that capacity. What's more, most contracts I have seen offering 2G data are close to extortionate. For people using feature phones, they're unlikely to afford routine data use. Also keep in mind that it is estimated that 50 million in Pakistan don't have electricity, which demonstrates the level of financial difficulty they're in. Meanwhile, calls or SMS messages are usually much cheaper. This doesn't have to apply to everyone; 82 million users is still a lot of users, but if there are such areas then we can't really claim to have worldwide penetration.

That still doesn't prove that "There are very few people in the world now who don't have internet connection." For one thing, those numbers don't state how many of those phones are capable of using internet. If it's a feature phone with 2G, then it probably doesn't do internet at all. If the local network is 2G, it's likely that the population isn't using it for internet even if their phones technically could.

However, the original statement was about the world, so let me find some countries that are more to your liking:

Chad: 1 fixed line per 100 inhabitants, 52 cellular per 100, 16 million people in total

Haiti: 1 fixed per 100, 61 cellular per 100, 11 million total

Kiribati: 1 fixed per 100, 43 cellular per 100, 110K

Pakistan: 1 fixed per 100, 71 cellular per 100, 208 million total

Papua New Guinea: 2 fixed per 100, 55 cellular per 100, 7 million total

Sudan: 1 fixed per 100, 77 cellular per 100, 43 million total

Those are some large gaps. In addition, it's worth keeping in mind that a bunch of connections can still be in one particular area, particularly if people are using multiple devices or if companies deploy cellular infrastructure (usually, the number of connections counts individual devices, and they don't care that a company may have deployed a hundred devices for corporate purposes and people aren't using them). For context, it's worth keeping in mind that, even with it's 1.27 connections per person rating, the Philippines is estimated to have 20.6 million people without access to electricity. I'll grant that a phone is easier to keep charged when power is problematic than most other electronics, but I'm still guessing that some of those people can't afford a phone any more than they can afford mains power.

Qualcomm flexes latest Arm chipset for laptops: Snappy performance and battery life if you can put off your upgrade long enough

Re: A notepad anyone

By "notepad", do you mean tablet with handwriting recognition capability? If so, why this chipset and why Chrome OS? First, you're likely going for lots of battery life, but the 25 hour estimate/exaggeration is for a laptop which can have a bigger battery. Most tablets don't have as large a battery because they have to put that all in the screen bit, making it really thick, whereas the laptop can put it in the keyboard bit. If someone wanted such a tablet with really nice battery life, they could of course make it thicker, but it might be better just to use a lower-power processor since it's designed mostly for writing.

For the software, Chrome OS seems ill-suited for, well a lot of things, but this task is one of them. If it's intended to be a notepad, I.E. frequently used to write documents, you want as much availability, speed, and power conservation as you can get. For example, interpreting writing in real time can be complex, so a recognizer that runs in native compiled code on the device itself is going to be fastest and use the least power. One which runs on the device but in JavaScript is going to be slower and will put more weight on the processor, whereas one that runs on the cloud will only work when the device is online, include a latency problem, and unnecessarily use power communicating with said server. While Chrome OS has been grudgingly adding some of this, most other OSes are significantly more eager to use the device for processing. Linux, Android, or Windows would all be able to do the job much better, especially given that the latter two have both had manufacturers who have invested in handwriting input quite a lot.

Maybe you want something which is different from what I think you want. If I'm correct, however, I don't think your suggested approach is the best one.

Re: There won't be laptops with 25 hour battery life...

I guess the main benefits are these, in increasing priority:

1. You might want to have a laptop with you while you're away from power for an extended period. For some reason. I'm not sure exactly why. The best option I have is that your residence might lose power intermittently. If it was longer, the battery would still die on you, so it still has to be a short time (maybe a week maximum) away from power.

2. You have a long journey ahead of you but the vehicle which is involved doesn't have a mechanism to recharge your device.

3. By using less power, you can reduce the number of times you have to recharge and prolong the battery's life.

4. They're lying their face off and 25 means 13, but at least with 13 hours of battery, it will last all day of frequent use.

Dating apps swiped left on Pakistan’s request to clean up their acts, bans followed

"Starlink (in future), or one of the myriad satellite providers in the area (if you can put up with the latency)"

No on both options. Starlink's operators have explicitly stated that they will comply with local regulations (I.E. censorship). So does pretty much every satellite provider, because if they don't, their equipment for subscribers to use the system is illegal to sell in that country. If you do find an operator willing to violate that (spoiler alert, you can't), there's another solution which has been time-tested on satellite TV receivers. Ask Turkmenistan to recommend a good hammer for destroying dishes and they might even throw in the good hammer to use against the users of those dishes as well.