Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Check out the brand new SourceForge HTML5 speed test! Test your internet connection now. Works on all devices. ×

Comment Re:Client side SSL certificates? (Score 1) 45

Why aren't we using more client side SSL certificates, these could be issued by Twitter or something for their purposes. Why are passwords still being used?

It wouldn't matter whether a third party had access to a password or a client cert; they'd still have access to the account. Passwords are only bad because of keyloggers and guessability. When neither of those two is involved in the hack, there's no benefit to using certs.

Comment Re:"Adding no Value" (Score 1) 252

Technically, they only know about the ones who didn't know that they'd save money buy buying the subscription on the website, which is likely to be a dwindling number over time, or at best, a continuous stream of newbies who eventually realize their mistake and start subscribing directly.

Comment Re: "Adding no Value" (Score 1) 252

Your proposed solution sounds good, but my issue with it is that the App Store does not cost Apple $0 to operate. So if they gave away their 30%, that would harm them disproportionately since they would then also have to provide for their App Store team.

The app store is entirely an orthogonal issue. Spotify is a free app, so Apple would still be obligated to distribute it for free even if Spotify stopped selling subscriptions in the app and presented only a login screen, thus forcing users to do a Google search for the website and buy a subscription there. So it doesn't make sense to treat those normal costs of doing business as part of the value added by Apple's in-app purchases.

So by my math, other than the fairly negligible benefit of customer convenience (only updating your credit card in one place instead of two every few years), the only real value that Apple's in-app purchases provide is the purchasing system itself, which is measured in the low single-digit percent territory, at least for customers the size of Spotify. (The benefit is bigger for smaller companies, of course, because they won't get the preferential merchant account rates and people will be less comfortable giving unknown companies their credit cards... but again, that doesn't apply in this case.)

So maybe not 30%, but how about 25%? That would roughly balance the scales.

Comment Re:Apple being excessive (Score 1) 252

Nice straw man. Apple could just as easily mandate one-click cancellation of subscriptions purchased through third-party billing systems as a condition for putting an app in the App Store. Nothing about such requirements requires using Apple's massively overpriced payment system.

Comment Re:"Adding no Value" (Score 5, Insightful) 252

The real problem here occurs when you have hyper-competitive markets like song-streaming. Spotify, AM, Pandora, they all fight every contract and deal they make for fractions of pennies due to the micro-margins and huge throughput in their business models.

That's not really the problem; I mean yes, it is part of the problem, but it is not the core problem. If every vendor had to pay that percentage for Apple's payment processing, those companies could continue to fight for every cent, but they would be competing on equal terms. The problem is that Apple itself is in the music streaming market, and by requiring everyone else to use Apple's payment service (on which Apple makes a profit) instead of much cheaper third-party payment services, it is effectively granting Apple an unfair competitive advantage over everyone else in that space, because they can choose to not take the extra profit on the transactions themselves, and thus can offer their own music service for considerably less money than anyone else.

Now if Apple is willing to give 30% of their Apple Music revenue to music charities to level the playing field, that's fine (and musicians everywhere would love it if they did), but otherwise, what they're doing is blatantly anticompetitive, in addition to undeniably harming consumers by tricking iOS users into paying a higher price for the same goods and services solely because they happen to have downloaded the app and subscribed through the app rather than going to the company's website and buying a subscription before downloading the app. (BTW, I reject the entire notion that Apple "brought those customers to the table". Most people who get apps do so after hearing about them outside the App Store. Almost nobody I know has ever discovered anything in the App Store without searching for it; the store is just too big for that to be practical.)

The thing is, lots of folks pointed out all of these problems way back when Apple first announced their in-app purchase rules, and it has been an ongoing battle ever since. Eventually, it is going to come down to an antitrust suit, which Apple is likely to lose. And that will also be bad for consumers, because the app store rules do have a valid purpose—making it so that users don't have to give out their credit card to every random app that they do business with.

IMO, the only solution that won't harm consumers is for Apple to change their rules so that the "in-app purchases only" rule has an explicit exception for subscriptions to services that are also available on non-mobile platforms via a website. This would basically cover all the interesting edge cases, would cause Apple minimal financial impact in the long run, and would prevent this from turning into an ugly lawsuit where everyone loses.

Comment Almost no surveillance concern at all, really (Score 1) 60

The ability to hide cameras is nothing new, and making them orders of magnitude smaller doesn't change that much; it was already possible to hide a camera in ways that would be almost impossible to discover accidentally. The reason we don't have cameras hidden everywhere is that you have to provide power and a communication channel, both of which take up considerable amounts of space, and thus make it much easier to detect the cameras. This doesn't change that.

Comment Re:warranty length (Score 1) 186

That's great for things that run in a controlled environment and for things you tend not to plug and unplug on a regular basis.

No, it isn't even true in those situations. There are certain classes of late-onset failure that are alarmingly common in computers, most of which involve thermal stress on solder balls (with the GPUs being the worst offenders). These problems don't usually start to show up in large quantities until after the first year, but reduce the machines to doorstops within the first three or four.

IMO, anything with a GPU should have a minimum five-year warranty, and the warranty should automatically be extended by a year every time they repair it. That's the only way it will ever cost the manufacturers enough to get them to put pressure on the GPU manufacturers to fix their designs....

Comment Re:Arr (Score 1) 147

Who pays to retrofit all existing shipping tonnage? How is your new "modern" ship with those unnecesarily complex systems and failure points requiring increased maintenance competitive in the market? Hint: Nobody will, and it isn't.

The question wasn't whether anybody would be willing to do it, but rather whether it was practical to create. If your definition of practical has to extend to whether it would be practical to replace all shipping boxes in the world, then clearly the answer is "no", but that would be true for any solution that anyone could possibly come up with, no matter how simple, because there are just too many shipping boxes in use.

And the reason it probably won't happen isn't technical, but rather statistical; the odds of pirates attacking an unmanned freighter are tiny, and the expected losses are so small relative to the value of the total cargo shipped that the ability to submerge to prevent piracy would not be worth the cost.

With that said, if the lack of security crew on the boats eventually results in piracy becoming more popular by several orders of magnitude, then at some point you'll reach a threshold beyond which it actually would make sense to submerge the cargo to protect it. And at that point, the market will decide that it makes sense, and they'll start building submersible transport ships and sea boxes that are waterproof down to a few hundred feet underwater. Unless and until that happens, obviously, it won't.

Comment Re:Arr (Score 1) 147

First, each container must be airtight and watertight (appropriate gaskets, etc.) to at least 200 feet, and locked using an electronic lock with no external access other than by plugging in a special controller that powers up the lock and uses RSA or EC crypto to authenticate itself to the lock controller. Second, to protect itself against pirates, the interior of the ship must be airtight and watertight to several thousand feet of pressure, and must have appropriate pumps and chambers such that when attacked, it can pump all of the air from one set of chambers into another, while allowing water to take up the space.

Pirates get on the boat. Boat submerges. Pirates end up in the drink.

Problem solved.

Comment Re: News at 5... (Score 1) 451

The truth is the computer has to assess the situation exactly like a human brain using sensors which have a finite accuracy and some uncertainity on the readings.

Yes and no. Yes, it has to assess the situation much like a human brain would by taking inputs and choosing a course of action. No, it isn't at all like a human brain. For one thing, an autonomous car has a lot more information at its disposal than a human brain does. In particular, it has multiple cameras, so it doesn't have to look to see if the left lane is clear for an emergency evasion. It already knows. Those critical milliseconds can often make the difference between a good outcome and a bad one.

For that matter, an autonomous car can have cameras where human vision can't penetrate. It has no blind spots in which other cars and pedestrians can hide. It can potentially see over other vehicles to recognize brake lights that a human can't see. And so on. It knows whether anybody is tailgating it, and can reduce the risk of getting rear-ended by blinking, then braking twice as hard a second later than it otherwise would.

And potentially, cars could communicate with one another, allowing them to know about bad road conditions when they're a mile away, allowing it ample time to slow down before the ice patch. And they could coordinate braking to make it possible for them to tailgate safely, thus increasing traffic density without the slowdowns and backups that we have now (which are a leading cause of wrecks).

So at least in theory, autonomous cars can be orders of magnitude better at driving than any human driver is capable of being, by virtue of being able to see things that people can't and communicate in ways that people can't. The only question is how long it will take before they exceed human driving abilities so much that we ban manual cars. :-)

Comment Re:Duh (Score 1) 168

I thought the USB 2 portion of the USB C cable was dedicated, while the four high-speed lanes are capable of carrying anything?

By my reading of the Wikipedia article, it looks like the alternate mode can take over the slow lanes, too, but it could be wrong, or I could have read it wrong. :-)

I thought the USB-2 only mode was when alternate mode took over all four high-speed lanes, which means you can still run Thunderbolt 40Gbps and get USB 2 accessories on the same cable.

That's correct. AFAIK, Thunderbolt 3 takes all four of the fast lanes, though. I mean, I guess it is technically a multi-lane setup, so if you were using only half the Thunderbolt lanes for Thunderbolt, you could ostensibly use USB 3.0 over it, but I don't know if that's actually an allowed configuration.

Comment Re:Duh (Score 1) 168

The alternate modes have the option of taking over basically every data pin. Whether they do or not is dependent on the alternates. If Thunderbolt 3 only takes over the four high-speed lanes, then yes, I suppose the port technically is still USB, though at that point, it is only USB 2.0.

My point was that in Thunderbolt mode, the protocol it is using to drive high-speed devices is Thunderbolt at that point, not USB.

Comment Re:Good for the Brits (Score 2) 1588

So for all the bluster of how the UK is doomed for leaving the EU, it seems several countries in Europe are doing fine

Doomed is too strong a word, but bear in mind that there's a big difference between never agreeing to join an economic community and telling that community to eff off. If this goes forward, they'll be at the mercy of the remaining EU members, hoping that the EU members don't decide to stick it to them out of spite, and I'd argue that the EU has little to lose by doing so at that point....

Slashdot Top Deals

Unix will self-destruct in five seconds... 4... 3... 2... 1...