Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×

Comment Re:100 mhz per core (Score 0) 44

A 100mhz might not seem like much until you give that to 12 cores and 24 threads.

Well you're not, at 12 cores loaded you're thermally limited and tests show no difference to speak of at all. You get a slight boost to single threaded performance but overall this is the dullest upgrade in ages and creating a whole new XT line for it is silly. Should have just called it 3910X, 3810X and 3610X, at least then they'd get points for honesty.

Comment Re:free speech (Score 3, Insightful) 140

Oddly enough I care less that China knows about my personal life than the United States. as China has no jurisdiction over me and what I say or feel. America while I am suppose to be protected by the first amendment, doesn't mean there are people with power who can actually make my life difficult.

Jurisdiction is not the only kind of power. China has repeatedly shown that they don't care if you're an American speaking out in America, if you're badmouthing China they will do what they can to make your life miserable. Of course they'll wrap their demands in weasel words and so will the corporate chain and your business partners but the real message is "STFU or we'll lose our business with China". Like any negotiations will be mysteriously stuck in a quagmire that'll equally mysteriously resolve itself once you're relieved of your position or the partnership is terminated.

Comment Re:Apple just doing what they are being told to do (Score 2) 72

Indeed but Taiwan is not exactly completely independent either. China can turn those screws if it needs to.

To be honest Taiwan's problem is that the had delusions of grandeur, in 1971 when they lost their seat on the UN Security Council they themselves strongly refused any division of China and the creation of an independent Taiwan. Instead they forced it into a vote of who was the "legitimate" owner of China's seat and long story short they lost. If they wanted independence back then, they probably would have gotten it. Their problem now is that China has flipped the script on them, because they were pouting over the loss of the mainland China they never cut their losses and got their independence so now China wants them to answer to Beijing.

They don't realistically have much political or public support in Taiwan though. China is pressuring every other nation to recognize China's claim to Taiwan if they want to have relations with China, but that only makes it a hostile territory they could occupy at great cost - and probably not without an actual shooting war with an army supplied by the US military. My guess is they're eyeing the situation in Hong Kong waiting for the right moment to say status quo is no longer an option, it's either independence or submission and we choose independence. They must see that there's no future in trying to appease China.

Comment Re: Time != Money? (Score 3, Informative) 44

Write no, compile yes. He's pulling code from all the subsystem maintainers (lieutenants) to create the master branch. And they again are mostly pulling other people's patches. They're supposed to make sure it's clean and tested before it reaches Linus but that's not always what happens. If it breaks or he finds bad code it goes back down the chain of command again so they can fix their code again and resubmit.

Comment Re:Going forward (Score 4, Interesting) 33

In rocket science I'd say 10+ launches is well into established already. The Falcon Heavy got their military certification after just three launches. The SLS is still planning to launch crew on their second flight. I suspect your sig is very relevant here:

If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...

My guess is one faulty part or installation brought the mission down. This was not a violent failure, the rocket just lost engine power and stopped. Which means it's probably the QA department and not the design department that needs to get to work.

Comment Re:That's an interesting coincidence (Score 1) 93

I think the most important thing I learned, though, isn't related to either of those fields, and isn't nearly as complicated as either. I've done plenty of dumb things. Most, probably nearly all, of my fuck-ups have been based on short-term thinking, doing what I feel like doing in the moment. (...) I hope to continue to teach my daughter (and myself) the value of delayed gratification, of doing things that are going to make you happy NEXT week, next year, or five years from now, as opposed to what we feel like doing right this moment.

Funny, I wish I had more of the first bit. Not my fuck-ups but the best times are those where I've been living in the moment and let go of all the rational, dull downers like that this junk food isn't good for me or that partying this hard will give me a hangover tomorrow or that this vacation is burning a hole in my wallet. Like I know that I'm overall fairly rational and got my shit in order, can't I get a break when I have a few excesses without that nagging party pooper in the back of my head. It's okay to have a plan for where you want to be five years from now but it's also important to accumulate happy memories on the way.

Comment Re:Bad Managers (Score 1) 79

Bad employees also tend to think they're irreplaceable and they're as common as bad managers. The difference is that bad employees are the manager's problem, a bad manager is your problem. Unless you become the manager, then you see the other side of that equation. I don't mind taking on a tech lead position but if I went into middle management I'd probably end up strangling someone. I have the feeling I'd be one of those technically bright but horribly micromanaging bosses who'd rather be doing my underlings job than doing mine. Which would be mostly herding cats, running interference and communicating upwards, downwards and sideways in the hierarchy.

Comment Re:Much respect for Linus Torvalds (Score 1) 63

Just because something is widely used it does not mean it is actually good. Prominent examples are x86 CPU's or Windows 3.1.

Intel drank that kool-aid and decided to make a CPU architecture where practically everything was exposed to the compiler, it's called IA-64 aka Itanium. They soon found out that a design grounded in reality beats ivory tower theory. ARM is also CISC with micro-ops these days, some instruction complexity is best left hidden. Forcing the compiler to deliver micro-instructions only hurts performance.

Comment Re:That's nice (Score 1) 285

The reason I use whitelist and blacklist is that they're not the same as what's allowed/denied. If you implement say a spam filter and somebody wonders why something does or doesn't get through then "it's whitelisted" or "it's blacklisted" are complete explanations that someone has overridden the regular spam filter that'll also allow and deny emails. If the rules are exhaustive there's no reason not to be more direct. I have no strong opinion about master and slave, I tend to use manager/worker anyway since that's been the most relevant structure for me. But as long as people don't go crazy on the euphemisms we'll be fine. Rudolph/reindeer anyone?

Comment Re:There's something else going on (Score 3, Interesting) 98

Ads are quite effective for the things you know relatively little about and in general we don't have the time or interest to research most things in our life. For these types of things I often don't even consider alternatives, it's like I have a problem and product X will solve that problem so I buy X and the problem is solved. Now maybe product Y or Z was better or cheaper, but I just reached for the first possible solution. That way I can get on with the rest of my life so I consider this rational even if it's sub-optimal.

What's the first thing that pops into your mind? It's a brand that you've seen a lot, with a memorable catchphrase or jingle. We like to tell ourselves that we don't care about the ads, but the practical reality is that we'd never bother seeking out their store on our own. If they didn't advertise they wouldn't be on the radar. At the grocery store a big poster with an introductory offer can be the difference between "meh, I'll buy the usual" and "okay, I'll give it go... doesn't cost me much to try". Notice the lack of enthusiasm, if I had that I'd buy it anyway. Ads is limping it in.

Comment Re:a little elaboration on the article. (Score 2) 87

There's nothing wrong with PGP as such, but one feature/flaw depending on how you look at it is that if the key is compromised your entire communications history with everybody can be decrypted if they have the encrypted messages. Which is not unlikely if it's sent from or received by any of the big free email providers, even if you "delete" them they could probably be recovered if you get them to dig through backups for you. The ratchet algorithm that Signal (and others) use is using something almost, but not entirely like session keys for the communication. Basically you update the send/receive keys continuously so only the keys to decrypt the next message is available and then they're discarded. That way nobody can do back and decrypt old dialog, what's gone is gone.

Comment Re:Statistics and damn lies (Score 1) 80

If your tool has a 50% failure rate, you can flip a coin and get the same results.

That's not how it works at all. If you got four people in a line-up there's a 1/4 = 25% chance you'll find the right person by coin toss. If I can make a tool that's right half the time I've doubled that. Along the way we're redefined the question from how hard the underlying problem is to how successful we are at it.

If I could make a machine that'd give me the lottery jackpot numbers half the time I'd be ecstatic. Sure, I'd only win every other week but given that it's one set of winning numbers and hundreds of millions non-winning numbers each week I'd gladly take the 50% failure rate.

If you take a database of ~10000 celebrities, your coin toss would have 0.01% accuracy by luck. Modern facial recognition algorithms get about 99.8% so like ten thousand times better. The problem is they want to find your face in a database of 640 million photos. One needle and 639.999.999 straws.

Comment Re:Disney+++ (Score 2) 57

Hmm...I wonder if they'll start "capturing" peoples' images, and then kill them so they don't have to pay them royalties

Why bother? There's billions of us, you can buy the rights off a nobody with zero acting experience for next to nothing or just create a generative model that's a blend of people but not anyone in particular. Once you got virtual actors you make your own stars.

Comment Re:different use case: apply to low res source (Score 1) 57

There's already PULSE for that, but the problem that you can see in figure 3 is that even with near perfect photo realism there's an ambiguity in facial expression, lip motion etc. which would probably go straight into the uncanny valley as being overly neutral in expression, lacking smooth transitions and lip sync. So you would probably need a more advanced model to fill in the missing motion detail. Also this model is only for photo realism not identity preservation, so it's the natural image manifold of people and not the natural image manifold of a particular actor. However adding an identity loss to the existing model shouldn't be that hard, assuming high resolution reference photos are available.

Slashdot Top Deals

He has not acquired a fortune; the fortune has acquired him. -- Bion

Working...