* Posts by isogen74

35 publicly visible posts • joined 30 Aug 2016

Boffins build the smallest transistor, controlled by an atom


Re: 1 atom switched junction <> 1 atom wide junction.

> We 'talk' across oceans at GHz with light that takes 30ms to get to the other side.

... but 30ms is an eternity; if you want to clock at 4GHz you only get 250 *picoseconds* to get from A to B ...

Wires, chips, and LEDs: US trade bigwigs detail Chinese kit that's going to cost a lot more


Re: "which is the thing that China has stolen out from under us."

Western populations have given this to China. Since the late 70s marketing based on price and maximizing shareholder profits became the only game in town.

Stop buying stuff made in China and corporations will soon learn what their customer base values.

You should find out what's going on in that neural network. Y'know they're cheating now?


Re: E.g. The infamous 'Compas' black box sentencing tool

The first neural layer in the AlexNet image classifier (now considered relatively simple) contains over a million parameters. It has 16 layers overall. How on earth do you intend to review that in a spreadsheet ...?

Industry whispers: Qualcomm mulls Arm server processor exit


Re: RISC-V is the future

I suspect the issue is exactly the same issue as this article. The problem isn't the instruction set - the problem is getting the software ecosystem built around it.

What does RISC-V bring that Arm doesn't have already? As far as I can tell, the only benefit is that it is "free", so the saving is a little bit of licensing cost. However, in the grand scheme of things the Arm fees are pretty inexpensive compared to everything else it costs to build a server grade processor (best guess, given their publicly stated revenue breakdown and shipped CPU volumes every year, so I doubt anyone with enough money to build a competitive chip here really worries about it - it's in the noise.

What does RISC-V not have? A mature architecture (it's fragmenting rapidly with extensions), or any form of mature software ecosystem. The Arm architecture is mature, and has been around at scale for at least 20 years, and yet according to this article is still struggling with a mature-enough software ecosystem in new markets.

Hell, it's exactly the same problem as mass-adoption of desktop Linux too, or VHS vs Betamax. Just because something might (and that's still unproven) be technically good doesn't mean it's going to be well adopted if the alternatives are easier to use and/or marketed better and/or better established.

Amazon, LG Electronics turned my vape into an exploding bomb, says burned bloke in lawsuit


Re: Batteries in the pocket, eh?

Shorts ... I.e. what happens when you put batteries in a pocket with loose change and keys ...


Re: Batteries in the pocket, eh?

Well, the article did say he was wearing shorts ...

Missed opportunity bingo: IBM's wasted years and the $92bn cash splurge


Re: Puzzled

They've been doing amazing stuff with silicon for years, and completely failed to successfully monetize any of it (to the extent they had to effectively pay GlobalFoundaries to get the foundry business and associated liabilities off their books). Silicon R&D is exceptionally expensive, and without fabs its hard to see how they will ever get any return on that investment.

Forget One Windows, Microsoft says it's time to modernize your apps


Fluent Design System

Lovely themes, all of which make some effort to add depth to the UI controls. However, I'm not sure that it's true to say this is reflected in W10 or many of the other Microsoft apps. They still seems to be stuck in the metro mindset of everything being flat 2D with no contrast, so it's impossible to know what's a UI control and what's not, which is the complete opposite of the "Fluent Design System" principles proposed here.

Yet another flip-flop reversal guaranteed to confuse users who have just started getting used to the last UI design "revolution".

US Congress mulls first 'hack back' revenge law. And yup, you can guess what it'll let people do


Machine != Hacker

The main issue as I see it is that most hacks are bounced via other compromised machines first. As noted on the classic movie "Hackers", you don't hack a bank from your house. ... because thats just stupid. If you allow retaliation attacks then really you're in all likelyhood just setting people off against machines owned by other "good guys". The "bad guys" are long gone.

Smart cities? Tell it like it is, they're surveillance cities


Re: Cough

If you took the entire cost of the the London CCTV infrastructure and staffing and invested it in the NHS you'd probably save far more lives. It's not about whether it saves lives, it's about the opportunity cost of what else you could do with the money.

Google's Hollywood 'interventions' made on-screen coders cooler


So what you want is civil engineers and planning officers then. Your average architect will spend their entire life doing loft conversions and garden grabbing extensions ... they don't need a degree, and the builder will ignore the drawings anyway.

Chinese smartphone cable-maker chucks sueball at Apple


Re: Beijing Intellectual Property Court

They also own huge amounts of natural resources (either in mainland China - like rare earths - or by buying up mining companies and mining rights around the world - like huge chunks of Africa, or by funding large project - like new UK nuclear power plans), and also own huge amounts of Western government sovereign debt, so it's unlikely any of the western governments are going to do much to directly challenge China here ... it will all be softly softly talky talky.

Raising minimum wage will raise something else: An army of robots taking away folks' jobs


Re: isogen74

> "On the other hand those "cheap imported goods" have been a lifeline for 4 billion people to lift themselves out of poverty in the last 35 years."

Oh yes, I definitely agree. I was really just pointing out the hypocrisy of people who buy goods from overseas where there is a lower standard of living (e.g. no pensions, longer working hours, less holiday, lower pay, less heath and safety law, fewer worker rights), insist on higher pay and better conditions for for their job, and then complain in the Daily Mail that all the jobs are going overseas. As seems to popular in politics these days, they "want to have their cake, and eat it".

> Most people I've encountered will quite proudly pay slightly-higher prices to support locally-owned stores if they can afford it and all else is equal.

In my experience it's never equal. Supermarkets and Amazon are easier and faster than treking around 5 smaller indie shops. Most people (including myself) fall into the "I'd love to buy local in principle, but I don't have (or am not willing to make) time" camp. I'll admit I'm lazy - in my case it's a "not willing to make time" rather than a "don't have time".


Re: corporate conscience

I think you took an overly narrow interpretation of the word "local". Given the context of my post "local" = "somewhere with the same standard of living that the buyer has".


Re: corporate conscience

Evidence seems to show that whenever a consumer is given the option to "buy local, but it will cost you 10% extra and take 25% more time to get to the shop and back", or "buy something manufactured in <insert foreign outsource destination here> where there is a lower standard of living than yours, from <insert mass market retailer>", they will nearly always buy the cheaper option.

They will then proceed to complain bitterly that jobs are going overseas.

The issue isn't the companies really - capitalism will generally give consumers what they pay for and if they stop paying the companies would fold quickly because few hold any significant cash reserves.

The underlying issue is that consumers are not really willing to pay the costs of goods produced where workers have the same standard of living that they think they are entitled to. If people actually stopped buying cheap imported goods, the companies would change PDQ.

ICO fines Morrisons for emailing customers who didn't want to be emailed


Re: "We sent out an information message"

It's not even uneconomical. Charging 8p a customer is less than it costs to send a postal marketing campaign, which Virgin Media seem to be happy dropping though my door on a weekly basis.

Huawei Honor 8 Pro: Makes iPhone 7 Plus look a bit crap


> The headline feature is the battery life. For the second time this year I’ve found myself trying to think of new ways to exhaust a phone, and been stumped.

For the most part use the wifi and 4G data connections and watch the battery vanish. Modern phones and display panels are marvelous things and increasingly well optimized, but the physics of shoving a high frequency signal at an aerial some distance from the device is always pretty punishing.

Linux kernel security gurus Grsecurity oust freeloaders from castle


Piling in against someone who doesn't seem to comprehend the ramifications of the license they released their code under seems entirely justified. If anyone releases products with a GPLv2 kernel they are *obliged* under the terms of the GPL to make source code available to those they sell the products to.

If you release code under a GPLv2 license it seems a bit rich to complain when people actually use it under the terms that the license allowed ...

Boeing details 'Deep Space Gateway' for Mars mission staging


Re: More USS Gravy Train than USS Enterprise

> Spot the difference between isogen74 and Edmund Hillary.

Hillary's trips were amazing, sure, but more or less self funded by the organizing committee, donations from suppliers. Getting humans to Mars is likely to cost a significant fraction of a trillion dollars.

You can do a lot of other useful science for half a trillion dollars, much of which is likely to be significantly more useful in the longer term ...

To be very clear, I'm not saying don't explore. I'm just saying understand the opportunity costs involved, rather than just pouring money into ULA.

> Assuming we get 66% savings from mass producing Curiosity, it would cost roughly $4.5 trillion to cover the same amount of ground a human can.

Why is covering ground a useful measure of success? If the ground is the same for miles in all directions (not unlikely given how Earth geology works) then the amount of new knowledge isn't proportional to distance covered. If you're unlucky (e.g. not enough scout rovers sent first, so you're effectively picking landing spots based on educated guesses) then you land in the middle of a region of very similar geology. Ability to carry food, oxygen, and radiation exposure alone says you can't actually go far from home base in terms of time, even if you could theoretically travel the distance.

> The problem with robots is they have to be built for a relatively narrow range of tasks or they become too unwieldy and still don't have as much flexibility as a person.

So build two or three specialized ones for different purposes. You can build 50 and run them for years for the cost of the manned Mars project, and at least you spread your bets.


Re: More USS Gravy Train than USS Enterprise

> If the goal is a permanent foothold in space - i.e., colonization - then robots have yet to equal humanity at making babies. ;)

The important part there is the *if* at the start of your sentence.

Colonization has to have a purpose - typically (1) a need for more space to grow population or (2) acquiring an environment or resource you can't get more cheaply elsewhere, while at the same time needing (3) not getting cooked and DNA scrambled in the attempt.

Inside the solar system, what's the point? Camping half a dozen astronauts on Mars doesn't have such a purpose. No solar body other than Earth really works for (1), and we've yet to find any evidence of a resource which meets criteria (2). Even if you found something dense and expensive, say a ton of pure platinum, it's only worth $24 million at today's prices. If you actually managed to find and haul back 100 tons of it (good luck getting that out of the gravity well), then you've just doubled world production and crashed the price. At best a manned trip to Mars is a very expensive photography trip ...

** About the only possible advantage I can think of for space industrialization is building microgravity manufacturing at scale. In this case Luna makes a lot more sense than Mars as it's lower gravity, and nearer ...

> Manned missions are also currently faster and more adaptable. The years of science from Spirit and Opportunity could've been generated by a human geologist in a few weeks. (The counter point is obvious: you can launch a lot of robots for the price of one human on Mars. You just need lots of patience with the robots.)

I think that's exactly the point though - the economics for manned missions just don't add up, you can launch a hundred complex rover missions for the same price. Plus, at the moment all you're going to get out of a human landing are some nice photos and some geology reports. All of your cost goes on keeping the human alive, not on science ...


More USS Gravy Train than USS Enterprise

... see title.

No one has yet explained what a manned mission will be able to do which a robotic mission cannot, except cost a lot more and therefore replenish the gravy train ...

US ATM fraud surges despite EMV


Re: Speed

> At one of the local grocery chains, payment went from barely longer than it took to swipe the card, to something like a 45 second wait for C&P

I suspect this says a lot about how much investment went into those C&P machines (or the online pin verification infrastructure provided by the bank). Honestly should be under 5 ... i.e. across Europe you usually have enough time to put the card in the machine and enter your pin, with no sitting and waiting.

* wireless machines not counted here - they are normally on some shitty GPRS side band for verification which is always as slow as shit. Anything with high throughput requirements should get wired EMV kit.

New plastic banknote plans now upsetting environmental campaigners


Re: "300 football fields of rainforest per hour is cleared"

> From this wiki entry, I find that a football field is 7140 square meters which, with the magic of the metric system, means 7.14 square kilometers.

Um. No. The magic is in the word "square".

One square kilometer = 1000 meters * 1000 meters = one million square meters. So a football pitch is 0.00714 square kilometers, which makes your whole argument fall apart somewhat. But hey, what's three orders of magnitude between friends. Don't let that get in the way of a good rant though - carry on ...

Tosh in deeper financial doo-doo as banks crank up the pressure


Re: This could be interesting for...

> eh?

CBS sold the nuclear power bit of Westinghouse to BNFL, which was wholly UK government owned and run. BNFL started trading at least partially under the Westinghouse Electric Company brand at this point. WEC was sold to Toshiba in 2007, under approval from Brown who was busy selling off anything which wasn't bolted down to at least try to cosmetically look like "prudence".

BNFL Plc itself is now mostly defunct (shell for the pension fund, nothing operational), the other bits have been either sold off too, or relate to the Sellafield clean up operations which has been spun off as Sellafield Ltd.

New Royal Navy Wildcat helicopters can't transmit vital data


Re: WTF?

"Prudence" Brown was quite happily borrowing £1100 per man, woman, and child in the UK per annum *before* the sub-prime bubble exploded as well as selling off government property and gold at rock-bottom prices. His ability to spend far in excess of tax receipts has nothing to do with sub-prime nor the credit crunch, the crunch just exposed it as the unsustainable "economics" that it was ...

Fresh Euro Patent Office drama: King Battistelli fires union boss


You have to wonder ...

... what skeletons he knows about, given that he's lost all ability to influence the staff he is supposed to be leading, especially given that this has been rumbling on for years ...

Brexit judgment could be hit for six by those crazy Supreme Court judges, says barrister


Re: Parliament must vote

Even if you managed to avoid a vote on Article 50, there is a huge amount of EU legislation tied into the UK legal system which would need new UK laws to determine what to do with, and thus a vote and an Act of Parliament to do something about. It's impossible to enact Article 50 without some vote on something with MPs, even ignoring the EU aspect of this, because triggering Article 50 without some plan to clean up the UK side of things would be nuts and leave the UK in a legal limbo ...

* Note that the current plan of record is "The Great Repeal Act", which actually amounts to "accepting all of the EU always already on the books and worrying about cleaning it up afterwards" rather than actually repealing everything. General legal consensus seems to be that it'll take a couple of decades to review and repeal individual laws (or keep the ones we actually want to keep).

Run a JSON file through multiple parsers and you'll get different results every time


Re: not parse JSON documents that I hadn't created myself

If you have a public facing website people can POST all sorts of shite at it, JSON or otherwise, whether you like it or not. Validating your inputs from untrusted sources still applies (there are good JSON parsers our there which can handle invalid inputs safely).

I'm not sure JSON is any worse in this regard than any other form of data upload; it's just an arbitrary string after all ...

New measurement alert. The Pogba: 1,200Pg = NHS annual budget


Ensuring future consistency

Given the time-sensitive nature of the value of 1 Pg, I propose he is cryogenically frozen and stored in the International Bureau of Weights and Measures alongside the International Prototype of the Kilogram. We can thaw him out, give him a clean, and let him have a kick about occasionally to ensure future accuracy of the value.

Intel XPoint over-selling criticism surges as Chipzilla hits back


SSDs didn't really change anything, other than adding a few interesting catastrophic failure modes when their firmware goes AWOL. Mass storage drives are always unreliable, and have been since they were first invented. Stuff on them is nearly always valuable. Therefore you have a decent mirroring / backup strategy, with recovery time dependent on the value / importance of the data.

XPoint won't change that - data loss is almost never an option, and even if the drive doesn't fail, the building might get hit by lightning, catch fire, etc ... so you still need backups and second drive.

Honestly there isn't really a distinction between a good drive and a bad drive - the difference between 99.99% and 99.999% reliability is immaterial if your business *relies* on that data you can't afford to be the people in the bad luck zone (either because of bad drive, act of God, or human error).

> as fast and reliable as the 3D XPoint is going to be.

Reliable. Irrelevent - see above.

Fast - either provide data, or sit quietly and wait. We've been promised the "next big thing" in both DDR and HDD replacements for at least the last 20 years. Most of it is expensive vaporware which never gets out of R&D and into mass production; even if the numbers are great, if you can't mass produce it on cost and with high yield then it's DoA.

Double-negative tweet could be Microsoft Surface Phone hint


Re: Um...

A 6W Intel part in a phone ... nope. In a phone form factor you get 2-3W out of the main SoC; much more than that the chip will either melt or throttle horribly. So you might get that chip running at half frequency, but it's going to be pretty unimpressive from a performance point of view ...

Google-funded group mad that US Copyright Office hasn't abolished copyright yet


Re: Your paranoia is showing, Andrew

> Believing that the job of the USCO, and all other the administrative and legal arms of the government, is to balance the interests of rightsholders and users, not to "further the interests of rightsholders".

No, the role of administrative and legal arms of the government is to implement the law as written. If you allow "balance" and "human opinion" in there, then your government will be swayed by money at the bottom level. If you don't like it, change the law (as I suspect it should be - lack of sense is a problem here), but don't give government offices choices about when or when not to apply the law - that's just anarchy under a different name.


Re: IIRC and Usenet

> The copyright laws have proven incapable of preventing this

Making murder illegal does not stop people killing each other, but it does mean that there is some legal redress/punishment when you do. That may encourage prevention (cost of punishment > value of crime), but that isn't the sole purpose of the legal system.

... and besides the point here isn't random individuals copying a book, it's big corporations wanting to effectively usurp individual rights to make a profit (at the expense of those who held the rights in the first place).

Deep inside Nantero's non-volatile carbon nanotube RAM tech


Re: All change

> The implications of memory technology which is a) faster than dynamic RAM and b) non-volatile, are considerable.

Yep. It means all of the technology will crash even faster, and now has the added benefit that turning it off and on again won't fecking help.

Intel's makeshift Kaby Lake Cores hope to lure punters from tired PCs


Re: Wrong target . . surprise, surprise!

"Hollywood bosses didn't want to stream 4K ultra high-def content from their studios' websites without mechanisms in place to thwart casual rippers"

... so they went back in time and screwed the telco network competition law so badly that few of the populous could even stream 4K even if they wanted to.

If you can watch it you can't pirate it ...