Posts by Frumious Bandersnatch

Re: Fast Integral Reactor.

I'm quite suprised no-one has thought of storing nuclear waste on the moon.

That didn't work out so well in 'Space 1999'.

Re: Transducer supply

I was thinking something similar when someone above mentioned radiators. They may be the lowest-power chips yet, but I guess we're not going to see these powered by thermopiles especially if they're an active part of a thermostat system.

The other thought that struck me was the ROTM angle. It's a little bit frightening that they now have the capability of running in "sleeper cell" mode for decades, just biding their time waiting ... waiting ...

Re: de Boise principle

I don't know this de Boise of whom you speak, but it was Cardinal Richelieu who said

If one would give me six lines written by the hand of the most honest man, I would find something in them to have him hanged.

Could mass surveillance possibly have any unintended consequences? Nah, surely not...

Re: Linux

Agree on the Linux comments, but you could still probably get stuck waiting for a disk check ("this disk has gone too long without being checked"...). Probably won't take 15 minutes, especially on a laptop, but it goes to show that Linux isn't totally immune from "randomly" locking you out of the machine for several minutes at boot time.

Kind of off-topic: kexec support in Linux kernel is cool, particularly on machines with very slow BIOS startup (eg, my old Dell PowerEdge)

This could rapidly become "convert an edited photo of their current mechanical meter into a meaningless number" and "cost customers tens of pounds rather than hundreds".

When I read this part of the article I immediately thought of something like SecurID. Have the display present both the units-used counter and some cryptographically-generated hash or time signature (like a HMAC involving the current time, the value of the counter and the serial number of the meter). Of course, once the meter is in people's hands, you're bound to find someone who has the skill needed to hack the device (eg, to find the secret serial number) but the vast majority of people won't.

I agree with the people who suggested the smartphone app idea. Done right it would be vastly better (and more secure) than "smart" meters that need to be networked.

Without wanting to blow my own trumpet (too much), I'm sure I could knock up a proof of concept for this in an afternoon with a Raspberry Pi, a small screen and a mobile phone. Just use QR code libraries on the Pi and phone and you've got a fault-tolerant reader (QR includes error correction) that can automatically send an SMS reading to wherever. Of course, this is probably way too easy. I'd have to massively over-engineer it for the government to have any interest in it.

Re: This wouldn't be (much of) a problem...

The problem with UEFI is that it is _far_ to complex for the problem it needs to solve

Don't worry. We'll all be running systemd on top of it. That'll help manage the complexity.

Right?

Re: Use all resources available

<groans>Yes, and my fly is now a walk</groans>

Re: Random musings

Then again there was at least one programming language that [allowed you] to specify any positive integer as the lower bound of an array.

#!/usr/bin/perl -wl

$[ = 4; # least-random number? or was that 2?

print $#foo;

("but don't do that")

Re: Just like the Jews...

If you listen and believe everything a flaming bush tells you

I haven't said a word for eighteen years till he came along!

Actually, still not bothered.

re: Scylla and Charybdis

Not a response to your post, AC, (though I agree), just a comment to say that this particular classical reference seems to be cropping up in the register quite a lot recently. Or maybe I've just got a case of Baader-Meinhof syndrome...

site:forums.theregister.co.uk scylla charybdis (151 hits)

Re: Financial models matter

(2) Devise a EULA that completely absolves you from any responsibility for the egregious flaws in your software.

Everyone does that. Even (especially?) free software.

Re: Very slippery slope...

are you going to stop a repairman that claims to have been sent by the hotel to your room?

If he's there to fix the porn, probably not.

If the feds had used that simple explanation they wouldn't have to learn any telco lingo at all.

Re: COPS! In the US, the beat you to death. In the Japan, you must confess.

Standard operating procedure over there, I hear.

I don't know why you're getting downvotes, as what you say does seem to be accurate. From the Wikipedia page on the Japanese Criminal Justice system:

Confession in Japanese criminal investigation

Many Western human rights organizations alleged that the high conviction rate is due to rampant use of conviction solely based on confession. Confessions are often obtained after long periods of questioning by police as those arrested may be held for up to 23 days. This can, at times, take weeks during which time the suspect is in detention and can be prevented from contacting a lawyer or family.[3]

And here's a link to a BBC article on the same topic that also suggests reasons why people confess.

These were the two top links when I searched for "japanese police confessions", btw, not just links I cherry-picked to support the argument.

Re: Said with love...

All well and good but how the hell did you get the Delorean in there in the first place?

One piece at a time?

Re: Not Again

And then you start thinking about embedded crud with problematic glibc libraries

Embedded generally means you're not using glibc anyway. It'll probably have newlib instead, or maybe μClibc.

Re: It is

出る杭、not 出る釘。

Maybe they're both right? I just looked up the expression in Jim Breem's dictionary and I got kugi/nail (or spike), whereas Wikipedia's list of Japanese proverbs lists kui/stake. Might be an issue of the phrase "doing the rounds" (going out of one language and then being brought back in in a slightly different form). The different versions might be due to mishearing, perhaps (an example of an "egg-corn", maybe?)? I can imagine "kugi" sounding quite a lot like "kui". The hard "g" sound might not be very distinct with some speakers, with only a slight glottal or nasal sound to distinguish it? Anyway, I don't really know because I very rarely heard the expression in Japan and I don't remember whether it was kui or kugi, to be honest.

"Like IPv4 addresses where one or more octects is often greater than 255."

Shhh! We don't talk about the hidden IPv4 addresses!

Re: So What?

think they may have called it the gambler's fallacy?

Doubling down on a loser is called a martingale (strategy). In probability, the word has various specific meanings, but the term used does derive from the earlier meaning in gambling.

Actually, with infinite resources (and no limit on the bet), doubling down on loser always wins eventually.

Re: And the alternative is ?

Doesn't your argument give weight to the fact that what you just described is actually what most people want from a language? For me, that would be the definition of good.

LOL. Yes, kind of. I guess it is a good language overall, but it's not a patch on Perl, IMO. I just find PHP to be too verbose and boring to actually like it. I think that the original context was about being good for security, among other things, and as I said, Perl's -w and -T checks put it head and shoulders above the competition.

Mind you, maybe I'm a bit perverse in my (programming) tastes. I love constructs like Duff's Device and the Schwartzian Transform and have been known to use them when appropriate.

Re: Timing...

all this will do is slow down the network

Yes and no. If you delay packets by a random amount, then yes, the network slows down. If, on the other hand, you replace a FIFO scheduler with one that merely randomises the queue order, then throughput is maintained*. Slowing down the end-to-end routing of packets through the network like this will impact the users, though the network throughput is unaffected.

* a simple example scheme which has a 1/2 chance of delaying the head packet in the queue if it's the first time it's been seen, and a decreasing sequence of probabilities 1/4, 1/8, 1/16, etc. each subsequent time it's due to be sent will mean packets may wait in the queue indefinitely (with infinitesimal probability), but on average will take 2x as long to get through it (sum of infinite series 1 + 1/2 + 1/4 + ...), not taking the probability of the replacement packet being sent into account...

Re: Hmm

How about some other suggestions via el reg users themselves?

Not a board game, but the Grass card game is fun.

Re: @Rabbit80 - Unfortunately not news ...

for a limited time

But 'cp /bin/bash /some/user/.randomapp/randomfile followed by 'chmod 04755 !$' hardly takes any time and the effects can last indefinitely unless detected...

re: Confucius says ?

Wise man never plays leap-frog with a unicorn?

I'm struggling to guess at the relevance of your comment, but maybe ...