* Posts by tfb

711 posts • joined 11 Jan 2010

Page:

Facebook blames 'server config change' for 14-hour outage. Someone run that through the universal liar translator

tfb Silver badge

Re: Optional

Yes, that.

tfb Silver badge
Terminator

Re: Not sure the comparison is valid

Oh the irony. You do realise that, in days of yore, there was only the Post Office and every company relied on it for all their comms needs?

There is nothing new in a business being dependent on single comms provider.

What is new is relying on a comms provider who you do not pay for their services, and with whom you therefore have no contract, because people paid the Post Office for their services that payment taking the form of stamps. And if you cared enough you paid for (and still can pay for) special magic extra-expensive stamps which meant that they promised to deliver your message by a given time, to the right person, on pain of financial penalty and also promised to pay you some amount if they lost your pack(age|et). And they did do that: I've claimed on those promises and they work.

But now we all, apparently, think that it's fine to complain bitterly when some service for which we do not pay, and with which we do not have a contract, falls over. Hint: it's not, we've just all been fuckwits to assume that there is not only a free lunch, but that it's an endless buffet, and no-one has poisoned the wine.

Here's the thing: the free lunch is not endless, and the wine is poisoned.

tfb Silver badge
Alien

Re: Not sure the comparison is valid

Their customers pay them. Advertisers.

tfb Silver badge
Boffin

The network is the computer

From the article:

The other big question is how a "server configuration change" led to not just Facebook but also its Messenger, WhatsApp, and Instagram services going down. That would strongly suggest that Facebook has either connected them up or attempted to connect them up at a low level, merging them into one broad platform.

While it's always nice to spread fear and paranoia like this, there is a rather obvious answer which does not involve such exciting scariness: it's pretty likely (and also well-known to actually be the case) that Facebook, along with other large organisations, don't send their backhaul traffic over the public internet. Instead they have a lot of physical connectivity that they've bought or leased, and they send traffic over that network. And, obviously, they send all their backhaul traffic over that network: Facebook's traffic, WhatsApp's traffic & everything, because why wouldn't they do that? If that network gets fucked up, then everything that uses it falls apart, and this outage has all the characteristics of some internal network outage: someone made some configuration change to a bunch of switches or routers and everything fell apart catastrophically, we've all been there, right?

Sharing a common network infrastructure is not 'merging [distinct services] into one broad platform' any more than the fact that my web browser and NFS client share the same network in my house is 'merging them into one broad platform': they're just sharing some infrastructure.

tfb Silver badge
Alien

Re: Optional

It comes from advertisers. And they, being Facebook's customers & having contracts with Facebook, are the people who might actually want to complain about this.

It's more than faintly absurd that people seem to be seriously suggesting that some service (WhatsApp, say) which they don't pay for, which has no stated availability level, and which could simply be withdrawn at any moment, is now so vital to people that is must be regulated like a telco. What kind of stupidity got us to this point?

Holy sh*tsnacks! Danger zone! Edinburgh Uni's Archer 2 super 'puter will cost a cool £79m

tfb Silver badge
Boffin

Re: For that money...

Just in case this is a serious comment: yes you could (and it's reasonably likely that some future HPC systems will have really stupid numbers of relatively wimpy cores), but what costs the money is the interconnect.

UK joins growing list of territories to ban Boeing 737 Max flights as firm says patch incoming

tfb Silver badge

Re: The reason that the Max series need MCAS

Oh, I didn't mean the pilots should not know about MACS: what I meant was that the failure mode of 'I give up, it's your plane now' tends to result in bad things when people become very used to having the system help them. I'm not sure there's a way around that: I expect the answer is that very automated planes do have these nasty failure modes and will occasionally crash because they drop into some manual mode and the pilots fail to cope. That does not mean they're not still safer, of course.

In the case of MACS people seem to be saying there are only two sensors so there's an obvious split-brain possibility: having three would make it a whole lot safer it seems to me.

tfb Silver badge
Trollface

Re: The reason that the Max series need MCAS

Heh, I like the silent downvotes!

tfb Silver badge

Re: The reason that the Max series need MCAS

If, in order not to hit the car in front of you (or have to steer round & get killed by whatever is coming the other way) you have to pump the brakes or equivalently if you have to brake hard enough to trigger the ABS so it will do that for you then you are too close.

tfb Silver badge

Re: Panic

Bah, 0.5% sorry.

tfb Silver badge
Boffin

Re: Panic

Two accidents is 1% of the aircraft that have been built (376). In two years of service. That is, in fact, numerous.

tfb Silver badge
Boffin

Re: Panic

Grounding the aircraft is not jumping to conclusions: grounding the aircraft is saying 'we don't know what, if anything, is wrong with them but guven the statistics there might be something, so let's do the safe thing and not fly them'. And this will, in fact, 'stop more shit happening'.

tfb Silver badge
Alien

Re: The reason that the Max series need MCAS

That's like a worse version of what killed AF447: plane suddenly panics and drops into some manusl mode, pilots don't understand what's going on, everyone dies.

tfb Silver badge
Alien

Re: The reason that the Max series need MCAS

So, wait, you were close enough to the car ahead of you that you needed special braking techniques not to hit it? Hmm.

This is the Send, encrypted end-to-end, this is the Send, my Mozillan friend

tfb Silver badge
Big Brother

It makes it significantly harder to track what's happening. For instance if I'm snooping metadata on two people communicating using some end-to-end encrypted system (say mail with pgp), then if I see fucking enormous messages I know something interesting is going on. If I see only small messages I may not: I also have to notice the corresponding uploads & downloads. It doesn't make it impossble for snoopers, but it makes it harder.

tfb Silver badge

Re: The devil with all this stuff is in the details

It does not require you to sign in. You get additional storage if you do, but that's it.

You've been dying to know. Here's the answer: The Milky Way tips the cosmic scales at '1.5tr' times mass of the Sun

tfb Silver badge
Boffin

Re: MOND ?

It's not the case that we've looked at 98% of the places WIMPs might be: we've looked at 98% of the places WIMPs predicted by supersymmetry might be: I don't keep up, but supersymmetry is either dead or nearly so because of this.

I'm not claiming GR is correct: obviously it's not. I am claiming that fundamental theories tend to be pretty, and that TeVaS is anything but.

tfb Silver badge
Boffin

Re: MOND ?

Inspired by your comment, I looked at the Wikipedia page for MOND, and then followed that to the Tensor-vector-scalar gravity page, because I was once a GR person so I'm obviously interested in that.

And, yuck. I mean, seriously, that makes me itch. 'Let's just add in some extra fields, some of which don't take part in the dynamics and a great mass of extra parameters to hold the thing together and call it a theory'. GR is this beautiful jewel-like thing which you can never really forget once you've seen it because it's just so pretty and right, and this is ... what, I don't even know, some kind of machine someone has bolted together from bits and some of the bits have imperial threads and some have metric and they've just used a really big spanner to force the thing together in some horrible, cross-threaded way and oil is leaking out of it and it's also on fire in various places. If physics works like that then God is mad in a psychopathic-axe-murderer kind of way.

tfb Silver badge
Boffin

Re: Supermassive black hole at the center of the galaxy

I don't think Sagittarius A* has ever been thought to be billions of solar masses. There are certainly SMBHs with masses in that range, but not ours.

tfb Silver badge
Alien

Re: MOND ?

I think MOND is a perfectly reasonable hypothesis. It seems to me to be a lot more radical than the hypothesis that there is dark matter, because if it turns out to be true then GR must be in, at least, big trouble. As best I can tell MOND also turns out to fail experimentally pretty badly.

So, like you, I'm left wondering why it keeps coming up. Obviously there is still room for doubt (I have read articles by people with good credentials explaining why the bullet cluster might actually not kill MOND), but I don't think that's why. I think it's because MOND has become a secret poster child for the whole anti-science conspiracy-theory thing that is slowly killing us all: if MOND is correct then a huge chunk of modern physics falls, and, of course, modern physics is a giant conspiracy by liberal elites to suppress the truth and keep the white working class poor I mean look at that Einstein what was his family background eh people with big noses knowwhatImean its a conspiracy I tell you they don't want us to know the truth it's just the same as the climate change they lie to us about nothing wrong with coal they should be rounded up and put in camps the lot of them the germans had it right you know although of course they never did what people say you know people where's the proof [...].

So I think that's why.

tfb Silver badge
Boffin

Re: Aren't planets dark?

They are relatively dark, but I think they're not candidates for dark matter any more. Collectively these things are known as MACHOs (massive compact halo objects), which includes planets not attached to stars, black holes and various sorts of very dim stars or failed stars. You look for these things by gravitational lensing ('gravitational microlensing') effects as they pass in front of luminous objects. And we've done that and we don't find nearly enough of them to account for the missing mass.

So, this is a good, testable, idea, people looked, and they weren't there.

tfb Silver badge
Boffin

Re: MOND ?

Dark matter obviously hasn't been observed directly, but the bullet cluster is a lot of nails in the coffinof MOND theories.

UK banking was struck by one IT fail every day for most of 2018

tfb Silver badge
Boffin

Re: IT Failure?

They also have a problem which many technology companies don't have: legacy. Systems whose direct ancestry (not the hardware, but the platform) dates back to the 1970s are common, and those systems must work. Compare that, say, to Google, who might have platforms dating back to 2000 or so, and for almost all of whose systems occasional failure is fine (who even knows if a search returns the 'wrong' result, or if gmail is offline for a few people for a little while?). And Google are also completely happy to just burn their users if they decide something is boring & have done so repeatedly: a bank can't just turn off the system that supports some saving account they once offered, perhaps for a generation after they closed it to new accounts.

Banking is actually a hard IT problem. The banks are not doing a great job of solving it, but that doesn't make it easy to solve.

tfb Silver badge
Pirate

Re: Cashless society?

Well, either you entered into whatever arrangement it was knowing that this could happen (or you just did not read the fine print) or they have broken their contract with you / the law in which case you have recourse. Since you state you have no recourse I must assume the former.

tfb Silver badge

Re: Cashless society?

IceSave is kind of the canonical example of why people should bank with banks covered by the FSCS scheme.

tfb Silver badge
Terminator

Re: Cashless society?

Well, banks do a huge amount of work to make sure that doesn't happen very often. Sometimes it does, but it's actually pretty rare, despite this article: not very many of these serious incidents will have involved ATMs being down or inability to handle transactions (some will, just not very many). It seems less rare than it is for the same reason people being eaten by sharks seems less rare than it is: every time it happens it's in the news (an average of a little over 7 people were killed by sharks per year over the last 60 years).

tfb Silver badge

That's a good point. Pretty sure that NW and RBS share infrastructure more-or-less entirely, Ulster I think so although not quite so sure.

tfb Silver badge

Re: Cashless society?

If your bank is absconding with your money and not letting you have it then you should probably be taking legal advice.

tfb Silver badge

Re: IT Failure?

That's why banks generally are not doing this for core systems, or if they are they're doing it which very fierce contracts around it.

UK's beloved RNGesus machine ERNIE goes quantum in 5th iteration

tfb Silver badge

Re: Hang on a minute

They don't in fact need to be in any particular order at all: if you just have a list of them, in any order, and then pick random entries from that list you're fine. I suspect the problem is knowing which bond numbers are still alive, although if that's the problem then I don't know how they know how to prune the dead ones after the numbers are generated anyway.

tfb Silver badge
Boffin

No, such a thing isn't possible. A photon interacting with this thing has a perfectly deterministic state (wavefunction) after the interaction in fact: it's just that that wavefunction is non-zero on both sides of the mirror. When the wavefunction is 'collapsed' (old-fashioned terminology) by the photon interacting with some macroscopic object (so, incrementing a counter or something) then the photon 'decides' which side of the mirror it's on. That decision is 'truly random' in the sense that it's not predictable in advance, even in principle, if QM is correct (see below).

Before it collapses the wavefunction may be skewed towards one side of the mirror or the other, which would mean the thing is biased one way or the other: this can be dealt with by measuring & correcting for the bias as other people have said.

'Truly random' is a bit subtle. We know, thanks to Bell, that if it were predictable in advance, then, if QM is correct, we can bargain that into being able to send information into the past. If we can do that then we can predict it anyway, just by sending the answer back in time. We can also, for instance, win all games of chance, like the stockmarket, and causality generally falls apart. There is strong experimental evidence that QM is correct (look up 'tests of Bell inequalities'), and assuming causality isn't violated then we can assume it is not predictable in advance. But this is subtly different than 'truly random': perhaps 'God' has a big list of all the future states of things and everything is determined in advance, just not from information within the system. This is called 'superdeterminism' and it's pretty easy to see it makes no practical difference: there still is no access to any information which allows us, within the system, to predict the state after the wavefunction collapses.

tfb Silver badge

I think that the prize money is the interest the government pays on the bond (ie on the loan we make tothe government).

The only way the article makes sense is if you think of the premium bond bit of the government as beinga separate organisation which takes in money from us and then lends it, at interest, to other bits of the government.

(This comment reads as if I'm some kind of anti-premium-bond person: I'm not.)

tfb Silver badge
Boffin

Re: Random or is it

That's true. But you can do the statistics on events like that and they are stupidly rare. In particular if a long enough sequence of purportedly-random numbers fails a randomness test, it is enormously more likely in practice that they're failing it because they're not random than not.

tfb Silver badge
Boffin

Re: Removing bias

It's better because it's faster. The spurious 'quantum' thing is just marketing bullshit: both processes rely on QM.

Correction: Last month, we called Zuckerberg a moron. We apologize. In fact, he and Facebook are a fscking disgrace

tfb Silver badge
Alien

Re: Wow

The problem is that these companies share 'second-order' business models -- business models where instead of paying for something their users get it 'for free' and the real business involves some different, second-order process, which in the case of these companies is harvesting data & selling either access to it or algorithms based on it to their real customers (for whom the business is first-order: they pay for what they get).

Second-order business models (more generally any non-first-order model) hve inherent conflicts of interest which drive people to behave in a shitty way. This, I think, has three results: people who are not initially shits are pressured to become so (I think this is what has happened to Google); people who are not shits and won't become them leave (probably all of the big internet companies); and finally people who are already loathsome shits tend to start such businesses & actively drive out the non-shits (Facebook / Zuckerberg).

One interesting thing to look at is other, older, businesses which have second-order models, to see how they have been dealt with, and how successful that has been. There's an obvious example: modern retail banking. For almost everyone banks provide accounts 'for free', providing a very useful service for no cost to the user. They then make their money elsewhere by reusing their retail customers' money &c. And there are terrible conflicts of interest involved, of course, and banks similarly attract shitty people & convert non-shitty people into shitty people. To deal with these problems retail banks are absolutely covered in legal regulations and internally have a mass of processes to enforce these regulations and try and prevent the conflicts of interest and shittiness-conversion from taking over (I've worked in a bank: banks are really frightened about the possibility of losing their licenses). And this works ... most of the time. Mostly, retail banks behave OK as a result, but sometimes they don't and sometimes they behave badly enough to destroy or badly damage the global economy. None of this is helped by regulatory capture, which is why the problems that caused 2008 didn't really get addressed I think.

So retail banking is covered in regulation and this solves the conflict-of-interest problems to an extent which is, optimistically, barely adequate. Facebook &c essentially are entirely unregulated, and actively oppose regulation on the grounds that it will suppress innovation.

There was, probably, a time when the internet businesses were small enough that regulation was possible, but they are now so large and have so effectively captured the people who run the legal frameworks in which they operate that this is not realistically possible: governments are not going to regulate the organisations they rely on for communication & propaganda, still less the ones that feed them the stream of surveillance data they masturbate over, or the stream of backhanders they need to keep themselves in drugs and prostitutes.

This is not going to turn out well: Facebook is a parasite which has avoided any meaningful regulation & will continue to lay its eggs in the body of its host until the parasite load becomes so large that the host dies and the parasite is all their is.

Watchdog asks UK.gov to reissue freedom of information guidance after councils are told to STFU about Brexit plans

tfb Silver badge
Trollface

Re: Why worry?

What advice is that?

Look up, for instance ScenarioMIP.

I'm not going to bother replying to the rest of it: arguing with people who live under bridges is never productive.

tfb Silver badge
Big Brother

Re: Why worry?

If a political act will, or is likely to, have economic effects (beneficial or detrimental) does this mean that the Bank of England should not report on the possible effects, as that would 'not be politically neutral'? What exactly are they allowed to report on then? Perhaps we should just not have them, or perhaps their opinions should always be made in secret to the government, as clearly it would be bad for ordinary people to know what might happen as a consequence of their decisions.

Similarly what about these terrible civil servants we pay so much money who keep giving us advice on the likely impacts of various political choices on the climate and pointing out that some of them are really bad. That's just clearly not politically neutral, is it. We have to stop this dangerous 'science' stuff which is even worse than the wretched economists, or at the very least make sure the public doesn't know about what the boffins are saying.

Yes, this is the knd of thinking we need: don't tell the public anything as that would 'interfere with the planning process'.

tfb Silver badge
Terminator

Re: "New updated guidance has now been issued"

After killing the leopard and opening the cabinet, you find some dirty sheets of paper, some of which have drawings done, in crayon, apparently by a child, and the remainder of which are blank.

tfb Silver badge
Pirate

Re: Nothing to hide, nothing to fear ...

Can we impale their heads on pikes afterwards please?

tfb Silver badge

Re: "New updated guidance has now been issued"

Of course it's not public. I mean, be serious.

tfb Silver badge
Alien

Right, and that's why we should be doubly alarmed: we should be alarmed because we're fucked if there's no deal and we should be alarmed because the government is systematically concealing how fucked we will be from us.

Without expressing a point of view about whether we would be better off in or out of the EU, triggering the countdown to leaving without first having sorted out the plan was an act so stupid it's hard to describe.

tfb Silver badge
Big Brother

So, apparently the government's advice has

[...] balanced the government's commitment to transparency with the need to protect the planning process.

What, exactly, do they need to 'protect the planning process' from? Is there some reason we should not know what no deal implies for us? If knowing what no deal implies would cause alarm then, well, we should probably be alarmed, not pacified by a convenient lack of information. Indeed, merely knowing that knowing that what no deal implies would cause alarm so we are not being allowed to know what it implies should cause alarm. And we do know that now, so we should be alarmed.

It was the best of times, it was the WFIRST of times: How NASA's next exoplanet hunter could find 1,000+ worlds

tfb Silver badge

Re: I hadn't realised

I got it from Charlie Stross. I had assumed it was common usage!

tfb Silver badge
Boffin

I hadn't realised

that WFIRST had been brought back to life after being a victim of the orange shitgibbon. That's good news (although I bet most of the Earth observation stuff he also killed is still dead, which is very bad news).

Nokia 9: HMD Global hauls PureView™ out of brand limbo

tfb Silver badge

Re: Monochrome sensors

Yes of course, but unless you fancy removing it yourself, you need someone to supply such a thing before you can put it in a camera. And unless you're Leica or people flying themin spacecraft, you need them to do so at a reasonable unit cost. This phone proves that this is possible.

tfb Silver badge
Happy

Monochrome sensors

I'm looking at this and I can only see one thing:

[...] three are 1.25um monochrome sensors [...]

I don't care at all about the phone but this means someone is making cheap, plentiful, good-quality monochrome sensors, from which someone else could make a not-absurdly-expensive monochrome digital camera. It might be small-sensor but it will be a dedicated monochrome camera for a tiny fraction of the more than £6,000 (much more if you don't have suitable lenses already) that Leica charge for the privilege of owning one.

This is very good news for a very tiny number of people. Perhaps just me.

Twilight of the sundials: Archaic timepiece dying out and millennials are to blame, reckons boffin

tfb Silver badge

Re: I'll Bite.... Yes, I know he was trolling

Well, anyone who wants to understand gravity...

tfb Silver badge
Terminator

Re: Innovative sundials?

The patent-pending tfb night-o-matic sundial also works in the dark: you merely need to cut away the chunk of the planet which is occluding the Sun. Our luxury leather-plated 'evil genius' edition comes with free giant lasers with which to do the cutting. Sharks and mountain lair extra.

A once-in-a-lifetime Opportunity: NASA bids emotional farewell to its cocky, hardworking RC science car on Mars

tfb Silver badge

Re: Little Green Men (with boots on)

Pretty sure they are its own tracks: it spends some time going backwards (or: the other way, anyway) around then.

(I mean, obviously what it isn't is a boot print: I'm only pretty sure that whatever else it is is wheel tracks.)

tfb Silver badge
Alien

Re: Little Green Men (with boots on)

I'm sure you have a link to this video.

Page:

Biting the hand that feeds IT © 1998–2019