* Posts by Lee D

4261 publicly visible posts • joined 14 Feb 2013

A 5G day may come when the courage of cable and DSL fails ... but it is not this day

Lee D Silver badge

Re: 46.2Mbps fiber?

Fibre = VDSL in the UK because OFCOM are pathetic and let it slide.

VDSL = ADSL + some knobs on. Max is about 75Mbps in general.

VDSL2 gets up to 200-300Mbps.

Lee D Silver badge

Three SIM

Pay Monthly "contract" (i.e. month to month but there are 12/18/24 month options available to make it cheaper).

4G.

30Gb / month, not counting TVPlayer and Netflix traffic.

30Mbps or thereabouts in peak times.

Vodafone do similar - 40Gb, available for £35 a month with a pass to let all kinds of junk that don't count towards your data limit (YouTube, Facebook, WhatsApp, Netflix, etc.).

EE do up to 100's of GB but they get expensive.

If you wanted to do it now, today, then it's a £50 Wifi 4G router (tiny little box), a £22 a month contract or £30 if you don't want to be tied in.

Just an aside - I have had that for at least a year as my only Internet connection because BT etc. generally charge at least £19 a month line rental on top of the broadband prices, so it's cheaper to 4G, and I have no choice for other providers in my area.

5G being even cheaper - more than possible.

Try a quick google search for "Mobile Broadband SIM" before you start throwing around assertions.

Lee D Silver badge

Do you think that every street cabinet has a 5Gb dedicated line back to BT?

I highly doubt it.

Contention is an inherent part of Internet provision anyway. Doing it over wireless media doesn't change that.

The difference is - that street cabinet likely serves several streets (dozens of properties). The 5G likely serves a thousand people using no-data-at-all and a few dozen houses at peak periods. Anywhere where you have more people, you need more bandwidth / coverage / poles anyway.

Azure, Office 365 go super-secure: Multi-factor auth borked in Europe, Asia, USA

Lee D Silver badge

Didn't say that.

But when it does, your data is sitting RIGHT THERE. Available to you.

Literally, pop down to PC World, buy a machine, access your backups, bam... data.

On-prem is no different to cloud or anything else - but you have your data in your possession. So even if it means "ARGH! Quick... install an exchange server quickly so we don't lose email on our domain", you can do that. Without having to wait for some cloud provider to switch you over. Or you can pull all your existing files and refer to them. Or you can literally make your own mini-network and get things running again.

While you're tied exclusively into "I have to log in to Azure and it's not working", you can't do a damn thing, even as an admin.

People with brains do on-prem AND cloud, so failure of one doesn't affect the other.

I speak as someone who only two months ago had 450KW of three-phase cable arced together at the main transformer station supplying the entire site. Our servers went down hard.

And I had a copy of the data... just there... literally in arm's reach. Given an absolute emergency, I could have used iSCSI and ANY MACHINE WHATSOEVER to power up some Hyper-V images direct off the storage and booted things up. As it was, it was a weekend and power came back on the Monday, so I just powered up the 50% of kit that wasn't damaged, checked data integrity and carried on as normal.

But if off-prem doesn't let you even log-in... you're stuffed.

Lee D Silver badge

Another one to add to the bookmarks list I keep for "This is why you really don't want to move off of the on-premises stuff".

It's getting quite crowded in that folder, to be honest - everything from Azure and Office 365 to Google Apps.

Sure, use them. But don't rely on them.

Vision Direct 'fesses up to hack that exposed customer names, payment cards

Lee D Silver badge

Tell me... why are they storing CVV for any purpose at all whatsoever?

https://blog.pcisecuritystandards.org/faq-can-cvc-be-stored-for-card-on-file-or-recurring-transactions

Not least:

"It should also be noted that PCI DSS Requirement 3.2 applies regardless of any permission the entity may have received from their customer to store the sensitive authentication data on their behalf. A customer’s request or approval for an entity to retain the card verification codes/values has no validity for PCI DSS and does not constitute an allowance to store the data."

I hope they lose the ability to take payment cards, because it's not only unnecessary, it's downright stupid.

Washington Post offers invalid cookie consent under EU rules – ICO

Lee D Silver badge

And it's basically useless as even without cookies they can track enough to link all your information together.

As I tell the kids in my school, when they think that clearing browser history or using a incognito window will protect them from my wrath - all it does is keep the records off YOUR computer. Not anything further upstream.

As Chrome itself says right on the Incognito window:

---

"Now you can browse privately, and other people who use this device won’t see your activity. However, downloads and bookmarks will be saved. Learn more

Chrome won’t save the following information:

Your browsing history

Cookies and site data

Information entered in forms

Your activity might still be visible to:

***Websites that you visit***

***Your employer or school***

***Your Internet service provider***"

---

They can tie you into any of your other records without even needing anything more than a vague browser fingerprint, a webpixel image with a particular filename, or any one of myriad identifiers that you're giving out.

Lee D Silver badge

Rule #1: You want to make my life difficult with fake options and deliberate obfuscation? Then I don't use your service.

Lee D Silver badge

1) Wouldn't use a news website that tried to force a subscription on me and/or limited my article views (completely counter-productive if you're then going to shove ads into those views... it's like clamping a car that's parked across your driveway... the person you hurt the most by doing so is yourself).

2) Wouldn't use any international site that, even for a moment, wasn't up on GDPR - most of the US news sites basically just blocked EU access for the first few months, which isn't a solution. They've since caught-up for the most part, which I'm assuming was driven by seeing 50% of their traffic disappear overnight.

3) If they took money from a single EU citizen / EU-registered card to access their site - then they are trading in the EU and need to offer EU-compliant services. Yes, it's complicated in the modern era, but that's how it works. If you are taking EU money, you need to abide by EU law and - also - pay EU tax.

Microsoft sysadmin hired for fake NetWare skills keeps job despite twitchy trigger finger

Lee D Silver badge

I was pushed into an interview by a well-known job agency whose name is something you might get growing on the side of a river (they've previously bought articles on here, and my comments then were censored because they are FAMOUS for being useless, especially for IT and medical staff).

What I thought was an IT position turned out to be:

- Phoning around Vauxhall car dealerships, asking if they needed anything from the main base.

Now... don't get me wrong, it wasn't Vauxhall's fault. The job was quite clearly specced but the agency had obviously struggled to fill the role and so they were deliberately obfuscating when it came to ringing up other people on their database (I was newly job-seeking at the time) and trying to shoehorn them into an interview "because it's last-minute" and they had nobody suitable.

So I ended up falling for their ruse (should've known *) and wound up in an interview where I'd been completely misled about the whole industry, let alone category of job I was interviewing for.

This all came out in the interview (obviously) with more than a glancing nod of "fecking job agency" between the people on the panel when the full story came out.

Long story short - they offered me the job. They said I was the best candidate by miles, presumably because I was honest about it and because I didn't let it worry me and still gave a good interview based on what I can provably do (i.e. pulling in customer service experience from my actual, real career). Even the guy from the agency who phoned me up sounded surprised.

P.S. I know feck-all about cars.

* My history with the company is thus: They did that to me. They did something similar to my brother (who sat through series of interviews before being informed that it was a FIELD technician role, and he doesn't drive and nobody had bothered to ask/tell him that - so he aced the IT part and then got blank stares when asked for his driving licence). My ex worked in a genetics lab and went to their healthcare agency, who do nothing more all day long than put people into healthcare jobs. She was asked "What's genetics?" and "Can I just put that down as nursing?"... so they wanted to shove a qualified Dr of genetics into a bedpan role...

Court doc typo 'reveals' Julian Assange may have been charged in US

Lee D Silver badge

Re: Well he's a bit of an arse, but...

Er...

Yes,.

And the problem with that is...?

The *RUBBISH* was that he would be forcibly extradited against the UK's will, or attacked or shot or killed or something.

But charges were always possible and pretty much inevitable.

To apply them to him, they have to apply for extradition. To the UK. Who approve or not based on law. Like everyone else.

What you've done is taken his hyperbole and spin and applied it to "US charges someone who revealed US classified information, and then MAYBE asks for their extradition to face US charges". Which is totally non-shocking, non-news, and non-exciting.

Assange, however, was claiming that he'll be "extraordinarily renditioned" or murdered. Which has been utter tripe. And that's obvious because the bloke is stuck in a VERY VERY public place that he cannot ever leave and gives interviews out of his balcony... hardly the actions of someone who fears being sniped.

If at first or second you don't succeed, you may be Microsoft: Hold off installing re-released Windows Oct Update

Lee D Silver badge

Nothing wrong with 8.1.

Slap Classic Shell on it, turn off all the Metro junk, it's pretty much Windows 7 with knobs on and the only real difference is when things want to jump into Metro for no good reason (e.g. some settings options).

Literally, this place was a Windows 7 shop, that had tried to upgrade to 8 "naively" (i.e. not tried to customise it at all) and it was a disaster and reverted and "never again". I took over the IT, silently started rolling out 8.1 with Classic Shell and NOBODY noticed or cared and still haven't 4 years later. The only thing they noticed was the organisation logo in the start menu button, because I put it there as a bit of branding, and they all loved it. That's something that, as far as I know, you can't do with basic Windows alone with all kinds of messy editing.

It's about how you manage it, not the fact that it's 7 / 8.1 / 10. The problem is that I consider 10 too big a leap for my users at the moment so I will lump it together with a load of other big-leaps so they just have to bite the bullet and change the way the work once, rather than 3-4 times over for different things. But that will happen when ALL my suppliers support 10 properly (and have done for a whlle, not just said "we now work on 10"), and the banks etc. are dragging their feet on that.

Lee D Silver badge

Yeah, this is why I stuck on Windows 8.1 (which gives me a few years before upgrade yet!) and don't automatically install any updates.

Sorry, Microsoft, but if even 1% of computers fail your update with errors like "you can't map network drives", then that update isn't going to get applied. Simple as that. I'm not running around 1% of my machines diagnosing mystery issues that won't get fixed until next year.

Have you not learned the lesson yet?

Monolithic updates are bad.

Updates which affect thousand upon thousands of unrelated files are bad.

Untested updates are bad.

Forced updates are bad.

If you'd broken this out into lots of small patches, you could have a) not affected your bandwidth but let users download it piecemeal, b) isolate individual fixes into individual updates with almost no difference on the back end, c) allowed power users and your own testers (laugh) to isolate which parts of the updates are failing easily, d) let your users download all the ones that aren't broken and revoke just those that are, e) seen interactions such as update X only goes wrong when also applied with update Y and saved everyone a lot of time, f) let users isolate using binary-search and remove individual updates that happen to cause them problems even if you don't care a jot about doing that yourself.

I really hope you continue to work like this, break so many critical updates, and get one of those "Whoops, you updated now and it's broke and there's nothing we can do" scenarios that you can just step back from, and start breaking people's machine irreversibly, so that you're made to stop with the forced updates. If you could manage that before 8.1 comes out of extended support, I'd be so happy.

Bright spark dev irons out light interference

Lee D Silver badge

My dad once ground down the edge of a SIP memory chip (usually only used in printers, but in this case it was for a 386 motherboard that only took SIP memory) with a power-tool to make it fit.

It over-hung the SIP socket and touched a component, so it wouldn't sit in there properly. He literally just took it to the garage and ground it down to the edge of the PCB traces.

Not a big deal, you might think. But we'd paid nearly a grand for that memory upgrade (1Mb to 2Mb!). My brother and I just watched in disbelief as he ground a grand's worth of state-of-the-art chippery to within a slither of the traces with a power-tool...

Worked though. We sold the machine fully-working many years later still with that chip in it. Not sure you'd get away with it nowadays with the multi-layer boards and extreme sensitivity of the chips.

Nvidia just can't grab a break. Revenues up, profit nearly doubles... and stock down 20%

Lee D Silver badge

Re: Oh please

You mean those people running mechanical, engineering and physics simulations who can't afford to have errors creep in from the implementation of something that might be there "because it makes this game run faster"?

Yeah, I think they care immensely. Especially when they're doing months or years of calculations and need a stable base to run it on, a predictable and consistent interface to do so, to squeeze every inch out of their hardware, and to do so without unrelated gaming/media functions or driver bugs rearing their head.

Brexit: UK will be disconnected from EU databases after 2020

Lee D Silver badge

Lessons to be learned:

- Never ask a question you don't want to hear the answer to.

- Never promise something that relies on other people's / country's / continent's co-operation.

- Never assume it's simple to undo decades of political wranglings and legal paperwork.

- Never take on a job that you don't actually want to do.

- Never stay in a job any longer than necessary once it becomes clear it's not what you want to do.

- Never piss off your allies without good cause.

Lee D Silver badge

Re: Is it just me or...

It's closer to that "time is slowing" sensation you get when you're the one hurtling towards the wall yourself, though.

iPhone XS: Just another £300 for a better cam- Wait, come back!

Lee D Silver badge

Re: Famously???

If you don't want people to call it an XS, don't use those symbols. You can't mix and match Roman numerals with ordinary letters if they're touching like that, or even if they are next to other things that are designed to be pronounced as letters.

And especially never name it such that the Roman numerals and the letters form a word (Excess! Most fitting) when you do pronounce them differently.

Mac OS X. Show me anyone who calls that "OS ten".

You'd think a company that's "all about the design" (cough, choke, splutter*) would know that.

(*Like their famous book on design... which has a white spine... with the name in white... on a white background... with no outline or anything else, just some very subtle "embossed" letters - so in almost all lights, it just looks like an blank white book spine... and is just a photobook that obviously costs something ludicrous like £3 per photo / page inside it).

Lee D Silver badge

£999 is enough to pay for my entire year's broadband (over a 4G wifi box), phone and tv, Netflix, Amazon Prime, a trip to the cinema once a month, and still have enough left over to buy the 4G box, a PC and a TV to do that all on. Not to mention, buy a mobile phone. Hell, if I skimped in the right places, it would also pay most of my household electricity usage too (not just that of the above).

And that's BEFORE you even connect the damn thing to a cellular provider.

Honestly, I've paid less than a quarter of that for CARS before now, that worked just fine for years afterwards, and then scrapped them when they needed HALF of that cost in works to pass the MOT for something else.

How the hell do people justify iPhones to themselves?

Douglas Adams was right, ish... Super-Earth world clocked orbiting 'nearby' Barnard's Star

Lee D Silver badge

Re: Getting a proble there?

The other objects are all in orbit around somewhere. It's cheating to consider orbit a speed, because then you get ridiculous values when planets are moving away from each other, etc.

The Pioneer's might qualify but seeing as we haven't contacted either in over a decade, it's pushing it. They were last doing 0.000041c and 0.000037c (same number of zeroes as Voyager).

Let me clarify: Consistent speed relative to Earth. That knocks all of the top 10 out.

Lee D Silver badge

Re: Getting a proble there?

Indeed.

The fastest man-made moving object is currently Voyager, and it's doing 0.00005c (3.6 AU/year).

It would need over 17,000 years to travel a light-year at that speed.

You just have to hope that space is incredibly empty.

Lee D Silver badge

Re: As for ion drive, where do you propose to get the energy from?

If you can use a planet's gravity to slingshot (i.e. accelerate) a probe, why couldn't you use one to slow it down too?

All you do is fly AGAINST the motion of the planet, rather than with it.

Sure, the maths is damn hard, and you'll want to make sure you have a good view of the planet / system in question long before you need to do the manoeuvre but there's no reason you can't use the same trick we use to launch probe quicker and further than ever before (Voyager) to slow them down at the other end in the same way.

Literally... fire it towards the planet "in front" of it, by millions of miles to be safe... aim it at the planet, as the planet flies past, it drags it back a little, slowing it. Take some photos while you're there. Readjust the orbit as you leave (because you'll now be in an elliptical that includes the planet's position, so you'll go out of the system, and then come back in towards it for another go) and keep doing that until you're in a perfectly circular orbit.

It's the kind of celestial mechanics that you can leave a computer to do nowadays, just updating it with more accurate observations as you get closer, and it'll work out an orbit with however many dozen deceleration slingshots you want until you get to a nice stable orbit. Minimal thrust to adjust the orbit would be required, and presumably you'd be in range of a whole new sun or still be powered nuclearly by that point.

You never go TO the planet, you go past it, let it swish you back around, have another go, and another and another and another. It's perfectly viable and we did it around Mars for every launch - nobody's yet invented vacuum-brakes or a way to get geostationary (or even stable) orbit perfect first time from millions of miles away in a straight line.

A new Raspberry Pi takes a bow with all of the speed but less of the RAM

Lee D Silver badge

Re: Naming Standards clearly designed to confuse...

Yeah, either name consecutively (1, 2, 3, 4) or have a clear distinguisher (RPi 3+ has more on it but from the same series as the 3 - why do you need A or B AND a plus?).

Otherwise it's gets too tricky.

Basically:

Zero = very tiny one.

Pi 1 = single-core 700MHz, 256Mb/512Mb, composite video (later ones had HDMI - grr!)

Pi 2 = quad-core 900MHz, 1Gb, HDMI

Pi 3 = quad-core 1.2+GHz, 1Gb, HDMI

The + models have more USB / Gigabit Ethernet / faster speeds.

But they really mess it up by randomly changing things between models, including three models of zero, the compute modules, putting pins only on certain models (e.g. the PoE pins?), and changing the form factor and even layout between each one.

It would be much easier to standardise the modules, and just not solder the extra components onto the smaller models, just like every other manufacturer does.

Basically, stick with the highest numbers, largest-lettered model available (RPi 3B+ at the moment) unless you have a particular pressing need to optimise power / space.

Lee D Silver badge

I'm beginning to like the RPi's again.

I was an early adopter / tester for the original and their hardware was quite bad, and their educational focus was all-front-no-action (still is).

But the RPi is now a cheap, viable, fast, powerful, multicore, PoE-powered, small, ARM Linux machine with Ethernet, Wireless, Bluetooth and HDMI.

I now look at the things my workplace use on the network and the Pi is capable of replacing half of them - PoE powered tannoy, PoE powered phones, PoE powered CCTV, vehicle GPS trackers, hell, you can make a Pi-powered smartphone if you like.

I have one with RetroPie, also running Kodi, using the DVB-T hat (that was just released), plugged into an aerial, connected to a projector, offering TV on-screen and out over my network (using tvHeadend), PVR functionality, and just one button away are all my old games and a ton of open-source ports. It uses an XBox 360 USB dongle to connect a bunch of gaming pads, plus I've wired in a couple of "arcade style" buttons/joysticks, a Logitech G27 wheel-set and a Bluetooth keyboard (because you can't play Speccy games with only a joystick!).

If I ever fulfill my dream of buying an arcade cocktail cabinet, then that machine is more than capable of running it and controlling all those inputs, for all the games I like, and being the "TV" by offering HDMI out.

Oi, Elon: You Musk sort out your Autopilot! Tesla loyalists tell of code crashes, near-misses

Lee D Silver badge

Re: No way ready!

The difference is quite obvious.

I know that the shadow is cast by a person. The computer doesn't. It has to be told. There's no driving instruction / test where they tell you "watch out for shadows", specifically. A human, though, is able to look at the situation (blue lights / shadow / kid running from the fields several feet off the road / the sight of a ball crossing the road in front of you which is likely to be followed by a small child retrieving it, etc.) and infer things about it that aren't present in the raw data.

The computer *cannot* do this. They can't learn like that. They can't infer anything that's not absolutely 100% inherent in the data or programmed in. They can't slow down every time you pass a police car with blue lights, nor can they track every object to the point that it realises it's a ball and cast the trajectory back to its likely origin without also slamming the brakes on for a paper bag blowing across a motorway.

The machines do *not* infer data. They are incapable of doing so. All of them. Even the "AI" ones. They don't infer. They are told, they try to find a marker within that data which is semi-reliable, and then then guess. They have no idea WHY they have to brake, they don't know why the road is suddenly all shiny and rainbow-coloured and why that means you should probably slow and make no sudden lane changes or steering at that point. They can't infer it back. They can only react to specific data they've been told to look for.

And you CANNOT tell a computer to look for every possible circumstance, with any accuracy. It's just infeasible.

I put my life in a computer's hands every time I get into a car. ABS. An ECU controlling fuel pumps. Even electronic engine timing can blow up a car if it goes wrong. But they are NEVER required to guess. If the oxygen is below this reading, signal failure. If it's between this and that, then you're at this point in the stroke and you should do X, Y, Z. They never "guess". They can't "infer". They don't know why the oxygen sensor suddenly returns zeroes, they just get told what to do if it ever does. This is why most cars with oxygen sensors just stall if the sensor is faulty. They can't infer that it's faulty and ignore it. They just sit, splutter and stall. Disconnect it, and the engine KNOWS it's not there and slips into "limp home" mode. But you have to know to do that.

And that's the entire problem... current tech can't even stop files being deleted, people breaking into websites, or properly autocomplete an English sentence. And you expect it to be even vaguely safe to interpret what is possible the worst scenario ever for a computer vision system?

Never rely on a computer to infer. They can't. They don't understand the world and thus cannot predict it or even notice when they themselves are failing. You give a computer instructions to do far more rapidly and perfectly than you could ever do. That's what they are for, that's what they are best at, that's what they do. You do not get a computer to ever infer anything, certainly not in any life-threatening scenario. These things can't even write a decent paragraph of English text with years of supercomputing efforts behind them. They have no concept of the data they are acting on. They are just following instructions.

Those sufficiently complicated instructions can work wonders, yes, but they cannot generate any sort of intelligence (nobody has ever proven that and, no, a Turing Test is nothing more than a psychology test for a human, not an intelligence test for a computer), and they cannot infer anything that's not present in the data.

If you can't infer, you can't understand the situation, or adapt properly to it, or deal with any situation which you don't have explicit instructions on how to deal with.

Uber is a great example - that self-driving car that killed the woman with the bike? Within a few seconds it detected her as nothing, a wall, a bike, another vehicle, a pedestrian, and then didn't know what to do about any of them. A human would infer from all those instructions what the situation actually was. The car wasn't trained on it at all, and wouldn't have stopped in time even if the braking hadn't been disabled.

Inference is an inately human / animal skill associated with intelligence. If I bop myself on the head with this stick it hurts me. So that means if I bop THAT monkey on the head...

Inference is a vastly different skill and not present in any computer system that I'm aware of. Not one of them tries to trace back the reasoning for the data being classified as such. They just operate on statistics and heuristics. Don't trust your life, quite literally, to chance and what-some-bloke-wrote-down for every situation.

Lee D Silver badge

Basically, driving via Twitch,,

"Left! No right! Run him over! No don't! Mount the pavement! Spin the car in the middle of the highway! BRAKE! Burn the tyres off!"

Lee D Silver badge

Re: No way ready!

Computers are singularly unable to infer.

We know why there's a shadow on the ground coming from between those parked cars in that lit street... we know that's there's a school on the right and the kids are playing football against the very fence that we're driving past... we know that the pillock in front is driving contrary to all road laws and that means we have to be much more careful interpreting the situation around him... we know that the thing that moved out from between the parked cars might just be a child, or it might be a paper bag that would look solid but not actually be so... we can infer that the blue-lights around the corner ahead mean we should approach with more caution than normal... we know that though we *could* make that gap, it's probably not safe or sensible to do so because it relies on everyone else continuing to drive exactly as they are, yet they are human too.

Computers cannot infer. They react only. Have you ever been in a car that is driven by someone purely on reaction? We're not talking rally drivers (they are scary enough, but because they are reading the situation and taking calculated risks that we wouldn't), but people who literally have no sense of the road and just drive based on what's six-inches off their bonnet? It's terrifying, no matter how skilled a driver you think you are personally, the one thing that will make me get out your car is that you're only *reacting* to what happens to you, not *predicting* or *infering* or *adjusting* to the scenario.

That's what the cars are doing. Reacting. They have no way to infer even the basic properties of the objects around them (e.g. the paper bag scenario... there is no sensor on a Tesla that can distinguish between a paper bag and a rock... so does it swerve to avoid the paper bag, or does it drive straight into the rock?). They certainly can't infer anything about the wider situation. That makes them dangerous. More so in that, in ideal conditions, they operate fine and gain the user's confidence. Nobody cares about that. Ideal-condition driving is both pleasant and fun, personally. That's the bit I don't *want* the car doing. But it's not capable of the bit I don't want to have to do at all... adjust to a rapidly-changing scenario or infer the intentions of other drivers, or even infer the properties of the road ahead.

Lee D Silver badge

Re: Say what you like about Teslas

Chart 6 / 7: Casualty / fatality rate per billion passenger miles by road user type: GB, 2017

<-- mathematician.

There's no point in comparing cyclists who cycle 200 yards to motorbikes who drive 200 miles.

P.S. That's GB only. Worldwide, it's INCREDIBLY higher.

Lee D Silver badge

Re: Say what you like about Teslas

Though I utterly agree, I do wonder why you continue to put your life at risk.

I used to be a cyclist. I stopped. There's no way I can justify being a tiny little unprotected, invisible thing sharing a road with 18-ton lorries, whether they look or not. People in cars get wiped out by such things every day, but at least they have a ton of steel around them designed to do nothing other than save their lives. On a bike, you don't even have that. Even putting a huge motor between my legs wouldn't improve that situation noticeably at all.

I'm not against cycling in any way - I actually think cyclists should be allowed to cycle on pavements if it's safe to do so. Motorbikes I think are fecking suicidal. Not only can you get hit, but you can go fast enough that it's instantaneous death, do not pass Go, do not collect £200.

But I cannot, for my own sake, justify being on a road in the modern era on anything approaching a two-wheeled vehicle. We laughed the Sinclair C5 off the road for safety reasons, why haven't we done the same for bikes?

Some 75% of road deaths are people on bikes. Literally, your life expectancy is lowered substantially being a regular user of one.

No matter what "should" be the case, why would any sensible person continue to partake in such a venture? I "should" be able to walk through the streets of London at 3am, counting ten thousand pounds out into my hand. I don't because that's just such an incredibly stupid thing to do, even if I simultaneously campaign for change in that regard.

Honestly... how do you justify, after several near misses / actual hits, continuing to use a bike? "Out of principle?"

I know I gave it up and wouldn't go back, and it's entirely unrelated to how well *I* can ride one / drive when I'm around one.

Lee D Silver badge

Deathtrap v0.98

Sorry, but we just don't have software capable of this kind of thing reliably, and despite Tesla using every trick in the book (including "hard-coding" instructions to do things like "ignore that bridge, because it's not actually a bridge" in certain geographical locations that they know it has a hard time interpreting) it can't ever be safe like that.

These are the diligent few, the ground-breakers, the ones willing to put up with flaws in the system to say they have a new. When it gets into the budget/hands of the general public it's going to come to light quite how dangerous these things have *always* been.

This is why the software claims were looked on dubiously from day one. We just don't have systems capable of doing the things they are claiming, and likely won't until some actual proper AI revolution (not the current "AI" fad which is just statistics and brute-force until the software plateaus doing a half-job and takes longer to un-train it from bad behaviour than anything else, and which we have no idea how it's actually inferring/making decisions, yet we pretend we can still make it do what we want).

'My entire company is without comms': Gamma's Horizon cloud PBX goes DOWN

Lee D Silver badge

We're on Gamma, and we're entirely SIP.

Not seen any problems yet, however.

P.S. This is why you ALWAYS keep one analogue line...

FPGAs? Sure, them too. Liqid pours chips over composable computing systems

Lee D Silver badge

Quite.

I'd like to tinker with these but it looks like a few hundred before you get anything worthwhile, and that has to interface with a bunch of stuff to do anything useful or you have to spend an inordinate amount of time adjusting.

For me, there are a few things in that kind of range: FGPA, SDR and CUDA.

I'd be after a "microbit" of FGPA, if that kind of thing exists. A teeny, tiny version that runs off a USB stick and which you can stick GPIOs etc. on.

https://tinyfpga.com/ looks promising but it'll have to wait for my Christmas funds as likely it won't just be as simple as clicking Buy Now, getting it, plugging it in, and starting to code it up.

Likely by the time I tinker and get there, things like FGPAs will be an inherent part of every processor / motherboard in some fashion anyway. A few people already make RPi boards for them.

It's November 2018, and Microsoft's super-secure Edge browser can be pwned eight different ways by a web page

Lee D Silver badge

Re: Never mind the flaw, look at the *pattern*.

If you violate the EULA, you lose the right to utilise the copy-right for the software.

Thus anything past that point is illegal usage of the software, because you don't have a licence to use it (and you can argue until the cows come home, but usage of software is legally classed under "making a copy of").

No different to how the GPL is enforced - if the software is only licensed under the GPL, and you violate the GPL, you violate the only agreement that gives you any rights whatsoever to the software. Hence, you have no rights to the software. Hence you've broken the law.

The only people who can change that are the ones who licensed it to you in the first place - by choosing to offer you another licence, another chance, overlooking your violations, etc. or otherwise giving you explicit permission to continue using the software.

Software use is a right (literally, a "copy-right") given to you by the creators of the software that describe how you may use it. If you fail to abide by their rules, or don't agree with them - then you have as many rights to the software as anyone else does... zero. The "purchase" and acceptance of an EULA is the only thing that gives you the right to use/copy the software in the first place.

There are complications (e.g. unfair contract terms, having to execute code in order to accept the agreement, etc.) which make what a lawyer would call "interesting questions" (i.e. gimme a few grand and I'll think about how I'd argue it in court, no guarantee of success). But pretty much you abide by the EULA or you literally have no right to the software whatsoever.

Don't believe the hype that EULAs are unenforceable either. It's not that simple. It's like saying that a tiny flaw in one particular huge contract makes the whole concept of contracts unenforcable. It's not true. And pretty much there's a clause that says "If one thing in this contract isn't allowed, all the rest still apply anyway" (and, amazingly, it doesn't even NEED to say that... that's very much a "your statutory rights are not affected" statement... of course they're not... they're STATUTORY rights...).

Data-nicking UK car repairman jailed six months instead of copping a fine

Lee D Silver badge

Re: Far too common

Next time give them a PAYG phone number and a forwarded free email address that you have never given anyone else.

Cost: Pence.

Impact: Nice letter winging it's way to them and the Data Protection ombudsman saying you've totally abused the data I've given you without permission and/or failed to secure it and not notified me of a breach... which is it?

P.S. Have done exactly this. Have screwed company to wall who used a stolen customer database. I know they did because (in my best Del Boy impression "That's not the hand I dealt you...") this case... That's EXACTLY the email address that I only dealt SOMEONE ELSE and not you...

Lee D Silver badge

Good.

The more cases like this I can point at, the less chance I have of any resistance to my "least privilege principle" processes.

Question: Why does the software allow blanket access to names and addresses of customers that he's not even dealing with?

I actually would posit that almost all call-centre software should be illegal under GDPR because you have no need to actually KNOW what the customer's address / phone number actually are. You just get put through to them by the system, and unless they ask you to change or confirm the address, you have no need to do so much as request it (via, say, a "Request" box on each database field), and so any blanket-requesting of customers would flag up under auditing rules, and any attempt to "mass export" the customer list would just fail and set off the flashing red lights.

Remember: If it's not REQUIRED for your job, you shouldn't have that access to that data. 99.9% of the times I've called up any utility companies, taken a call from suppliers, etc. there is literally zero need for them to personally have access to any of those details.

"Shall I ship it to you home address sir?"

A) "Yes please". Done. No need to do anything but "deliver".

B) "Hold on, I moved recently, which address do you have?". Call operative presses Request on the address, the grey box for address only gets filled out from the database, operative reads it out, confirms it. Done.

We honestly need to start designing systems around least-privilege (again) before the law catches on that it's own definitions require it.

We definitely don't need more towers, says new Vodafone boss scraping around for €8bn savings

Lee D Silver badge

Re: no more Towers?

5G doesn't need them.

In fact, the first 5G rollouts won't even be able to. The licence auction hasn't happened yet.

The high-freq stuff is merely a nice extra - if you're in range, you'll get "even faster" but existing towers and even some existing frequencies will always be used (and not just by "fallback to 4G").

5G doesn't "need" anything beyond what's already there. It just makes it slightly faster and have clearer channels.

It's like saying you "need" 5GHz for 802.11n. No. You don't. It'll still be faster on 2.4GHz than b and g ever were. But if you're able to use that higher frequency too, you'll go even faster.

NBN satellite user waiting for extra gigabytes? Keep waiting

Lee D Silver badge

Ya cannae beat the laws of physics, Jim.

And I've run an entire school for two weeks off 4G sticks before now and nobody was any the wiser.

Between you, me and that dodgy-looking USB: A little bit of paranoia never hurt anyone

Lee D Silver badge

Easy it's called "I can implement that change, but it'll cost you one IT Manager and a lawsuit about trying to get rid of them for providing adequate data and system security with a reasonable, demonstrably-effective, proven and already-in-place system".

Also, that in any proper workplace, such people DO NOT have access to the IT system whatsoever (physically or electronically), in any way, to implement such a change behind your back - even if they got IBM/Microsoft themselves to come in and try to do it.

Hint: The triggering of any one tripwire which suggests intrusion - whether by my own employees (IT department), other employees, outside entities, management, or any of their consultants - will result in the correct response in the case of such potential compromise. A full system shutdown until the situation can be determined.

Other hint: Every workplace I work at is made aware of a simple rule. If I ever discover that the master password lists / backup devices are accessed by anyone other than authorised personnel in the reasonable execution of their jobs (and I will know), I walk.

You really need to read GDPR. Unless your boss has a reason to have the domain administrator's password/access (hint: They don't, unless that boss is the domain administrator), then it's illegal for them to have it. They can *request* it. They can *instruct* me to hand it over. And I guarantee that it'll cost them one IT Manager and a lawsuit unless it was absolutely required (e.g. I'm in a coma in hospital somewhere).

P.S. The best way to stop such things is to say "Sure, I'll do that. But it's against my advice. Just sign here to tell me that you understand that and accept the consequences". I've actually used that. It's incredibly effective. No, my boss does not have any IT rights beyond that of any other member of staff working in such a position (e.g. he has a PC with office, rights to the documents he requires, but can't even rebuild his own machine or log into a server).

Lee D Silver badge

First action upon starting at my current workplace:

A blanket ban on all USB sticks and any mass storage devices, and any "unauthorised" USB devices in general.

You want that, it has to come through IT who will scan it, and copy it to normal storage for you. If it leaves site at any point, it has to be scanned again. No exceptions. Not even for the big bosses. USB is just disabled and alerts us when it's attempted.

That's held for 4 years, and I'm regularly able to demonstrate why it's in place (with speakers, presentations and visitors all the time, there's ALWAYS something on a stick, and more often than not I have to refuse them access).

Number of virus infections: 1. Contained to a single PC. Introduced from a dodgy download, which the user persisted in attempting to run despite it being a file-inside-a-file-inside-a-file from a personal webmail from a spam from someone they didn't know, etc. etc. etc.

(Second action on starting at my current workplace: Stop all the password expiry nonsense as per all modern password guidelines.)

Honestly, you have ZERO NEED to use USB sticks, or even devices. The hindrance is literally "Hi John, nice to meet you, can I just take that stick from you to give to IT, they'll put it on the system for you and give it back, cool, let's go get a coffee and get you set up, eh?". You're just introducing the potential for everything from keyboard loggers, wireless access that bypasses your network security (or even shares out the local network to the Internet!), etc. etc. to anyone.

You need a piece of software that lets you block categories of USB drivers (e.g. mass-storage, etc.) and also whitelist authorised devices. Even then, there's potential for serious compromise (e.g. nothing to stop a USB keylogger looking like an authorised keyboard by offering fake USB PIDs).

Huawei Mate 20 Pro: If you can stomach the nagware and price, it may be Droid of the Year

Lee D Silver badge

Re: Can anyone tell me the advantage of face/print unlock?

Having something in your brain (currently) leaves no physical impression or trace that can be detected or copied.

Having something on your body doesn't. It may be "complex" to copy, but it's there. And can be fooled by things like Gummi Bears or just a high-quality print-out.

Nobody has yet managed to pluck a thought from someone's head (though Derren Brown can show you quite a few tricks), so that's the ONLY way to be secure at the moment.

Anything else is security snakeoil. In the same way that your briefcase doesn't need to withstand a bunch of safe-crackers (given 5 minutes, a fast hand, and the opportunity and I'll open any 6-digit briefcase combination lock for you), some people are happy with "no security but a slight inconvenience".

If you care about security, a PIN / passphrase is the only way to go.

Lee D Silver badge

SD Card and battery, personally.

The headphone not being there is annoying, but the other two you KNOW are going to fail or fill up and you'll want to replace at some point.

Western Digital: And when I pull the covers off, behold as NAND becomes virtual DRAM

Lee D Silver badge

Ah, the old "let's run a swap-file on a RAMDisk" trick of old.

Honestly the two technologies really need to just merge so that you're "reserving" a space for the swap-file is really just "use the first 16Gb of the storage for main RAM" on a device where you can treat everything equally.

We need to just replace "RAM" and "disk" with "memory" and use it for everything.

Wanna run a server? Run several chips of memory in a RAID and thus have them verify each other (ECC, effectively).

Upgrade the disk, you upgrade the RAM. No suspend/hibernate junk. No "I ran out of RAM so I'm just going to fall over and crash". And "persistent RAM".

We need to get rid of hard disks, which will make everyone focus on SSDs and Flash, which means we could easily get their longevity / sacrificial sectors to the point where they compete with RAM, and they are already fast enough that you need a direct-bus connection to get the most out of them - and creeping their way towards RAM speeds all the time (I believe they are in the DDR2 ranges for speed already).

Brit boffins build 'quantum compass'... say goodbye to those old GPS gizmos, possibly

Lee D Silver badge

Anyone else reminded of The Big Bang Theory episodes?

4G slowcoach Three plans network and IT overhaul to get foot in the door with 5G

Lee D Silver badge

Re: But... what will it cost?

https://mobiletechtalk.co.uk/4gee-home-router-review/

I pay-per-month. So that's a £100 up-front cost, for a device I'll literally never use, then £50+ a month, for lower speeds, with dodgy traffic and being backed by BT (who are the very reason I can't get decent local broadband in the first place).

For 100Gb on that, I'd have to pay £100 + 12 x £45 = £640 a year, which is £53.33 a month, or sign up for 18-months (which I can do with Three but deliberately avoid doing even though I'd save).

It's an option, sure, but it's not one I'd choose, and it's still just-as-cheap to buy two SIMs from Three instead, and kit of my choice, and use them in tandem. Or just buy a smaller data SIM from each and then use EE when Three falls over, etc.

Lee D Silver badge

Re: But... what will it cost?

I do this.

I have no landline (telephone or broadband), cable, TV, etc.

I live in London. It's cheaper to buy a 4G box (I use a mini Huawei 4G Wifi router that runs off a mobile phone battery, is based on Android and is basically a mobile hotspot with 8 hours of battery life, so I use the same connection when out-and-about, on holiday and when at home for my broadband) than it is to get even an ADSL line installed.

I get more than adequate speed to do all my TV through it (TVPlayer/Netflix/Amazon Prime/iPlayer, etc.). I get more than adequate speed to do all my browsing through it. It get more than adequate latency to play online games through it.

Literally, the only blocker for me is their pathetically low data plans. The best you can get in the UK if you're tethering (they cut you off for such things, if you think you just can use your "unlimited" mobile phone data package) is about 50Gb a month. I'm on 40Gb and they don't count Netflix (which is another 30Gb for me). It's more than adequate for me, and I'm online all day long. I have my CCTV on it, I have my phone on it (phone uses it over Wifi, which may seem odd but then I can even out my data usage), and my entire local network runs off it (including CCTV, printers, games consoles, RPi, DVB-T streamer, Chromecast, etc.).

I do *not* get an amazing signal (I bought a little aerial to plug it into when I'm at home), but it's already viable on 4G and held back ONLY by the stupidly low data limit which as far as I'm concerned is entirely artificial. 40Gb is fine for me but if I had more people in the house I would have to bump it. And weirdly it would be cheaper to buy several 4G routers and SIMs on Three than to pay their data over-charge (my Huawei has an option to piggyback off another Wifi in preference, so you just make a chain of them and turn one off when the data runs out...). I set the SIM to not allow overuse so if I over-use it, it just stops until the next month rather than charges me.

Also works in Europe when I go over there, so I'm not using up anyone else's data.

I have it on Pay Monthly but on an annual contract you can get it down to £22 a month. Cheaper than literally EVERY broadband offering available to me, once you include line rental. Also I get no telephone spam, because the only phones I have are mobiles and SIP.

I would actually pay £50 a month for Three to let me use, say, 100-150Gb of data on that same 4G SIM. I wouldn't care if it was 4G or 5G, to be honest. I get more than enough speed to cope with everything I throw at it (including 1000 Steam games and their constant updates).

The only alternative is Vodafone (who charge £30 a month for 50Gb, and then another £15 "pass" will remove all your Facebook, Whatsapp, Netflix, Prime Video, etc. data from counting towards your allowance). But they're too stupid to send me a SIM or realise that I can't get back in to order another with the same details because the first never arrived, and I can't be arsed to argue in a shop with a digital telecoms / Internet company who can't work how to get me to order a SIM online).

All the other I look at have LUDICROUS limits when tethered / mobile broadband. The average is about 9Gb a month. But a Three Mobile Broadband Sim-Only 40Gb w/ Netflix traffic built in can be had for £22 a month, and I get 30Mbps and more out of it all the time.

Lee D Silver badge

Re: G5's higher RF frequencies

Stop spreading FUD.

The mmwave stuff is an optional extra, not a strict requirement.

"Initial 5G launches in the sub-6 GHz band will not diverge architecturally from existing LTE 4G infrastructure. Leading network equipment suppliers are Nokia,[23] Huawei,[27] and Ericsson.[28]"

If you're close enough to get mmwave, you'll get greater speed. But 5G will also do you better than 3G/4G out in the field from the same transmitter that did those for you, on the same frequencies as those use.

Spammer scum hack 100,000 home routers via UPnP vulns to craft email-flinging botnet

Lee D Silver badge

Friend bought round XBox 360. It worked. No UPnP.

I have 1000 games on my Steam account. They all work. Online.

Skype, Whatsapp, hundreds of apps, phones, other people's consoles on games nights, you name it. They all work.

You only EVER *NEED* a port-forward if you are HOSTING content. You do not need it to game, join servers, browse servers or anything else. All major consoles have matchmaking services that can handle that side for you, no port-forwards required. And that's because only when you are actually being a server should you be punching holes in your firewall to let others in (rather than talking to a matchmaking server, or talking over an ESTABLISHED connection to another person which is what matchmaking servers set up for you).

Seriously. Turn UPnP off now. Play any game you like. See what happens. At absolute worst, XBox even has a term for it that shows up in the settings that nobody ever looks at... it basically means "you're behind a NAT, so I'll use a matchmaking service that knows that".

UPnP has several functions - one discovers things over the local network using local broadcast/multicast addresses. That's fine, and is on the client. One tells the local network that there is indeed a way to get to the Internet. That's fine, but often runs on the router and is entirely unnecessary on any modern operating system. Some advanced routers (e.g. Draytek) will have an option to leave that on, if you like. It's called "Connectivity Status". The other thing UPnP does is the port-forward thing. Every client asks for port-forwards. If your router grants them, this is by far not the first security problem with that. If you turn them off, the clients carry on regardless. Even weird stuff like videoconferencing, Steam matchmaking etc.

Before you start spreading nonsense saying that you "have to have UPnP", turn it off and see what happens. It's literally one click on your router.

Then tell me why you would ever want that functionality enabled on, say, a corporate network either, and why they turn it off from day one, and who's likely to be the biggest user of things like port-forwards and SIP / H232 / etc. protocols that all "need" that... yet it all works without UPnP.

Honestly, just try it. Nobody is even suggesting you have to ditch your local wireless devices, because they can use mDNS and UPnP etc. discovery over your local network, and connect to the Internet to do everything they need, without EVER HAVING to use it to punch as many holes in your firewall as they like.

TURN OFF UPNP ON YOUR ROUTER. Seriously. Not your clients, they can do what they like, because they can't punch holes in your security without the router's assistance and will just discover each other and work around it. And if you *didn't* know this, you really need to think why you're on an IT forum.

Lee D Silver badge

P.P.S. and turning off UPnP on your router will NOT stop local devices discovering each other via it.

Just turn it off, because having it enabled on any router will basically give all devices a free port-forward of their choice.

Lee D Silver badge

Nonsense.

UPnP allows ANY network device to request ANY network port on ANY external connection be forwarded to ANY internal IP/port combination, with NO AUTHENTICATION. Not one vendor has properly implemented authentication.

Turn that crap off, on all your networks, because even just "internally" it's not safe, and not necessary.

P.S. I have 1000 Steam games, a Chromecast, and all kinds of kit and none of it complains one iota about not having UPnP enabled.

British fixed broadband is cheap … and, er, fairly nasty – global survey

Lee D Silver badge

Re: Downloads

Google Play Store can throw you some doozies if you have a bunch of phones and a handful of apps that need updating.

An iPad / Windows / MacOS update can easily be gigabytes on its own.

Hell, the other day I downloaded a couple of movies from Google Play onto a mobile device and they were gigabytes before I even got a few on. Let alone, say, an entire TV series.

If you can burn through 50-100Gb in a month just browsing as an ordinary family, you can be sure that the speed at which you can download a 5Gb file matters a lot.