* Posts by Lee D

4259 publicly visible posts • joined 14 Feb 2013

Imagine being charged to take a lunch break... even if you didn't. Welcome to the world of these electronics assembly line workers

Lee D Silver badge

Re: "the cost of which would be automatically taken from their wages"

What kind of jobs are you applying for?

I've literally never signed on the dotted line without a full and explicit contract, with working hours, pay, break entitlement, a job description, etc. often signed *months* before I actually ever start work for someone.

All the other fancy stuff confirmed orally is - to quote a famous person - "not worth the paper it's printed on". Until I sign, I'm not committed. Before I commit, I *must* know the hours, pay, holiday and every detail. To do otherwise basically makes the contract negotiable, I just say "Nope, sorry, that's not what you said, and my contract does not mention those hours, so no... we don't have a meeting of minds, so no full contract, so no notice required".

My current contract is no different (and I have an electronic copy sent to myself via a verification service in case they try to claim "their" copy is somehow different years down the line, which basically guarantees what the contract said and when that was evidenced by an independent third-party) - it explicitly lists working hours, venue, breaks, responsibilities, holiday, pension, pay, and all kinds of things. It was signed and dated by both parties before my notice was handed in at my previous place.

"Not required by law"... that part's just nonsense. You do not EVER sign a contract without those details specified explicitly.

And you don't EVER work for anyone that insists that you do.

Lee D Silver badge

Re: "the cost of which would be automatically taken from their wages"

I have explained to several employers that if they start arguing over a minute here and there, they will be arguing over a lot more than just minutes.

I'm not a work-to-rule kind of guy, which is exactly one of the reasons it's probably not a good idea to try to make me one... I might well turn up early / go home late / skip a break if I'm busy. You don't want me working-to-rule. So complaining that I'm a minute late (so long as there's no urgent specific need to be there at that exact time that I'm aware of) is likely to lose you a whole lot more minutes that are being voluntarily donated to you elsewhere throughout the working year.

I have had this conversation with a new "office manager" at a previous workplace. They thought that being the office manager when *nobody* else was around meant that they could instruct the IT Consultant hired by their employer what to do. They had a go at me one morning for coming in five minutes "late". I explained that I didn't have a start time. They didn't listen and decided to waste ten minutes shouting at me.

Next morning, I was 10 minutes late. They shouted at me again, threatened to "tell the boss". I shrugged. No problem. I'm not paid by the hour.

Next morning. 15 minutes. They threatened all sorts and went to the big boss.

Who politely explained that not only was I an external contractor and - during that time - could pretty much turn up and go home whenever I liked, but that I was hired for a project delivery not for a time period, that I was far more valuable than the office manager (my daily rate was something like 5 times theirs, and all I had was a *deadline* to give the project by), that I wouldn't be bound by the office manager's hours anyway (I arrived later and left earlier than her every day for years, but she wasn't in charge then so she had never noticed), that the project was not only delivered on time but actually early (and thus more cheaply than they were expecting), and that I'd actually already saved the place more money than the entire project cost anyway by popping in on a few of my day's off to fix other things, clear out old equipment, get it recycled / sold on, etc. for free for them.

She never liked me much after that. But at least from then on I didn't have to stop on my bike a few hundred yards up the road to make sure that I arrived consistently later each morning.

Vivaldi to give abusive sites the middle finger with built-in ad blocking

Lee D Silver badge

There is absolute no reason for a website to hijack a back button. If they can't handle double-entry of forms by checking for a previous submission with a unique token, they are idiotic. Session control is not optional, especially when someone *actually* being able to use the back button can duplicate the form (i.e. your "protection" against such things is fake if you're just trying to fudge with the user's back button - a rogue party could easily resubmit their form another way and overwhelm your system with duplicate orders, etc.).

With HTML5 and modern scripting, there is absolutely no reason for it. Google Docs can save my form on the fly with no effort - to the standard of outperforming Office auto-save on a local machine. So can your form. Your poor design is no reason to ignore the user or fudge with their entries (because they may want to go back BEYOND your site to the search they had up before and messing with that just hinders the user.

Your poor session control is no excuse for breaking the user's browser functions.

Lee D Silver badge

A rule that *every* UI I've ever used breaks:

- The user gives you the orders, carry them out.

If they click End Task... end the damn task. If they click on a button, make that button perform the action. If they are clicking on a Stop button - that takes 100% absolute priority over the process they are trying to stop. If they are clicking a button in a UI... you make it show they clicked it IMMEDIATELY... and then queue the fractions of a second of task it takes to do, rather than "wait".

Linus Torvalds once expressed something similar - user actions are the highest-priority request on a desktop operating system. The file-copy can take a fraction of a second longer - but the button that says copy should depress immediately to let you know it's happening. No process on the machine should be so overwhelming prioritised that the error message takes a minute to appear, the machine chugs, the screen doesn't draw properly, the mouse stutters, the UI doesn't update, or anything else.

Not one desktop operating system that I know follows this. Not one mobile operating system, either. When I press the Home button on my Android phone - take me to the home screen. Everything else can wait until I'm there. When I tap an icon, initiate that action, everything else that's happening can delay until I'm not tapping the screen again. When I'm drag-dropping, don't let ANYTHING pop up or stutter the process, they can all wait for me to do what I'm doing - giving the computer a command.

When it's a server, we don't need to worry, but even there - if root says do something, on the command line - then just do it. It's more urgent than any hour-long RAID resync in the background and might well be critical to that RAID-resync and/or cancelling it if it's slowly trashing data.

By way of a bonus, this makes the computer feel a billion times more responsive even if it's actually taking slightly longer to do the background actions. Which is the right way round. I can bring a computer to its knees by scheduling the Windows Search database update when a user is in the middle of doing something. That just shouldn't be possible.

Nobody does it. And yet "user-initiated action" is very, very, very easily distinguished from background processes and running programs and every other type of input.

The user is king. Not doing what they say is failing at being a useful tool. Every example of a rogue action on a computer is a prime example of not respecting this - pop up windows, persistently annoying notifications, swipey-slidey-hidey menus, viruses being able to bring a computer to an absolute halt and stop the user removing them, Cancel buttons and file copy dialogs that just do their own things when busy and totally ignore the user, forcibly changing the default browser, etc.

All the UI/UX designers in the world have somehow forgotten / ignored this for 40 years, despite being frustrated in their own use of machines by the exact same thing.

Lee D Silver badge

Yeah, like all the other things they were going to do with it to make it "like Opera".

It's just Chrome. In fact, currently it's "Chrome that doesn't work for WhatsApp Web because we can't be bothered to revert the changes made in the last version". They literally won't do anything, and haven't released an update in weeks despite acknowledging that new problem within hours of the last release.

Vivaldi are an absolute, 100% complete disappointment. It's literally just Chrome with slightly different menus. All kinds of "Philip Hue colour" nonsense. They've changed the program icon 4 times. You still can't drag/drop arrange bookmarks (and it took years to get to the point where you could drag/drop bookmarks at all.

But, hey, yeah, you're gonna change the Internet with this new feature... just like the promised-and-then-years-later-ignored email client portion of Opera you were going to replicate.

The weird thing is, Vivaldi was born out of distaste with Opera which after v12 went Chrome too and stripped out all the interesting bits. Vivaldi basically did *exactly* the same, despite supposedly being the alternative "old Opera".

Sorry, but they can try to grab all the free publicity they like - they have made an average Chromium clone which they barely manage.

PowerPoint to start telling you that your presentation is bad and you should feel bad

Lee D Silver badge

Re: I have yet to see a Powerpoint of any worth.

A slideshow of images is a different thing altogether. The subject then *is* the image. Rather than repeating what the image clearly shows.

I found proper exceptions: Dave Gorman. And Prof. Brian Cox latest tour that I intended (purely because humungous 8K images of the deepest universe are pretty... but even he knows to just have them as background while talking about something that needs no informational slides for him to explain).

Lee D Silver badge

I have yet to see a Powerpoint of any worth.

Whether it's a (entirely third-party) YouTube video per slide, or just a bullet list that someone reads out, or just long blocks of text that have nothing to do with what's being said and are skimmed/skipped, I see no value in Powerpoints whatsoever.

The presenters who need them just need a prompt, that's all. The presenters who don't, don't need anything at all really.

There is nothing worse than having read the entire slide, getting the point, but being held to that wait-for-the-clicker moment to read the next one while people waffle on.

Honestly... if you want to do it... say "There'll be a summary Powerpoint sent to you after the talk, so don't worry about missing anything, let's just dive right in".

Similarly, however, I actually can't remember the last time I was in a room where someone was powerpointing where the *entire meeting* wasn't worthless. Either full of lies that never materialised, contained no useful material of interest and/or was not relevant to the majority of people there and could have been handled better with smaller individual meetings or even emails.

As someone whose job involves *setting up* for those kinds of presentations, that just baffles me. I've yet to load something up for someone who's giving a talk and thinking "Wow, that's pretty cool/useful/interesting".

Were I a CEO in charge of a company, emblazoned in place of the corporate "mission statement" would be a set of commandments instead:

- Thou shalt not Powerpoint.

- Thou shalt not conference call.

- Thou shalt not Skype/Whatsapp/Videocall.

- Thou shalt not call a meeting involving more than 5 staff without prior authorisation.

- Thou shalt confirm everything in email, even if a meeting took place.

- Thou shalt use email groups appropriately and in a targeted manner.

Ubuntu says i386 to be 86'd with Eoan 19.10 release: Ageing 32-bit x86 support will be ex-86

Lee D Silver badge

*Looks at every machine under his control*

*Fails to spot a single machine running 32-bit OS*

32-bit apps, sure, for backward compatibility. But 32-bit OS? Come on... the first machine that you ever used 4Gb on you should have realised. My minimum spec for *office workstations* (the least-heavily-used of all the computers I've ever deployed) is 8Gb. Whether or not you're playing tricks to access that remaining RAM, you should know that it was the end of the line, and the only excuse to continue was if you were upgrading the hardware of a machine that had started out 32-bit.

But if you haven't rebuilt or reimaged your machines in... what... 10 years or more? (And that was just Windows 7... let alone all the predecessors which had 64-bit support of some kind back to XP at least) What the hell are you thinking?

You can't even virtualise properly on a 32-bit OS, not really, and Windows Server has been 64-bit only for a while now. Those hints were all there. They've been there for a decade or more. Are people seriously buying new machines and installing 32-bit OS on them? Honestly? Where's your common sense? I can forgive not jumping into the "first-day-bandwagon" but come on!

I met resistance 5 years ago when I deployed this network (64-bit on every machine since day one, even the 5+ year old recon machines that they were using at the time) and I found that unbelievable - they honestly thought that 32-bit Office wouldn't work on 64-bit Windows (at the time, it was actually *recommended* to do that as 64-bit Office wasn't quite there!). I ignored them, deployed it site-wide, nobody has ever had any related issue whatsoever. I found it incredibly backwards to even be having that argument at that point, let alone now.

And Linux was *way* ahead of the game 64-bit wise. And had much better 32/64 mixed app support from day one. The last 32-bit only processor that Intel made were the Atom in 2010 and the Core Duo (not Core 2, which was 64-bit!) in 2006. That's literally the LAST THING that you *had* to run 32-bit OS on, let alone the last mainstream thing that you should have been buying to deploy anything serious on which would have been years earlier.

This is why security fixes are never deployed - people are still running OS that don't have support for the last 10+ years of processor instructions (and, no, 32-bit did not save you from Spectre or Meltdown neither!) including memory protection, virtualisation, etc. etc.

Youtube Queue Chrome extension booted out of store for search engine hijacking, revealing Google's lax dev checks

Lee D Silver badge

How many times:

Honour based security is worthless. "I'll ask for permission to read every website, but I promise I'll throw most of them away" is a stupid pointless exercise as even with the best analysis in the world you can still miss something that's doing that you didn't realise.

Fine-grained permission control is literally the *only* thing worth having, and doing it properly means that your app approval process is greatly simplified. Have a bar... on that bar is a tiny little icon for each extension (whether the user wants it or not). So when they are on the bank and that little icon is still lit up THAT EXTENSION IS READING THAT WEBSITE. They are the ones stupid enough to have said "Yes to All" without thinking it through but maybe the reminder will prompt them to realise that that wasn't the brightest idea.

If you have "permission to read" and "permission to talk to the Internet", you have implicit permission to "do what the hell you like with my data and spread it over the Internet". It's just that simple. An app approval process doesn't even hinder that.

The way to stop it is to make it so that it's just not possible.

There will come a time when Microsoft, Apple, Google, etc. will learn this - likely the first "hard" lesson of one will teach the others. But everything from UAC to iTunes app permissions to Chrome extension security models are messed up and just wrong. We learned this with Java, with Flash, with ActiveX, with the operating systems of the 90's ("everything single user as admin"), etc. but we keep committing the same mistake over and over and over again.

Don't "look for people who might be trying to jump over the fence". Build a fence that they can't jump. It's ironic that the Android app permission model is so much better than the others... but still useless unless it has a "don't give them that permission, but let them believe you did" so you can use the app and it has no say in whether you can or not depending on what security you gave it.

Yes, people are dumb and will Yes to All if they have the option. This isn't surprising. But then the blame is on them. You make it big clear warnings, including the word "THIS SHOULD NEVER BE NECESSARY, WE HIGHLY RECOMMEND YOU DON'T DO THIS" and flag warranty bits if they do that taint the OS like Linux taints the kernel if you're silly enough to tinker with untrustable code... we do it for "rooting" the device, why not for loading it up with potential malware? Maybe people would think twice and app manufacturers would flee from ever asking for those permissions if we did that? In the meantime every "approved" extension/app is just another reputation-loss for whoever approved it in the first place, and none of the major players are immune from that.

Parliament IT bods' fail sees server's naked OS exposed to world+dog

Lee D Silver badge

If that allowed read-only access to, for instance, the NTDS or other password files.... then ouch.

John The Ripper can take those and crack the passwords offline, and then use them to login elsewhere, most likely.

And I'd guess that any web server is probably holding at least SSL private keys... again... ouch.

23. 712. 3. 608. 45. 89. 11. 332. 841. 255. You want more? Cloudflare and pals are streaming 'em live from new RNG API

Lee D Silver badge

Re: The best RNG so far is

Clearly they were either so pressed for every single instruction, or someone couldn't be bothered to read Numerical Recipes and code up even a basic one.

This from the same brains that later used a wonderful mathematical optimisation (https://en.wikipedia.org/wiki/Fast_inverse_square_root) when they had buckets more processor power to use.

Lee D Silver badge

Re: What's it Good For?

Yeah, because the people who run the lotteries don't already do that.

Lucky Dip anyone? From the people who run a website basically premised on the idea that they generate sufficiently random numbers to use on / allocate you a scratchcard from a pool of available ones?

UK industry calls for delay of IR35 off-payroll tax rules to private sector

Lee D Silver badge

I would severely question what software you're using to do payroll if it can't update between November and April along with the entire rest of the country, and the government systems, and everything else.

I get it's not instantaneous, it may not even be bug-free years later, but to support an "IR35" set of fields, calculations, and records... that doesn't take six months. Especially not if your industry is software that has to legally keep up with the moving-target of tax legislation.

I know where I work, I have to update the finance software whenever it changes, and that happens a lot just before April to make sure it's right for the new financial year. I'm sure you can push an update in six months.

Those people *not* using up-to-date (or unable to be updated) software to run their payroll... well, that's their problem. At worst they should buy into one of those online services that does it all for you on a website.

If your development process is that bad that a well-documented taxation, the details of which will be available long before November, is approved in November, and then you're not at all sure how to get that into your software for next April... when the governments own online taxation portals will similarly be updated for 70m people, maybe you should be taking on a few of those ex-public-sector programmers!

When customers see red, sometimes the obvious solution will only fan the flames

Lee D Silver badge

Re: Dolt

"If your company has produced software with a (for example) timing bug which can't be reproduced at will by the user, you're still responsible for fixing it. Just because you can't reproduce it doesn't mean it's not a major PITA for your customers."

And the first question they're going to ask me on the support line? Can you reproduce that for us, because we can't.

Again *I* cannot fix a problem that I cannot reproduce. Others may be able to - and if it's a supported piece of external software, that's up to them.

Yes, in my career, I have had to say "I'll call the company. You know their turnaround on things like this is about 10 weeks, don't you? No, I'm sorry, I can't recode their program for them. They have to fix it. That means we have to wait for them to fix it. I'll try to find you a workaround, but other than that, it's out of my hands." In fact I've said it on a regular basis.

Just because it's a major PITA for my customers doesn't mean I can reproduce it, and even if I can reproduce it, actually fixing it is often someone else's problem.

(Say Word crashed every time you opened a particular file - no matter what you do, that file never opens, even on a clean install, even on an up-to-date install, on every computer, every test, it crashes - do you think you can *do* anything about fixing that? I don't just mean "recreate that file" but actually fixing the problem? No. Only Microsoft can. And even if you recreate the file, what if it's what you want to do that causes the crash, i.e. it doesn't like the data pulled from the production database? IT are not programmers - actually some of us are - but we're certainly NOT programmers with full development environments, complete source code to every program we deploy, and the ability to create patches for every problem as part of our normal working day.

The reason we pay for support contracts is for others to fix the problems they cause. I currently have at least three support contracts and one of them has something like 20 outstanding and reported issues, one of them from 4 years ago, that has never been fixed. There's *nothing* I can do about that, while the company wants to continue to use that software. In this case, the upheaval of removing that software is orders-of-magnitude worse than the outstanding issues, and the outstanding issues I cannot fix - I even produced an Excel that, using the data in the associated database, and after much churning, produces the output we desire. They still haven't added that function into the database themselves, and they probably have no intention to. And, yes, we have threatened to leave. They do nothing.

Sometimes, all IT can do is say "That doesn't work, don't do that."

Lee D Silver badge

Re: Dolt

Welcome to English.

"absent other X" means "without anything else X".

Google examples and you things like "In other words, it is considered to be true or at least adequate on first appearance, absent other information or evidence." from law sites and all sorts.

I didn't just make it up.

Lee D Silver badge

Re: Dolt

1) Can I reproduce the problem? No: then it's specific to the user / computer.

2) Does the problem occur for that user on other machines? No: then it's the machine in question, Yes: then it's the user account or user themselves.

3) What exactly is that user doing / typing / clicking to make it do that?

Not saying that it's 100% infallible, but pretty much every single ticket that falls through that path on those questions will end in the results listed.

If I have spent my life in IT and learned anything, it's that 1) on its own usually means the user is doing something really dumb, and every "actual problem" fails that first test (i.e. I can reproduce it).

I also have it well-documented and advertised to all users: I cannot fix a problem if I cannot reproduce it. It's that simple. Telling me that last week something funny happened and not be able to tell me what that was or show me it again means I can often do absolutely nothing about it.

Sure, maybe I need to be you, sitting in your chair, on your user account, using your computer to reproduce it. But if I *cannot* reproduce the error, I cannot fix it for you. If you, as a user, *can* reliably reproduce it... then I can probably fix it. Or tell you why I can't (e.g. "Microsoft didn't write it that way").

File a ticket. If I can't reproduce it elsewhere, I may well ask for a screenshot, or take remote control. Or walk to your office next time I'm passing. Pretty much, at that point, you just *know* it's going to be something the user's doing, absent other flags (i.e. "The computer is saying the hard drive is failing").

US can try extraditing Julian Assange next year, rules UK court

Lee D Silver badge

Re: Cant see

Cheaper to incarcerate than man a police operation to surveil him for seven years, you mean?

Lee D Silver badge

I'm gonna hazard a guess and say that every judge sitting on criminal trials has been called a lot worse almost every single day of the year, to their face, by the accused, and had to remain entirely impartial.

UK Home Sec kick-starts US request to extradite ex-WikiLeaker Assange

Lee D Silver badge

Re: Sweden

I'd like to know how you think they obtained a sample of his DNA to test if his DNA was on the items, as you've alleged.

I can find no reputable link that mentions any nonsense like that. In fact, they *think* they found his DNA on one of the condoms from the other women - but they can't confirm without a DNA test. Hell, the other one could have just been the wrong one pulled out of the bin.

If he's that innocent, it'll all come out in court, she'll be made a fool of, end of story. Hell, if it was *that* much a cover-up all that nonsense you spout wouldn't be out in the public domain, would it?

P.S. DNA is far from infallible.

US border cops confirm: Maker of America's license-plate, driver recognition tech hacked, camera images swiped

Lee D Silver badge

Re: Subcontractor’s network compromised?

"the unnamed contractor"

When it comes to DNS over HTTPS, it's privacy in excess, frets UK child exploitation watchdog

Lee D Silver badge

In other news, telephones can be used to organise crime or save lives, paper can used to write a novel or commit arson, and cars can be used to transport emergency victims or run people over.

Could you just pop into the network room and check- hello? The Away Team. They're... gone

Lee D Silver badge

Re: At home

I work at a school. I had been explaining the reasons behind having fibre between buildings, even if they were small local runs of only a few feet, especially if the electrical phases between those buildings could be different (now or in the future!). Pretty much, everyone ignored me because the cabling had been in place for years and they didn't see it was worth changing it.

One night, a *huge* tree that is on the site was hit by lightning. It coursed down the tree, found some aerial cables including the cable that ran between a wooden maintenance shed and a permanent building... blew the telephone in the shed into smithereens. The other way, it entered the building and damaged a 48-port switch (which I still have to this day... the middle 24 ports are absolutely dead, and some of the others won't give PoE but everything else works, including the cloud management... we swapped it out and it's now our "test bench" switch).

From the building, it also managed to find some random and exotic bits of hardware, another telephone line that went through the air to some staff accommodation a hundred metres away, blew that up, took out a DSL router and a computer connected to it (not just the PSU, but the whole machine).

However, that same switch also ran a fibre back to the next building where several dozen machines and other critical pieces of equipment were all copper-cabled, not to mention several dozen phone lines, connected to all kinds of electrical equipment, all underneath where the boarding pupils lived. Not a dicky-bird, because the power couldn't get down the fibre. Everything else just stayed up. Fortunately, that building was only fibre all the way out to any other building.

And, again fortunately, I had the sense to run not just fibre between buildings, but also fibre between distant buildings (so if A is connected to B is connected to C, via fibre and switches at each, then I also connected A directly to C with more fibre so that when B was offline A and C could still talk). And then also connecting C to D and D back to A so that STP could do its job and provide a redundant route.

Basically, the fried equipment was more than enough to potentially start a fire in three separate buildings, but luckily didn't. The network just saw a switch go partly offline and routed around everything else. And from then on, every connection that went through open-air or the ground was changed for fibre.

It's official! The Register is fake news… according to .uk overlord Nominet. Just a few problems with that claim, though

Lee D Silver badge

Honestly... do these vanity-plate domains generate money for anyone other than the registrar?

Who's going to "food.uk" instead of waitrose.co.uk? Which search engine is up-ranking sites because their domain name is "rare" even though they're just a bunch of ads? What typo-squatter is actually doing anything with their domain except put up a boring page of almost-links that don't actually even link to the domain they're typosquatting (for obvious "passing off" reasons)... why, yes, I did mean to go to theregsiter.co.uk, and click an advert for Nike trainers, rather than the tech news website...

I knew someone many years ago who used to "snap up" domains they consider valuable... I'll change the name slightly with synonyms but do you really think student-flat-search.co.uk is a valuable domain? What about financetohelpyoubuy2let.com? Now, that was the 2000's, so they actually made a small profit by selling those domains to other mugs who thought they were valuable, but they also got lumbered with a bunch of unsaleable domains that they had to keep registration on for years and then just dumped.

Who's buying these things, what are they doing with them, why, and are they actually making any profit whatsoever? I hold a bunch of domains that the kind of sites that sell them keep contacting me about. Because they were just... "Oh, that sounds nice, I'll use that for my site", not because I wanted to profiteer. They *claim* my domains are worth thousands. Not once has anyone offered anything even approaching that. Still, today, if I put them on that big domain auction site, they get high valuations and *zero offers* at any price.

The only one I "kinda" get is buying up the misspellings of facebook.com if you're Facebook, but even that seems a bit of a pointless exercise and will only encourage people to register even more misspellings in the hope of a payout. I imagine most of those are actually the result of a court case where Facebook sued the backside off someone who was passing off, and won the domain as part of the settlement.

BT to axe 90% of its UK real estate, retain circa 30 sites

Lee D Silver badge

"BT employed 83,000 people in the UK and 23,000 overseas as of May last year when the process started."

100,000 staff? Doing what?

Can't quite cram a working AI onto a $1 2KB microcontroller? Just get a PC to do it

Lee D Silver badge

Re: Emperors new clothes and neural network humbug

ANN is just "a statistics based magic box".

You plug things in. It makes some kind of spurious and random correlation between what you're teaching it and the input data, which you can't really interrogate, understand, improve or modify.

If you plug enough things in, it might train itself enough to work a percentage of the time. Then untraining, retraining or anything else? It's pretty much throw it away and start again - it's existing "superstition-based" intuition will trip over every exception to the point that it becomes painful to make any significant change after the learning plateaus.

It's been an unsolved problem for decades, we just get to throw more and shinier hardware at the same problem.

You would otherwise notice, for instance, that Google searches would become personalised, almost-psychic, tailored to every user... Siri would know what you want before you ask it... because millions upon millions of users are training it daily and it doesn't sleep - it should be learning exponentially. It's not. And over time everything you interact with that has "AI" would get better and better... it doesn't. Improvements are microscopic at best after the initial plateau.

Literally Siri and Google should be Skynet by now. The reason they aren't... AI and ANN in particular just doesn't work like that.

What we need is a really, really radical re-think of the whole thing with something entirely different. Because genetic algorithms are the same, ANN, everything we try hits the same problem. And for every million images you train it on, it takes a significant percentage of those images again to retrain it for the ones it gets wrong. And again. And again. With diminishing returns.

Lee D Silver badge

Re: What is new here?

It's now cloud IoT hyper-convergence AI neural nets with fabric... and chintz and...

Pretty much, nothing's changed in the meantime except we have slightly faster computers but pretty much the end result is no different... we can just "afford" to bundle this junk into little speakers and your search menus where we couldn't before.

In terms of things actually *learning* or doing useful things, we're still stuck in the absolute dark ages of the technology where really the problem is "unlearning" - i.e. having a machine that trains itself on a million images is all well and good, until you spot something it's doing wrong and you have to basically retrain it from scratch because it has 1,000,000 entries that say it's doing the right thing, and one that says it's doing the wrong thing and it has to resolve that somehow without losing all the subtle nuances that it was trained on (i.e. you can't just weight the error 1,000,000 more than the others).

Ai, since the 80s and through to today, learns and then plateaus just on the cusp of usability and then *stays there forever*. It's almost a perfect PhD research topic - do it, write it up, hope nobody ever asks you to apply it to anything else that doesn't involve literally starting from scratch every time a change is made.

And, even then, it's generally only 90-something % accurate which is pretty useless compared to even a trained dog.

Lee D Silver badge

"Ah.... so you're a waffle man..."

(Red Dwarf, for those who don't get the reference)

Apple kills iTunes, preps pricey Mac Pro, gives iPad its own OS – plus: That $999 monitor stand

Lee D Silver badge

Hold on, I can't comment properly, I haven't stopped laughing yet.

$6k for a $2k basic model, god knows what the top model costs but I bet I can match it for less than that "basic" $6k. About the only thing even vaguely sensibly priced is a VESA mount... Apple really know their stuff... and a 6K monitor that I can buy an 8K equivalent for cheaper just shopping on Amazon.

The cheesegrater PC is about the only thing that looks different to being just a box, and that's not a good thing.

Everything else... all that software stuff... long overdue and not gonna help against prices like that.

Hopefully this is the beginning of the end. There's a reason that professionals are steering clear.

If you work in a company where they think they can justify this kind of expense... sack your IT guy and find another company.

Sunday seems really quiet. Hmm, thinks Google, let's have a four-hour Gmail, YouTube, G Suite, Cloud outage

Lee D Silver badge

That's what secondary DNS providers are for.

It's almost like what I say is applicable to everything AND can be done at the drop of a hat AND for minimal outlay, if a day's GMail outage has, say, more than a £10 a year importance for yourself.

If you're using it because it's free, you don't really have much weight complaining that it's down for less than 1/1000th of the year on one occasion.

Lee D Silver badge

To be honest, long before the cloud, Hotmail used to have regular outages... sometimes several a year for a day at a time.

This isn't "cloud" problems... this is a "you chose a single supplier" problem. If you had, say, a domain name that you could redirect your mail to, or which copied your mail to two separate accounts at different providers, you'd have avoided all such problems. Or, at worst, redirected mail somewhere else.

The second you chose *one* supplier, you're at their mercy. Of course you shouldn't believe their "we're cloud, so we're always online" spiel, but past that you have nobody to blame but yourself.

Apple's privacy schtick is just an act, say folks suing the iGiant: iTunes 'purchase histories sold' to data slurpers

Lee D Silver badge

Re: This could be interesting

I can't.

The guy writing the court summary can if you read the PDF:

"These factual allegations are corroborated by publicly-available evidence.

For instance, as shown in the screenshot below, the Personal Listening Information of

18,188,721 “iTunes and Pandora Music Purchasers,” residing across the United States

(including in Michigan and Rhode Island), is offered for sale on the website of Carney

Direct Marketing (“CDM”) – one of many traffickers of this type of Personal Listening

Information – at a base price of “$80/M [per thousand records]” (8 cents each): "

"SRDS, another list brokerage company, offers for sale the same or a

similar list as the one sold by CDM, at the same price, and additionally offers a finder’s

fee to brokers who are able to find purchasers of this Personal Listening Information

(offering “20% commission to brokers” and “15% commission to agencies”), as shown

in the screenshot below of a publicly-accessible webpage on SRDS’s website:"

And it doesn't matter if Apple *weren't given* your age, income, education, etc. - they correlate it from all kinds of other sources and link it against your iTunes account once they've identified you. Think "buy the same kind of data from Facebook, Google advertising, etc. and then correlate the IPs":

"First, Apple discloses its customers’ Personal Listening Information,

identifying the names and addresses of its customers and the particular genres of music

they have purchased from its iTunes Store, to data aggregators, data miners, data

brokers, data appenders, and other third parties, who then supplement that information

with additional sensitive personal information about each of Apple customers,

including their age, gender, purchasing habits, education, household income, and

(when applicable) the number, age, and gender of the subscriber’s children. "

Lee D Silver badge

Re: This could be interesting

It's in the article.

There's literally an API where you can buy that data. That's the point. As a developer you can just pay Apple, query the API and get thousands of names.

Whether anyone *does*, only Apple knows, but that's not the point. It's like literally having a "download any of our customer's data" button on their website, with search fields and a Buy Now button.

And, I would presume, that to initiate the lawsuit and make those claims, this guy - or his lawyer - pressed such a button, paid the fee, and got back real data.

Game over.

Tesla's autonomous lane changing software is worse at driving than humans, and more

Lee D Silver badge

Re: Crikey. Where has common sense gone...

I tried literally everything I could think of to make my electronic handbrake fail. It steadfastly refused to do so. It also deliberately failed to activate inadvertently no matter what I tried. Whoever designed the switch in the new Ford Mondeo's did their job - just a perfect amount of debouncing/whatever to mean that a snap of it doesn't do anything, but a press-and-hold causes you to beep the horn using your forehead and nose. Same for the anti-rollback / auto-hill-start functionality. I can only make it release by doing really stupid things.

By law, it has to have an entirely dependent activation path. Sometimes that's an entirely independent ECU just for braking, sometimes it's part of the main ECU. Sometimes there's an entirely independent emergency power supply or power reservoir of some kind just for the emergency brakes. There are a lot of ways to do it, and a lot of regulations around it.

Read this, looking for "parking" in the PDF:

http://www.unece.org/fileadmin/DAM/trans/main/wp29/wp29regs/R13hr2e.pdf

The paragraph you're probably looking for:

"In the event of a failure of the energy source of the electric control transmission, starting from the nominal value of the energy level, the full control range of the service braking system shall be guaranteed after twenty consecutive full stroke actuations of the service braking control. During the

test, the braking control shall be fully applied for 20 seconds and released for 5 seconds on each actuation. It should be understood that during the above test sufficient energy is available in the energy transmission to ensure full actuation of the service braking system."

What that means is, even in the case of total electrical failure, I should be able to put the brakes on at least 20 times from the handbrake. A large capacitor can do that. It's just a solenoid, after all.

Lee D Silver badge

Re: Crikey. Where has common sense gone...

I purchased my first ever car-that's-from-new about three years ago.

I deliberately specified none of that crap. There were options, way within my affordability, for lane-assist, road-sign assist, auto-cruise-control, etc. etc.

In my opinion, those things don't belong on a car. If you need them, even for a brief moment, you should have pulled over long ago.

The basic model, though, comes with a dangerous-by-default config in my opinion. It has cruise control. But that cruise control is not-auto-brake. Hence it will happily accelerate to the set speed, no matter what's in front of it, at the press of the button. But it won't regulate that speed or prevent an accident by stopping before it runs into the back of something.

Now I know that's "traditional" as such, but it bugs me. Auto-brake is a paid option. Why? It should be the other way around. Especially seeing as "speed limiting" is a built-in default option that can actually operate on a per-key basis (presumably so you can give a key to your kid and it stops them going mad).

However, I have tested it extensively, and the brakes always override, and the brakes can override the engine, and - being a manual - you can slip it out of gear even if it goes temporarily insane (assuming you react in time). Same as the electronic handbrake switch, which I do not trust and tested extensively before any serious use on the road*. It's placed so that it can be activated by the passenger accidentally, but I couldn't do it accidentally, even "simulating" a bag-strap getting caught - at least, no easier than if it were caught on a manual handbrake.

I judge people whose choice of car includes such options (except where they are by default, but even then!), and especially those who don't endeavour to disable them or avoid their use by habit.

*(My forehead still hurts. Seriously, don't mess with those things. Handbrake turns are impossible but for sure I like the way that system brakes independently even with no engine power... it does NOT mess around.)

Yeah, you're not having a GSM gateway, Ofcom tells hopeful operators

Lee D Silver badge

Re: You mean, like, Caller ID?

I just whitelist.

Unless you present to me a known CLI that I have chosen to accept in the past, nothing gets through. You can't contact me? Shame. Maybe that's because I never gave you my number in the first place (or didn't want to but was forced to, or someone else gave it to you, etc.).

For personal use, I just mute the default ringtone and make known contacts have a ringtone that sounds.

For business use, I only bother with SIP and then it's just a matter of pushing them through a SIP-based menu if you don't recognise their number from your whitelist. They don't even connect to my site unless they can get past that menu (which has an option to leave me a voicemail, which I treat just like unknown spam email).

If you're hiding your CLI, I don't want to talk to you (the old joke: You want to hide your number, fine I'll 100% respect your privacy by not taking your call...). If you need to contact me, you'll be given an option to leave a voicemail. If you can only get through by calling from a proper CLI-number... oh what a shame... that's like only being able to send me an email if you come from a proper domain, with proper reverse DNS, that's not on a blacklist, etc. etc. etc. No different. At worst case, I google the CLI phone number and if nothing comes up, I assume you're a spammer.

And if it's anything even vaguely important, like an unpaid bill, a contract you need me to sign, or anything along those lines... you'll send me a letter. And I'll discuss it with you at my convenience.

I've cut the landline. I have a 4G box instead. The number of that is my "spam" number because... well... who cares, it can't even ring anyway. My mobile phone (20+ year old number) operates a whitelist like above. I get less spam on that than anyone else I know. My inconvenience factor of "someone couldn't get hold of me" is basically zero - people I know have no problem getting through. And my backup plan is that I have a Draytel SIP account in the background, connected to my router. If I need to, I'll move to that, and send you through a verification switchboard menu first.

Fact is, the last 10 years I've had that option and I've not been pushed to it yet.

I'm on the TPS, I don't sign up with my numbers to anything, and I get almost nothing anyway - maybe a number that people online say is a "have you had an accident" caller once every three months or so. I'm presuming they just try every possible 10-digit code or whatever and hang the TPS. They don't get through. 99% of the time they don't even get to a voicemail. I don't have the time to mess with them or play games or even answer them so I don't... if my phone is ringing from an unknown caller and I need to use the phone, I just press hang-up and dial the number I need.

I've not yet missed anything important in 10+ years. I don't see why others would suffer the way they do. A friend of mine just bought one of those services/phones that blocks so-many-CLIs and things... it's useless, and the blacklist fills up very quickly, and they have to PAY for it... to the same people who are delivering those calls and being paid to do so in the first place. Sod that.

If it ever really becomes an issue for me, I'm only about 2 days from getting a bunch of 07 numbers from a SIP provider, or something like Skype, and just let them get spammed to oblivion and never answer them to anyone I don't know. Cheaper all round. To be honest, I think I'd just do it through my router or a PC with a load of 3G dongles... and just load Asterisk, FreePBX or similar to send them in circles if they are unknown. It's not yet come to that, though.

Why a telco thinks they'll see anything of my money while this is happening and they aren't taking the simplest of actions against it, though, i can't fathom.

Never let something so flimsy as a locked door to the computer room stand in the way of an auditor on the warpath

Lee D Silver badge

Re: First step, tell them to fire security

Much, much more valuable than any amount of "security" stopping people doing things is to *let the only person who cares* know that something's happening. If you got an alert the second a door was opened and shouldn't be, you can stop it in action, deter future attempts, etc.

The problem is that even most business have "an alarm" that doesn't actually alarm anyone - it annoys the neighbours and that's about it.

If you care about security - whether student flat, shop front, or large warehouse - you have to let the right people know. An alarm that's monitored and can invoke a police response is good, but if you were to text the guy who owns the company, they can do things like log into CCTV, check it's a crime in progress and ring 999.

Your "security" only needs to be a security canary - enough for something to fall over and trigger when something happens that shouldn't. (same goes for antivirus, by the way, which I refer to as the "virus canary" when users asks what one they should spend a fortune on, and why their expensive one didn't notice the quite-obvious virus on the machine... when it falls over, you have a virus, it's that simple, and that's its real job). Hence most big locks / systems are pointless.

A system that alerts the right people (which includes the burglars but notice that it *doesn't* include the neighbours as they are useless and will only be annoyed by any noise if it goes off), in good time. Everything beyond that is really pointless.

I watched a police programme the other day where a huge Mothercare was broken into. It was the *police* arriving, after a member of the public reported the shop being "left open", that called the keyholder to let them know they'd had their entire front glass door smashed in, and clothing nicked.

At that point, I really wondered what the point of them having an alarm at all was.

Lee D Silver badge

Re: First step, tell them to fire security

Insurance companies only care about forced entry.

All they ever need is something that prevents casual entry, and has to be damaged in some way to gain entry. Then it's a crime, it's insured, etc. etc.

Your insurance standards for security are incredibly sub-par because of this. Most datacentres will want to vastly exceed such nonsense, because the insurance of "office computer got nicked" is very different to "datacentre holding government data was ransacked".

Our insurers are the same - they just need enough to show forced entry. Some of our main servers are literally a hollow plasterboard wall away from being stolen. That doesn't bother them and, while they will pay out for anything nicked, it doesn't bother me.

The bigger issues of "data security" should *anything* be nicked is another matter entirely, and quite clearly IT's remit, i.e. if you nick my servers, you can't just log in and pull all the data off them.

Think about most offices, businesses and houses. One smashed window and you're in, literally seconds and a brick and a bit of bravery. But it's quite difficult to gain entry without causing visible damage in some fashion.

Lee D Silver badge

Well, the door did its job - it prevented entry long enough against a determined attacker until entry was forced through it.

The security guys didn't. Blame them.

Oh 4G, I'm speechless: EE network outage smacks rare breed of customer that talks into their mobile phone

Lee D Silver badge

I'm still wondering what idiot named a network "3" thus making quite a lot of things un-google-able, not to mention look outdated the second 3G was superceded, making itself difficult to understand to the uninitiated ("Why would I want to use a free network?").

It certainly wasn't the person who named their customer app "Wuntu", but it might well have been the person who recently gutted it of pretty much all the useful benefits.

Bug-hunter reveals another 'make me admin' Windows 10 zero-day – and vows: 'There's more where that came from'

Lee D Silver badge

If a user can create an arbitrary scheduled task to run any given executable, it's game over anyway.

That Windows helpfully repermissions your file that you want to run from a legacy imported scheduled task is really just icing on the cake.

Presumably pci.sys is used because it's a "known" signed file that Windows trusts anyone to activate?

Ordinary users should not have the capability to schedule tasks, nor should they have the ability to access the folder where scheduled tasks are kept, nor should they be able to execute arbitrary executables. Hell, they shouldn't even *see* the scheduled tasks panel, there could be privileged information in there!

The problem is not some "new" exploit... it's just the same old complete lack of security on basic features, and "trusting users" the same way people did back in the days of Windows 3.1 or DOS.

Backup your files with CrashPlan! Except this file type. No, not that one either. Try again...

Lee D Silver badge

Re: Just change the filenames ?

To be honest, it takes seconds to run the "file" command against any file and it'll tell you what the real type was, whether the extension is correct or not. It'd certainly spot VM images.

Of course, all you need do is what you SHOULD BE DOING ANYWAY. Encrypting anything you "backup" to random third-party servers somewhere else in the world.

Proper encryption, resulting in a ".bak" or similar file, rather than just copy/paste your VMs as they are for anyone who works at that company to poke through.

Swedish prosecutors request Assange detention: First step to European arrest warrant

Lee D Silver badge

Though the presumption of innocence is required in a court of law, it's not something that meshes well with fleeing from the law and breaching bail conditions. This guy ran away from a UK court's custody because he didn't want the possibility of having to stand before a Swedish court.

Sorry, I probably couldn't be impartial in that case. Same as when people resist arrest without good reason, or mouth off as they are being arrested for something (or, indeed, not being arrested at all until they start mouthing off and get themselves arrested).

There are certain things that innocent people just don't do. My sympathy plummeted to zero when he did that, and further down when he decided to hide in an embassy and preach from the balcony every week.

On another note, it's been the quietest I've seen in a long while for Assange-related tweets and soapbox moments. If I start a kickstarter, can we keep him banged up for the full 50 weeks instead of letting him out half-way?

Exclusive: Windows for Workgroups terror the Tartan Bandit confesses all to The Register

Lee D Silver badge

Re: I'm boring

What tool are you using that allows remote reboot/shutdown as an option from the desktop screen without using a CLI to override?

Get a better tool. Even RDP just removes that option, and you have to "shutdown /r" in a console to make it do anything like that.

It's a technique I use - a nice red background for the admin users suffices to make things clear that there's danger ahead or you're in the wrong window. Because, obviously, you don't work day-to-day as an admin user, right?

If only people who worked as administrative users on server-grade hardware remotely (e.g. for customers) could take a second to, say, double-check before they press "Restart Now"...

Lee D Silver badge

Re: I'm boring

Every desktop background I have is a solid colour from the bluey-green range.

Plus there are about four icons on the desktop and everything else is under a categorised and alphabetised start menu (Thanks, Classic Shell!).

I honestly don't get the reasoning for having a bitmap constantly loaded on the back of the desktop at all times where you only ever see it when it's in the way of reading your damn icons.

Wallpaper, screensavers, desktop icons, any sounds whatsoever (I know I clicked, my mouse clicked, I don't need the speakers to click too, and whoosh, and whatever other silly noises), side-bars, taskbar icons (except the essential). All disabled.

Strangely, personal or professional, I just want to run the program I need, have it start, and then use it.

Don't even get me started on that transparent window junk...

Bloke accused of conning ARIN out of 750,000 IPv4 addresses worth $9m+ to peddle on black market

Lee D Silver badge

Re: *COUGH* "Now if everyone would just move to IPv6... "

P.S. Any IP address of any kind is under no obligation to respond to ICMP anyway. It's nice, sure, but not necessary or required.

Google are telling people that over 25% of their accesses are coming in over IPv6... 1 in 4 people around the world are using it when they go on Google. Do you think they all sat there and deliberately configured it, or do you think it's just a base protocol that's part of the requirements of major technologies like 4G?

https://www.google.com/intl/en/ipv6/statistics.html

Lee D Silver badge

Re: Using IPv6

It wouldn't buy enough time.

Routers would NOT be simply adjustable to that format, they wouldn't have been designed with that amount of internal RAM and processing and all their processing speed would break down for even ordinary packets - they would have to be reprogrammed, packets sizes would increase, fragmentation problems and MTU problems would increase, old kit would never get upgraded, and most old kit *couldn't*... just a BGP routing table hits the RAM limits nowadays, suddenly increasing every IP and mask by doubling the size of it knocks it out of the park on anything but new hardware.

TL;DR: Old kit would break, slow down, or just never be able to accept the new addressing.

And in the meantime, you'd still end up in the same problem - in about 30 years we've exhausted IPv4, with exponential growth expected.

IPv6 was invented ~15 years ago, to solve the problem once and for all (because if you have to do all the above, you may as well do it once and properly so you don't have to do it again in another 15 years). So all equipment from ~10 years ago onwards was redesigned with IPv6 support. One redesign, decades upon decades of future-proofing, plenty of space for everything you ever need, change it once, spec the router/RAM/whatever accordingly and then you're covered. That happened in the WINDOWS XP/Vista ERA. Done. Gone. Happened. OS included.

And IPv6 already has shorthand... because even in hex, an IP is for example: 2001:0db8:0000:0042:0000:8a2e:0370:7334

In dotted decimal, that would be something the size of:

192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.192.168.001.001

(every two characters in hex is two dotted decimal numbers up to 255!). Easier to remember "ab" or "160.161"? You literally don't understand how big IPv6 is and the problem it was solving for all foreseeable future extensions.

And: 2001:0db8:0000:0042:0000:8a2e:0370:7334 That can be shrunk to: 2001:0db8::0042::8a2e:0370:7334.

Not that much longer / harder to say that your eight dotted decimals from 0 to 255 each. Imagine having to say "2 5 5 dot 2 5 5 dot 2 5 5 dot 2 5 5 dot 2 5 5 dot 2 5 5 dot 2 5 5 dot 2 5 5 dot" out loud to an amateur, with random numbers instead of nice repetitive ones.

Two thousand and one. "O" "D" "B" "eight": "double-O forty 42":skip it: "eight a two e", "O three seven O", "7 double-3 4". And I've chosen a very long, quite incompressible example (from a random website) that conveys an address from BILLIONS of your IPv5 networks, not just one. It really didn't take that much more, if anything.

And IPv6 also supports compression of the kind you mean (where any multiple 00 can be compressed out of the string by just removing ), the first 64 bits are allocated to you and won't change, and you do not need to worry about all the end digits - as the last 64-bits (8 bytes, or 8 hexadecimal characters, or 8 dotted decimals up to 255) are all yours to do with as you like.

So IANA gives a number using a bunch of the first bits to regions, ARIN etc. allocate numbers with their bunch of the second bits to ISPs, your ISP gives you numbers with all the above, and you get the last bunch of the bits to yourself to do what you like. The same way that MAC addresses work. The initial HEXADECIMAL! bytes are the manufacturer's allocation, the following ones are the individual devices. Note that a MAC address (12 hex digits) already conveys more information than an IPv4 address (4 hex digits), more than your IPv5 address (8 hex digits), and are bandied about, entered into device management portals, copy/pasted, read out over the phone, etc. already. And you don't want to confuse and IP with a MAC or even be close to doing so.

Hierarchical, ordered, future-proof (i.e. only 3% of the IPv6 addresses that exist in current plans will ever be used for the Internet itself, the rest reserved for other uses), and everyone along the chain gets SO MANY addresses that they can issue that any possible use they think of for billions upon billions of customers is covered into the future.

And yet your IP still just looks like:

FE80::0202:B3FF:FE1E:8329

where the last few blocks of letters for most everyone will be squished to zeros using :: and 1, 2, 3 etc. at the end.

For example, I have 65536 IP addresses at the moment on my network, i know precisely three of them (two DNS servers and a gateway). So, a user's IP under IPv6 might change to:

FE80:0202:B3FF:FE1E:0000:0001 or

FE80:0202:B3FF:FE1E:0000:0002 or

FE80:0202:B3FF:FE1E:0000:0003 or

FE80:0202:B3FF:FE1E:0000:0004...

But I still only need... the first three... which will assign names to the remaining billions as they are required. And the first bit never changes (copy/paste, computers are good at repetitive work) while giving positively BILLIONS of addresses / ranges to everyone along the way.

This is not a burden.

It's not difficult.

It's no more cryptic than IPv4.

And 99.99999% of people will never need see, touch, manipulate or remember that address at all (what is your external IP *NOW* as we speak? Do you know? How did you memorise your 192.168.0.x subnet when you saw it for the very first time and realised you needed it plugged into your network routing?) and the people that do need to? They'll copy/paste from their DHCP config, DNS record, an nslookup, their ISPs techy-only email, etc. etc.

It's ridiculous, in a world reliant upon billions of devices already, to expect a small increase to work for any significant length of time, to expect old hardware to just support that small increase as a matter of course (we're actually worrying about BGP routing tables of only 512k entries which nearly broke the Internet because old routers couldn't hold them!), to redesign ALL new hardware worldwide without future-proofing, or to hinder such efforts because of an address you have to copy/paste from an admin interface once in a blue moon only if you're an IT guy.

Giga-hurts radio: Terrorists build Wi-Fi bombs to dodge cops' cellphone jammers

Lee D Silver badge

If you're that determined, and even vaguely versed in electronics even to the standard of a secondary school electronics clubs, you can replace a GSM module with a satellite radio module (didn't The Reg use one for their paper plane thing? I remember an SMS-like interface, over satellites, that if you have a clear view of the sky received anywhere on the planet), a 433Mhz module, hell an FM radio module, or an SDR (which you can use a £10 DVB-T USB stick for) and a transmitter.

If you're that determined to blow something up, you do it the other way anyway - you do ABSENCE of a signal as the indicator. Text every 10 minutes, no reply before the next text and you activate. Then any jamming actually sets off the bomb. Or have a machine with a 4G module just poll a random Reddit page or whatever and if there isn't a post from a certain user within the time limit, detonate. Almost impossible to trace post-event, all you'll see if accesses to a page even if you can get a full Internet history of a smithereens-SIM... you'll have no idea what it was grepping for among the thousands and thousands of other posts.

Hell, just a GPS receiver with NMEA sentences that - if moved, hits a certain time, or loses triangulation... all kinds of methods.

For £50 I could wire up about ten different ways to activate a circuit board pin, inside something the size of a matchbox.

C'mon, UK networks! Poor sods have 'paid' for their contract phones a few times over... Tell 'em about good deals

Lee D Silver badge

My story is not from the industries in question but:

- My car insurance came up for renewal.

- I got a renewal quote that was more than previous year (no idea why, clear licence, no complexities, lots of NCB, etc.!)

- I went on comparison sites. Got prices literally half of the renewal cost.

- Phoned up my insurance company and tried to negotiate a deal. Quoted the exact same cover, the costs, the companies that were offering cheaper (basically all of them!), etc.

- They literally said that they were unable to knock even £10 off the annual renewal cost. Nothing. Nada.

- So I went with the cheapest from the comparison site. Who use the exact same underwriters, the exact same insurance portal (you can see the design is identical and the URL goes via the same centralise provisioning with online insurance docs etc.), provided more benefits, did it all online, and cost me half the price.

We had a contracted printer services supplier for something like 15 years. They used to lease us the printers, support them, supply toner, etc. throughout the entire site.

- One year the renewal went up nearly 20%.

- We argued. No movement.

- We got quotes from dozens of other services. No movement.

- We showed them exactly what was quoted, which was better kit, for lower prices, on shorter contracts. No movement.

- We offered to take lower services / buy out some machines and put them on the support contract only. No movement.

- We told them "We don't want to go with an unknown. But if you go down this road, we won't renew, we'll just go elsewhere". No movement.

- We went elsewhere. We kept one printer (it was nearly "ours" so it was cheaper to keep it for a few months than pay it off) with them.

- They witnessed the new kit come in. They had to change their kit to be compatible (firmware upgrades for newer Papercut versions etc.) at their expense. The new kit from the other place was basically what these people had been trying to sell us to replace our kit for years. No movement.

- The final lease finished. We asked them to come collect the dregs of it, thus removing the very last reason for them to ever visit site. They took it away. We replaced it with another serviced model from the new company. No movement.

My experience is that even if you make a huge effort, and give them all the chances in the world, most places have absolutely no idea of customer loyalty whatsoever and are run by "computer says no", and that any threats or actions are useless unless - and this is important - they could and would have given you that discount at any time anyway. If you'd literally just asked your mate, who was on a cheaper deal, and queried it, they'd give it to you just for asking.

Generally speaking, if you want my custom, the price is only a single solitary factor in that decision. Knowingly screwing me over comes higher as a factor.

As even Del Boy says, "they may not be cheaper, but at least they smile when they take your money".

If I wanted the cheapest deal, I'd be on TalkTalk broadband, driving a Fiat Panda, and do all my shopping at Lidl. None of those are true, for the vast majority of the population.

Tesla big cheese Elon Musk warns staffers to tighten their belts in bid to cut expenses (again)

Lee D Silver badge

Re: I don't get it...

It's easy to give away the farm and make a loss by supplying a decent product to people below the actual cost of production.

You just have to have a rich idiot bankrolling it... or pulling it out of bankruptcy a couple of times, or maybe a few investors convinced that "one day we'll be selling thousands upon thousands a day"...

No different to "Moviepass" in the US - that was a great deal for anyone who was a consumer. Pay a pittance, then watch as many movies in as many cinemas as you like! Wonderful! Doesn't mean it's a good business idea.

For centuries people have implemented business models that resulted in wonderful value for the consumer, and great product, but at a loss. It's nothing new. The whole dot-com boom was basically the same thing. Here, have everything, for free, because it's "online". Now... whoops... who's paying for that?

Musk bankrolls all his companies, he has legitimately insane amounts of money from his youth, most of his places have neared bankruptcy at least once, he's injected huge amounts of cash into them, because to him it's just a hobby / vision, not a business. Some of them "profit" in the short term (e.g. SpaceX) but most never see their full investment back ever. Even if they did, the investors would want their money back first before prices started to drop for the consumer.

There's a reason that the shareholders in Tesla aren't particularly sad to see him gone from the board, and why they immediately raised prices and/or cut costs once he'd gone.

Tesla, SpaceX, etc. are Musk's toy projects to spend his billions on. That's it. When the funding goes, the business underneath is basically unsustainable in the condition that it's currently run in. He's like the rich kid running a lemonade stand from his dad's Porsche and never making a cent. When he pulls out, and takes his funding with him, all that's left is a vastly unprofitable business. But, sure, he makes good lemonade with only the most expensive and finest Sicilian lemons...

Lee D Silver badge

Re: Servicing

Are we saying that electric cars don't need, say, oil, brake fluid, hermetically-sealed systems, moving parts, etc.

Most of what I've ever taken a car to be serviced for consists of:

- Consumables (brake pads, oil filters, air filters, tyres, aircon, etc. - almost all applicable to electric cars)

- Failures of the car itself (e.g. suspension, exhaust, starter motor, radiator - mostly applicable, or analogous to, parts in electric cars)

When a head gasket goes, or the gearbox fails, you throw the car away if the price is above what you consider acceptable for a non-MOT repair if you're a normal person, or get it serviced under warranty if it's vaguely new or the work still affordable. Few people rip out their engines, change pistons or anything like that in anything approaching a modern car, even people like my father who worked on nothing but car / lorry engines for 40+ years.

On an electric car, no matter how fancy, there are still the same forces involved, still the same failures involved, still the same structures and parts and functions involves, they just might be on a different system. No different to a Wankel vs a 4-stroke, "electric" is just another type of engine. I can replace a dead battery myself, for instance, but I wouldn't be touching a Tesla's battery. The service work has just shifted to proprietary systems under electronic control, serviceable only by Tesla, that's all.