Posts by Lee D

Re: Money talks

It's open to scamming, though.

You go away for a week, you stick a huge interfering electromagnet on the line.

The sync speed drops, they charge you nothing for the week, despite having to pay for all the infrastructure and capacity *IF* you'd suddenly decided to sync at full speed.

You come back off holiday, take the magnet off. You've saved yourself a tenner, they can't detect it, but they had to provide all the backend for it.

Same for each night, of throughout the day based on your desired usage.

Suddenly, they are required to give you 100Mbps backend 24/7 but you're paying precisely only for the a fraction of that, only in peak period, and nothing else.

You could save a lot of money, that wouldn't be visible as you turning it off, not using it, etc. for which they'd normally still be billing you, and it wouldn't be their fault.

Either cost it by megabyte with a "best efforts" line speed (the faster your line speed, the more money they make quicker, so it's in their interest to give you the fastest possible and give priority to HEAVY users!!), or cost it flat-rate per month for a given minimum line-speed.

Anything else is open to abuse, and it's already hard enough to recoup the pricing of telephone and broadband lines to rural locations, etc.

Re: How it made money is therefore a mystery...

And when they have a bill they can't pay, they just need to sit on your money for another week to pay it.

Sounds suspiciously true.

Train the rodents to attack the infrastructure of your enemies.

Nobody would ever suspect...

"A rat brought down Google last night" is much less attributable to malice than cutting a submarine cable in the middle of a war.

Re: I'll never buy Seagate again

It's all anecdotal, and I've been having the same conversation for decades (remember Maxtor / IBM etc.?)

But when I took over my latest workplace they had Seagate in all their servers and all their NAS (some of which were brand-new).

Within a year, I'd had so many failures that I was sick of it and replaced them all with WD. I'd literally never done more rebuilding of RAID arrays in any other job until I took over those Seagate devices (and, yes, they were "enterprise" versions and not just cheap consumer junk). Even the Seagate replacements (I literally order identical model numbers when the server-drives fail, to get an exact match) died at just the same kind of rates.

Meanwhile the clients were all WD Blue and never had a single problem.

Since I replaced them all, I haven't had a single failure in 3 years, and I estimate there are probably 200-250 hard drives on site for various jobs (from CCTV to NAS to RAID to workstation etc.).

I always avoided the arguments of old (back in the IDE / 20Mb drive days) of which manufacturer was better as it was mostly subjective and we only ever had a couple of hard drives anyway. But I have to say that this place has completely destroyed my trust in Seagate drives.

I actually get better reliability out of the cheapest-of-cheapio SSDs that are used 24 hours a day than I did out of enterprise Seagate hard drives.

And, just for reference, I have Samsung / Crucial SSDs in dozens of machines, WD Blue in hundreds of machines, WD Reds in their dozens, plus dozens of IBM-supplied (HGST really?) drives for the "serious kit" on the server / storage end.

Re: ??

Gosh, I wonder how they can do better next year?

Re: Chromecast

Any number of cheap knock-off Chromecast-like things will also do it.

But privacy isn't really an issue if - like mine - the Chromecast is only powered when the HDMI is selected, and is properly - like any client should be - restricted in what it can do (i.e. it can't see any of my local network). And, pretty much, I use it for showing Google Play movies on the screen.

If you're that paranoid, just use DLNA.

Needs broadband is ridiculous in this day and age. We're talking about streaming video, if you don't have broadband, your choices are severely limited anyhow.

Needs additional hardware - yes. That you can replace for £20 a throw rather than £2000.

For reference, I have a Chromecast, a VM box, a Blu-Ray player which can play DLNA, a Android-based satellite box that can do DVB-S for Freesat / Italian satellite. The TV, however, is as dumb as they come with only SCART and HDMI and an RF interface for analogue/Freeview that's not even plugged in.

TVs are display devices. Buying a TV because it runs the app you want or accesses the content you want is stupid, because someday it will stop working like the article. Buy a TV that has a port on it you can put video and audio down, put all your "content" on cheap, replaceable, throwaway boxes that you can upgrade and replace as suits the situation, that don't all need to talk to each other, and that you can add new ones of whenever you like. Even that Android will be out-of-date and unsupported in a couple of years, and then it's just a health hazard sitting on your local network.

Last time I counted, including games consoles, etc. I had about 10 ways to view BBC iPlayer on my TV. Everything from an app on a smartphone pushed over a Samsung proprietary link, to Chromecast from a browser, to the Blu-Ray/Wii having access to it built-in. When one goes "wrong", who cares when you have so many other ways to access, or so many other services to do the same. And my entire setup - with all those boxes and necessary cabling - doesn't come to half what that guy paid for his TV. Probably not even a quarter. And I've had the same setup for nearly 10 years now, and just added to it piecemeal (it is in fact the second Chromecast as the new ones do 5GHz Wifi, which the old ones didn't).

And I bet an Android-based Smart TV is much more privacy-infringing than a ChromeCast, if that's what you're worried about. You have absolutely no clue what that's doing with its data. At least a ChromeCast that you only use for watching ChromeCasted things you have a chance of isolating and seeing what it's doing.

Re: Thus neatly proving ...

Yeah, not something to put on your CV.

"I hold a world record."

"Really, what in?"

"Hallucinating periodically and nodding off, while sitting on a sofa eating finger-food for 48 hours".

Re: Why?

Most new TV's I've seen don't even HAVE Scart any more.

Multiple HDMIs.

And maybe a digital-only RF-in.

Honestly, go look at the back of the display models in John Lewis or wherever. An awful lot of them don't have SCART, analogue or composite at all. And VGA is almost dead too unless you're buying ones made for digital signage.

Also, the composite on Spectrum is a hack anyway. The original Speccy does not output composite without soldering. I think the +2/+3 had R/G/B in some fashion on a DIN plug (which tells you its age!).

Re: Introductory

I'd be concerned if it wasn't for my own computing degree.

At the time, Java was still quite "new" but it was the course language.

As part of the three year degree I had to do a programming course that covered the whole three years.

It literally started at Hello World (fair enough), but at that point I realised that I could do it all without having to be in the lecture at all.

I skipped three years of programming courses by the simple precept of downloading the coursework from the FTP server, knocking something up on the train on the way in, making sure it compiled and run (pretty much, always did once you take account of the missing semicolon or whatever - and I was the only person I knew who logged the dual-boot machines into Linux and ran it in there, the number of people who wrote programs that only worked on one OS was amazing, given it was Java), and then emailing it in.

I don't think once that I struggled to do what was asked, with zero reference to the course material. There was nothing fantastically difficult there, that wasn't covered in literally the only reference material I had - a copy of the O'Reilly book for Java, complete with 1/3rd of the book being nothing but a class / function reference (which is why I bought it, really).

As such, what language you choose to do THAT kind of stuff in, it really doesn't matter.

I still remember sitting in the IT labs as a 1st year, and being consulted by the master students on why their Java minimax implementation for a game of draughts wasn't working. I literally debugged it with one glance, it was that obvious from the code.

I'm not anywhere near an expert programmer, but it was quite worrying that Masters and 3rd year pure-computing students were struggling with that stuff.

Re: Stupid

Letting BT's monopoly steal Virgin's fibre-only customers sounds a great plan!

And then, obviously, what they'll do is sink millions into cabling up a couple of streets that weren't served by either company anyway, just to give people faster speeds at prices that they won't recoup the money from in decades (*).

I think you forget that, apart from new builds that pay for it, the only "cabled" areas in the country were put in by a company that went bankrupt and was bought out by Virgin for a pittance. Everything else is BT / Openreach over cables that have been there for 50 years and/or you have to pay quite a lot to install a new line.

And the rest of the "fibre", even on BT, is not really fibre at all. It's VDSL or DOCSIS 3 with a fibre backend. Unless you have a leased line and paid through the nose to install it, you're not getting fibre anyway (I know - I've bought three of them for workplaces, they aren't cheap but they are real fibre).

(* Do the maths - a leased line install to a cabinet costs on the region of £10k for the install alone.

Connecting that cab to the nearest cab/exchange costs the same because it's all wayleave and digging pipes, not the actual thing you lay inside them.

That cabinet will happily serve the street, if you dig the entire streeet up or run phone lines to every house, costing a lot more than £10k.

And you want to pay, what, even £50 a month? How much of that goes to the people who put in the infrastructure rather than the ISP? Less than a quarter? That's going to take them something like 2000-5000 monthly payments alone just to get their install money back, not counting ongoing costs and actually providing the service to you. How many houses in your street? 50? That means either every house for 10 years, guaranteed, or most of the houses for, say, 20 years. JUST TO RECOUP INSTALL COSTS.

There's a reason nobody wants to pay to connect up people, especially in rural areas where you run 10s of km of cable to service a handful of people. It just doesn't scale without massive subsidies)

Re: Maybe I'm not as smart as these card tech guys...

"That's not how the card, or mobile phones work"

I'll think you'll find that he means he can get an image of your fingerprint quite easily.

And your phone fingerprint sensor can be fooled by an sufficiently good image of a fingerprint, printed onto certain surfaces. You don't even have to get very technical.

Every smartphone fingerprint sensor (and this card sensor) on the market can be fooled with nothing more than a picture of the fingerprint smudge you left on the card as you last took it out of your wallet. It just depends how many times you want to try it to refine your technique.

Last year, someone pulled the fingerprint of a German politician from a photograph of them raising a wine glass. All the "temperature/heat/light/pulse/etc." sensors in the world can't do much that isn't easily fooled, and the actual "fingerprint ID" process is still - to this day - finding the edges on a high-contrast B&W image of the fingerprint in question as it lays flat on a surface. Whether the sensor is swipe, scan, optical, or whatever.

I have a bunch of Gemalto etc. fingerprint readers in my junk box if you'd like to play. They almost all have open-source software that presents the image as a B&W TIFF from the sensors to something that edge-detects and then hashes / stores the result. How they store it is irrelevant if you can present the same image to the sensor and the sensor then hashes that to the same hash as a real fingerprint would hash. The hardware doesn't do anything fancy, but a bit of image processing and maybe a particular wavelength of light / check for colour variation for pulse (and that's an "advanced" model).

There's a reason they're all in my junk box despite being "state-of-the-art" for banking security at one point or another.

Re: Where's my Ferrari, that's what I want to know...

They come to whatever the authors demand.

Even at a penny each, that's one pound per console and they haven't paid that.

But if the authors still say it's 10p or a pound for their game, they either have to pay THAT or cannot distribute.

Theft is not a case of value of the item.

Re: My exact thought

Modern android pre-compiles the app to a closer-to-native format. The Dalvik / Java thing is basically gone, and it's all JIT and ahead-of-time compilation, with on-the-fly profiling and optimisation.

I would put this down to being more about "we only have one phone with fixed hardware, so you can massively optimise all your apps towards that" versus "anything can run Android, your app has to check for everything, so your optimisations will never be perfect on all devices". There's also an Internet speed factor here, too, Super Mario Run and those other apps do a lot of network activity on startup (try it without data and wifi turned off - it just bugs out) that can slow such tests down. And I'm sure I can find a hundred tests that the iPhone "fails" on just the same.

I guess that by the time the S8 gets to the point that I would bother to touch it, I won't even notice anyway. The iPhone will basically just never come down in price.

To be honest, raw performance isn't what I buy smartphones for, though.

I hate pissing contests over raw numbers when, actually, things like "I'd like to plug headphones in", "can I change the battery", "can I change the SIM", "can I buy an approved charger for a non-ridiculous price", "can I expand the onboard storage", "do I *need* an account to make it work", etc. are much bigger questions for a smartphone to my mind.

Sadly, Samsung et al are following the stupid Apple answers for some of those questions even now. I stuck on the S5 Mini - fast enough (and hence, can't really see why I'd need it significantly faster and I do all kinds on my smartphone), stable enough, cheap enough, sensible enough, accessorisable enough (though USB host would have been nice), small enough and big enough, and lasts long enough.

And the H in this instance is for Hydrant?

Re: Webs!!!

Giant alien spiders are no joke!

Have you never played FTL?

Refuse anything with a CLI (option on most smartphones)

Install the apps that lookup the number as it's ringing so you know if it's got a high chance of being spam or not, they can even auto-refuse the call if you want.

Also, make the default ringtone "silent", and the ringtones of contacts (stick them all in a contact group, say) your normal ringtone.

Draytek.

They have very featureful routers, lots of regular updates, ADSL / VDSL / Ethernet and 3G/4G failover on the same device (e.g. 2860 / 2870's). They let you do proper VLAN, QoS, RADIUS, web filter and all kinds of things if you really want to. They have very good wireless. They can mesh nicely with other APs and routers in the same range (centralised management, etc.). They can handle IPv6 (shame most ISP's can't). They rarely do stupid things. They have the backend processing to keep up with anything you throw at them.

They are more expensive, but you get a lot more back.

TP-Link are good and cheap for what they are. Just don't trust that kind of stuff as your front end to the Internet.

*Virgin Superhub 2 in modem mode -> Draytek Vigor 2860Vn+. for reference.

Re: Pleas for extra

I would expect any administrative cost to be lost in the noise of the operations of the business. Otherwise someone is doing something very wrong.

Curing still needs the flies to be removed.

You could smoke it, but if you have a campfire, you could cook it anyway.

Smoking/curing/cooking a 65kg human is going to take you longer than it takes to go off, if you're on your own with no resources.

Re: "coffee please"

"I'll have a tea."

Re: They walk amongst us

Should we be funding any claimed "professional" suggesting treatments that are known to be no better than placebo, but cost more than any standardised placebo treatment?

My answer is no.

Placebo has a powerful effect. And can be given to you by a sugar pill. We shouldn't be PAYING PEOPLE to be nothing more than unregulated placebo-peddlers for patients without clear notifications of such.

(P.S. Note carefully the wording, as professionals WILL knowingly give placebos in controlled trials, for instance - and they tell you it's a possibility [which, interestingly enough, doesn't affect the placebo effect!], and they are doing it to compare against the real research going on in the same project).

No matter what study you cite, title you get, qualifications you have, or claim to have, all "alternative medicine" treatment is independently verified to be no better than placebo, time and time again - measured by scientists and independent medical researchers on your behalf. Just because a handful of GP's believe it, or Prince Charles, or even someone with ten medical doctorates, does NOT make it truth. Hell, they can even PRESCRIBE it, it doesn't mean it works, or that there isn't something much simpler that would work much better. I can point you to any number of people qualified in such areas who belief all kinds of nonsense.

Pay for your own sugar pills, and think me wrong as much as you like. Because when something "works", it gets refined, experimented on, analysed, enhanced, commercialised, tested and becomes... medicine. Aspirin came about from people chewing tree bark for 2000 years. The Mongols put mouldy bread on their saddle-sores. But until you can prove it works, it means nothing. And then when we prove it works, you get a proper medicine not some home remedy (aspirin, penicillin and it's thousands of derivatives, etc.).

If you have something that "works" and medical researchers aren't trying to buy the rights from you, it's a lie. Because they would happily take it, isolate the part that works, synthesise it, and sell you a pill of it in a second. The fact that they don't? It means it doesn't work.

Re: Dosh?

"Which is why Apple's kit compares favourably with the competition."

Does it? The latets Apple Macbooks start at $1,299.00, for a 1.1GHz ("up to 2.2GHz") mobile dual-core.

The Lenovo comes with 2.6GHz proper i5. The 3DMark scores alone beat the Macbook. For the sake of $30, getting beat in all the CPU and graphics benchmarks, and with a smaller screen than the Lenovo, isn't a good showing.

This is by far not the cheapest PC laptop on the market, either, whereas a Macbook is the ONLY Apple laptop on the market (legally).

You're kidding right?

GSM, Wifi and Bluetooth controlled relays are everywhere. You can pick them up in Maplin's if you want.

And, no, we're not just talking solder-together kits but proper pre-built and support modules where you just power the thing, and then put the "button" wires through the relay. Job done.

Hell, for £10 on Amazon, I bought a GSM-controlled device that monitors GPS (even when the car is off), texts me if the car moves, tracks the car in real-time, alerts on overspeed, etc. as well as allows me to listen in to a microphone in the car, text the car and cut off the fuel pump, etc. by relay (included in the box!). It was literally cheap Amazon junk that can do exactly what you're asking.

I also have an house alarm that does the same (text it to activate a relay as a side-effect of it's 8-zone PIR / reed switch, battery backed house alarm functionality).

There are also bluetooth sensors that can activate when a particular bluetooth device comes in range / leaves range - sounds ideal for that purpose, no messing about with apps, just driving up to the garage with your phone switched on and it opens up.

Hell, my garden gate is maglock with an RFID reader and my girlfriend and I both have RFID tags. She uses it from her bike (push bike, mind) so that she can cycle up, fob the reader without having to get off, then ride into the alleyway directly. It cost me £10 on Amazon too, and came with a bag of fobs. If she loses the fob (like if your phone app doesn't work!), she just types in a code instead.

There's no need for poking holes in firewalls or other junk if you buy the right stuff. Which is often the cheapest junk because it has no subscription or dependence on a massive server infrastructure to do a simple job.

It's been 45 years since we bothered to put humans outside the planet's immediate influence. At any point in the intervening time we could have done this. At any point WHILE we were putting people there we could have done this. We didn't. In fact, we didn't do an awful lot, really.

I see no reason to think than in a few years we can turn that around without a serious ramping up of competition/co-operation worldwide, determined to do something practical and useful and with some kind of return. Even SpaceX isn't in the same kinds of order of magnitude (and from what I just Googled it's barely profitable at all, and that was before the latest setbacks and with billions in investment?).

We're not doing it for science. We're not doing it to save humanity. We're not doing it because people want to pay the prices to live up there. We're not doing it to get people to Mars. We're not doing it to get one over the Russians or say we were the first. All of these are reasons (not necessarily good ones) to do such things, but that's not what it appears we are doing it for. As such, it's hard to see WHY we're doing it, and therefore why it would happen.

We've been able to do this since the 1960's. That we haven't says a lot. And the distances and scales involved are literally hilarious. The Moon is 0.1% of the distance away that Mars is, for example. And we're talking about getting ONE thing to just past the Moon as some major project for the next generation...

Re: Is it a bug?

If you have to insert an explicit rule, it's not AI. It's a human-written heuristic.

If the machine can't learn on its own, it's not AI either. It's a human-controlled heuristic.

If you have to spend your life telling it "Oh, and look out for this explicit thing that you get wrong", then you may as well just write a list of rules.

And the exact problem with these "deep learning" machine algorithm things is that you can't just say "Oh, take this into account", because they aren't written that way, they've learn from the data.

No, you have to go back, create test cases for every imaginable scenario, spend years training it on all of them and hope it picks up on what it was doing wrong. And then someone comes up with, say, picture-in-picture which confuses it again. Back to square one retraining on that too.

So you can't use it for, say... crowd-based facial recognition (as is often advertised as a use case for such things), or self-driving vehicle cameras, because it could flag ANYTHING at any point just by being sufficiently distracted - even with NO knowledge of its underlying training or algorithms. And you can't train it on every possible scenario well enough that someone trying to catch it out can't just make it flip.

Imagine telling even a 2-year-old that they're going to need to win the toy by getting the video right. And you show them a video with a thousand frames of tigers, and one frame of an Audi. Would you really ever expect them to press "Car" instead of "Animal"? This system is no better trained than a 2-year-old human, in that case, who can do a lot more besides.