Re: BOFH phone
> the phone charges up a 10kV 10uF cap and dumps it onto the external connections
They did that in "Tomorrow Never Dies".
Vic.
5860 publicly visible posts • joined 7 Dec 2007
> Or, how about this for a DIY low pressure switch ?
Here's a variant of that idea that might not need so much setup:
Put a condom in a plastic tube. Put a microswitch (or similar) at the end.
Seal the end of the condom.
As the pressure drops, the condom grows. By correctly calibrating the amount of gas in the condom / length of the tube, a firing pressure will be attained.
Make sure to debounce the switch heavily...
Vic.
> A friend had his insurance claim refused (third party fire & theft) after his car was stolen
Yeah, you'll notice I said that insurance companies are prohibited from avoiding *third-party* claims. I said nothing about theft.
Your friend made a claim against his own insurance because of theft. This is not a third-party claim.
Vic.
> at least give them a black-line demarcation that makes a moderate degree of sense.
And that's exactly the inability to understand speed I was talking about. No such black-line demarcation can possibly make sense, beause hazard is inherently a probability function, relying on many other variables as well as speed.
By convincing the population that 29mph === very safe and 31mph === very dangerous, we are making a huge mistake. The amount of sleep the driver had the night before is likely to be far more significant than a few mph. Ditto the state of his personal relationships, his job, ...
What we need to do is to get people to take responsibility for their own actions, not give them trite little "rules" which bear little or no relation to reality.
> perhaps you also might rather I limited my speed to mitigate any potential damage?
No. I'd rather you be regularly assessed for competence to drive, and I'd rather you be held personally responsible for any damage you cause.
I disagree with mandatory sentencing, but I would like to see an *expectation* that any driver hitting a pedestrian or cyclist will do jail time.
Vic.
> most insurance companies will use No MOT as a get out on their policies
Insurance companies are legally prohibited from using such issues to refuse third-party claims. They must pay if the claim is proven.
They will, however, attempt to recover such losses from the driver.
Vic.
> Do accidents happen because of behaviour that can be reasonably
> assumed or expected to occur
Almost always, yes.
The problem is that so many road users simply don't put enough effort into working out what can be expected to occur.
Once again, I'm going to plug "Roadcraft". If you haven't read it yet - you should. It's cheap, and it *will* save you at least one accident. And make you feel very inadequate[1].
Vic.
[1] Strangely, it also made me much faster. Although I no longer drive or ride at the same top speed as I used to, the observational skills I learnt from the course mean that my average speed is quite a bit higher.
> Speed ... is easily understood
Actually, I don't think that's true.
By far the majority of the people I meet do not understand a square law relationship. They cannot see that a car at 40mph has approximately twice the energy of a car at 30mph.
But more importanbtly, IMO, a vast number of people cannot see that safety is not a black-line demarcation; it is categorically *not* true that $limit-$delta is safe, whereas $limit+$delta is dangerous.
So although speed does need to be considered when deciding if someone is driving safely, there are many other factors[1] which seem to be overlooked once an observer has decided to focus on speed...
Vic.
[1] Many of these are probably far more important than speed when it comes to the probability of having an accident[2]. Tiredness is the most commonly-ignored factor IMO - and it is arguably as dangerous as alcohol.
[2] Note that I am considering the chances of having an accident, not the aftermath of having had one. It is obviously the case that more energy is likely to lead to more damage. But I would rather people looked at ways to avoid accidents, rather than just mitigate the damage. I would much rather be missed by a car doing 40 than hit by one doing 30.
> When was the last time your insurance even held steady, let alone went down? I
Mine went down this week (renewal was yesterday).
It helps that I hadn't had a car for some years, so last year, I had no NCD. This year I have a year.
My insurer from last year still put the premium up, though, and didn't bother sending me a renewal notice - just told me over the phone that they were about to charge my card anyway., The bastards shouldn't even have had the number. Grrr.
So being pissed off with that company, I went looking. One quote that beat my renewal by £80 was - exactly the same company. Bastards.
I ended up going with someone else, for a lower premium, and a lower excess. I saved over £100, and got a better deal all round...
Vic.
> the only time the EDS/HP board listen to an idea is when it comes
> from someone they are paying to come up with ideas
This is entirely normal. There is a class of people who can only value things - items or ideas - according to the price tag. For reasons I cannot explain, an enormous amount of management seems to consist of these people.
Years ago, I had a real problem getting people to take my ideas seriously. I had many opportunities to do the "I told you so" thing later[1], but that didn't stop the project going pear-shaped in the meantime.
My solution was simple: I doubled my price. And all of a sudden, my suggestions were adopted.,
Vic.
[1] I learnt very early on that actually *taking* such opportunities is a bad idea.
> it is intentional
Yes.
When the traffic industry says "managed", don't take that to mean "does something I consider useful".
Traffic management is about affecting driver behaviour. The actual intent might be different to what you or I would consider sane. That intent might or might not actually do some good somewhere.
Vic.
> Southampton, a city with such bad traffic management
Much of the traffic industry sees Southampton's management as *good*.
Quite a bit of the industry is around Soutrhampton/Fareham, so we get trials...
But the Bitterne Road junction is world-famous - traffic engineers around the globe know of it, even if they don't know which country Bitterne is in. Shame the whole design[1] went pear-shaped the minute they deregulated the buses...
Vic.
[1] The plan was to stop people buying the cheaper housing on the East side of the river, then driving in and out of the city every day,. So there is a complex set of junctions between Athelstan Road/Bullar Road and Bursledon Road that make driving private cars harder if you venture off the main road, but prioritises buses (top bit of Bitterne Road, Garfield Road). The idea was to make public transport *much* more viable than private. And then they deregulated the buses, so it isn't.
> nice habit of turning red on the main road, to allow side traffic to join
You want to try living in France.
Some parts of the road network still use the "priorite a droite" law, which means that slow traffic joining from minor roads has right of way over traffic already on the main road.
Other parts of the road network do not use this rule.
Vic.
> taping a couple of neodymium magnets to the frame or pedals.
I wouldn't bother with that.
The sensor loop is a parallel tuned circuit - you've got a big coil of wire in the road, and a capacitor across it. The whole is driven by an oscillator - you've got a high-Q filter, so when in tune, there is negligible current flowing.
When something metal passes over the loop, the circuit loses tune, so current starts flowing. This is a very sensitive detector.
Over time, the components and road environment age, and you'll get some drift in the tuning. I did read about an idea to allow the loop driver to auto-tune, but I've no idea if that ever made it into production - it's a long time since I've done anything in this field.
Vic.
> What this guy is proposing is a city-wide model
Existing traffic systems have been using a city-wide model for decades.
He *might* have come up with a clever new algorithm. He probably hasn't - most such "innvoations" have a fundamental flaw that has been overcome many times[1]. But what he has *not* done is to create something entirely new - a city-wide, flow-modelled control system with feedback loops of varying sizes is in existence in most major cities around the world, and has been for a very long time.
Vic.
[1] My favourite is the single signal head: every year, some educator uses his influence to get in front of the bods at one of the big traffic companies to show off his latest protege's idea. This idea will inevitably be a signal head with a single lens, and a multi-coloured illuminator behind it. This used to be separate bulbs and a bit of optics, now it's assuredly LED. He will go through all the supposed benefits of simplified maintenance etc. Eventually, one of the audience will crack, and utter that fateful observation "red-green colour blind"...
> they had those black rubber bars across the road
The black rubber tubes are actually a census point. They are connected to a box chained to a pole somewhere close, and are used for gathering information about road use for planning purposes - they can determine class of vehicle, speed, wheelbase, etc.
But you're right that VA junctions are old news - only the sensors are predominantly buried into the road.
Vic.
> Using sensors to measure variables like the amount of traffic already in
> a road section, how quickly it is moving, and how long to the next change
> of lights, Dr Helbing’s approach is designed to respond to unexpected events.
Traffic systems have been doing that for decades.
If you look at an approach to a set of lights, there will generally be two sensors embedded in the road - either a diamond or a rhombus shape, depending on which manufacturer put them in. There may also be IR sensors on top of the signal heads.
Each junction has its own local controller, and is also networked to all the others in the area.
I've not been near the traffic industry for about 15 years, but all this was established technology when I was doing it...
Vic.
> If that is the case, why did they teach the Green Cross Code?
Because being right doesn't help much when you're dead.
> pedestrians only have right of way when crossing a side street and you are
> turning into said side street
Rule 206 of the Highway Code says differently. RTA probably has a section to make that enforceable, but I can't be arsed to check right now.
Vic.
> Do a little research
Do your own research.
Just because someone built a monowheel does not mean that they are useful. They fundamentally cannot brake hard., as just a few minutes' reading or thinking would dsemonstrate.
It's possible to build a 1200cc unicycle. That doesn't make it a practical vehicle.
Vic.
> what would happen if VW did throw some of its GDP-of-a-small-country
> R&D budget at creating a hovercar
Someone would crash it.
The trouble with having no direct contact with the ground is that all acceleration needs to be done by way of vectored thrust - directed fans and the like. Unless your name is Apollo and you have a big "USA" decal on the side, that thrust tends to be comparatively low compared to the mass of the vehicle. So you get comparatively small acceleration.
That means you have to look much further ahead, and plan your manoeuvres much more carefully. Which won't happen, and so drivers will go careering into stationary objects. Frequently.
There are a few places around where you can go and drive hovercraft. It's excellent fun - but you'll come away with an unshakeable belief that such vehicles are not fit for the public road...
Vic.
> That is silly
Yes. That was the purpose of my post - to demonstrate that it was a crazy position and a good part of the reason that Sony's venture into 1394 went nowhere. Well done for getting that.
> Best to ignore that and carry on
It's much easier to ignore it when you're not in danger of getting fired for doing so.
> Same socket anyway
That would be because it is the same interface, as I mentioned in my initial post...
Vic.
> They insist on introducing proprietary formats
Actually, much of the time, the formats are not proprietary. It;s just the name...
Years ago, there was an article on ZDnet (IIRC). The author was making a fuss about Sony having introduced yet another compression format called i.Link.
i.Link wasn't a compressor. It was a data link. I was permitted by corprat to write a reply to the article explaining that - but I was *not* permitted, under penalty of $unspecified, even to hint at the fact that this magical i.Link was simply 1394...
Nobody bought i.Link because they were scared of being locked into a proprietary standard.
Vic.
> My last camera was in the bad CCD period, but deck was playing up
I've got a camera in my workshop. A customer left it with me, as it was BERed by Sony.
It's a Mavica camera with quite a nice lens on it. It writes images to 3-in CDs.
Well, this camera had stopped writing to the CD. It kept failing, which was of no use to anyone. My customer took it to the local camera shop, who sent it back to Sony for testing.
Sony said that they tested it thoroughly, and the drive was knackered. A new drive was £600. I shit you not. The camera was written off, and a replacement (not a Sony) purchased.
I put a decent-quality CD in the drive, and it's yet to fail on anything...
Vic.
> And those 2 things are really not that bad.
Yeah, they are...
> Sony make great products
Sony *used to* make great products. When I joined, the kudos was simply fantastic. I was proud to be associated with the Sony quality.
During my time there, the pennies were ever-more pinched. In the end, the kit we were shipping was at best just the same as everyone else's, but with a higher price tag. At worst, there were serious difficulties with what went to market.
Vic.
I downloaded the Android app mentioned in the article.
It requires a full-time Internet connection. Which I'm unlikely to have when outside[1].
Without a connection, it just throws up an error box until you reconnect.
So I uninstalled it.
Vic.
[1] I'm on a very cheap PAYG tarriff, which means having a net connection costs money.
> SSL encrypted comms - which are widespread even for home users using POP3 / SMTP
Encrypted SMTP is currently quite weak - although a key exchange does occur, it is rarely verified against a root certificate. Thus a MITM attack could supply its own key to the sender, and the data is effectively in the clear for that attacker.
I expect to see much more TLS verification in the coming years.
Vic.
> anyone thinks that old Liz actually has a clue about anything
She's far more clued-up than you might imagine.
She was a driver/mechanic during the war, and she was introduced to email before most of the rest of us.
That she talks such utter bollocks in the "Queen's Speech" is down to it being written by the government, not by the monarch.
Vic.
> I don't see why having an IPv6 enabled router with built-in
> firewall has to be more difficult/expensive than using NAT
Because people have to think about having all their machines exposed.
Many people have gotten used to the idea of one device (the ADSL modem/router) being the only thing directly exposed, and the core of their security is that their PCs are not routable from the Intertubes unless they have explicitly made them so.
Now it could be argued -and I probably would - that this was always the wrong way to be setting up a network. But tens of thousands of organisations across the country have exactly this model. It's where we're starting from.
To move to IPv6 as it currently stands means changing that model so that the sysad - if there is such a person - needs to think through each machine being accessible, and thus the firewall needs to be set up rather more carefully. Many organisations simply won't do that - and IPv6 will gain itself a reputation as a security problem.
What IPv6 needs is to accept NAT as a reality - just because IPv6 doesn't *need* NAT in the way IPv4 does, that's no reason to try to prevent it happening. That gives us the best of both worlds - those that want massive routability have got it, those that want to hide their internal machines have also got what they want. And NAT for IPv6 already exists - see nfnat66. Last time I looked, the IETF was less than happy[1] with NAT on IPv6 :-(
Vic.
[1] There has been some movement; NAT is no longer a dirty word, and the IETF has published RFC6296 for stateless NAT. This won't help the sort of user I've described above. nfnat66 additionally supports stateful NAT, which is what will have to happen before the general public accepts IPv6. But the SF page for nfnat66 shows an overwhelming 1182 downloads, so I don't expect to see this going mainstream for a little while yet.