Posts by James Cullingham 50 publicly visible posts • joined 20 Oct 2009
What a day, when amanfrommars made more sense than most other commentards.
It is truly amazing how the merest mention of generative AI brings out the pitchfork-wielding mob. It's just a tool, which like every other tool can be used well or can be used badly.
This problem has little or nothing to do with the technology.
We all, both people and not-people, learn from our experience of the world.
If the information we derive is relevant, accurate, balanced, insightful, etc then that is what we have to build on. And from that we can learn and develop. Positive feedback.
But the more that source is polluted with irrelevance, accuracy, bias, closed-thinking, the more we regress. Negative feedback.
On the stretch of road where the bridge was, the StreetView images are dated October 2012.
Approaching from one end of the road the images are dated May 2019. From the last of these you can see the collapsed bridge, and you can also see that there is a fairly flimsy barricade. Presumably the camera car could see that the road was impassable, as it went no further. That would have been a good time to update the map.
Approaching from the other end the images are dated May 2023, and now concrete blocks are visible, mostly blocking access to the former bridge from both sides. Again, the camera car would clearly have been very aware that the road was impassable.
So I think Google could and should have updated their records. But their failure should never have resulted in anything more than inconvenience.
As of 2019, at least, though, the signage and blocking of the road was pitiful.
But, as far back as 2019, and still in 2023, so clearly also at the time of the accident, there were (small) trees growing up in the space where the bridge had been. A competent driver, proceeding sensibly down an unfamiliar single-track road, should have been going slowly enough that seeing a tree in front of them should have been enough to cause them to stop safely.
I would pin a substantial amount of the blame on the blame on the driver and the rest on whoever is responsible for maintaining the road.
Following the link to the page on that Driving Test website, we can see that it correctly gives the thinking & stopping distances from the Highway Code. At 30mph, this gives 30' (9m) thinking distance and 45' (14m) braking distance. Total is therefore 75' (23m).
It also states: "Broadly speaking, stopping distances double when you’ve got wet conditions. On ice it’s worse: they’re ten times greater. So if, driving on a normal sunny day at 30mph, it takes you 23 metres to stop (that includes the time it takes you to see the hazard, process it, and react), then in icy conditions it could take you as much as 230 metres to stop."
Seriously? My thinking time is also going to be multiplied by 10?
No, my thinking distance will remain at 1 double-decker bus, not 10
My 'favourite' feature related to the storage of numeric constants that formed part of a line of BASIC. As background, you have to bear in mind that the machine was very storage constrained, but also very performance-constrained, especially when it came to FP arithmetic.
Say, for example, you had a line
10 LET A=1
10, as the line number, was stored in 2 bytes
LET was a keyword, so was stored as a single byte token
A and = obviously took 1 byte each
But then came the numeric constant 1. Due to the performance hit of parsing that into an 5 byte FP representation (1 byte exponent, 4 byte mantissa), this was done as the line was entered, so what actually ended up in memory was the code for the character 1, followed by the marker byte 0x7E, followed by 5 more bytes of data.
Horror! One character took 7 bytes!
It was actually more efficient to write instead
10 LET A=PI/PI
because PI was, again, a single byte tokenised keyword, so now you had 3 bytes instead of 7 - a substantial saving and one that would add up over the length of a full program.
Similarly 3 was INT(PI), and arbitrary constants were, IIRC e.g. INT("12345") (still saving 3 bytes).
You got the performance back by writing every-increasing amounts in machine code (process: write it out, then hand-assemble, then create a hex entry utility, then type in the entire codebase in hex)
And you saved bytes in your machine code by referring to your disassembly of the entire ROM (published as a book - https://k1.spdns.de/Vintage/Sinclair/80/Sinclair%20ZX81/ROMs/zx81%20version%202%20%27improved%27%20rom%20disassembly%20%28Logan,%20O%27Hara%29.html) to see if you could abything from a full routine down to a few bytes here and there.
Happy days!
Uber is acting as a marketplace - it brings service providers and potential customers together, and takes a fee for the privilege. As part of this, it may or may not choose to put measures in place to ensure that a satisfactory standard of service is being offered, as that protects its own reputation and therefore business.
It seems therefore that it is entirely valid to compare it to, say, eBay, which does exactly the same thing. If I start a business selling goods on eBay, and earn enough to make a living, that emphatically does not make me an employee.
The only difference is that in the case of eBay, payment is earned for selling items, whereas in the case of Uber, an argument could be offered that this is more of a personal services scenario. But it isn't - a driver is not being paid for their time per se, but for the fact that they have completed the service that was required, i.e. to move people from place to place. If they drive more slowly, they will earn less. This is not how employment works.
Please note that I makes no difference to me one way or the other - it just seems so clear cut that I genuinely don't understand how the GMB feel that they have any legal basis to their case.
Please correct me if any of the following are wrong.
(1) Uber drivers are solely responsible for ensuring they have a car. So they have the fixed costs of buying or leasing a vehicle, as well as paying for insurance, VED, maintenance/repairs, etc, plus the variable costs of petrol, congestion charge/tools etc.
(2) Uber drivers can choose when to work, and for how long.
(3) Drivers are rated 1 - 5 stars by customers, but drivers can also rate customers - implying that they are free not to accept a fare if, for example, a prospective customer has received poor ratings.
If the above are in fact correct, then it seems absolutely clear that the drivers must be self-employed, and must accept both the benefits and the risks that come with that.
In particular, how can (1) and (2), taken together, be consistent with the idea of a minimum wage. If a driver, in a particular month, has fixed costs in relation to their vehicle of, say, £300, and they only choose to work for 2 hours, should Uber pay them £314 just to ensure that their 'wages', aka profit, for the month corresponds to £7/hour for the hours worked. Why would this be fair when compared to the case of the driver who works 234 hours?
Use case for manual disengagement is a slow downhill start.
I found this shortly after getting a Renault Laguna with automatic handbrake. I parked in a short drive that sloped steeply uphill away from the road, such that to start I had to reverse into the road, first easing back to check visibility. At the time I had not spotted that the brake could be disengaged by applying the footbrake then operating the manual switch, so achieved this by selecting reverse, applying sufficient torque to get the parking brake to disengage automatically, then getting very quickly onto the footbrake before shooting out into the road. Much nicer, obviously, to disengage manually then roll back gently.
Even if they had looked up the answer to this one it is easy to be led astray. If, for instance, you visit mooreslaw.org — a site which certainly sounds authoritative — and the home page is headed "Moore's Law, or How overall processing power for computers will double every two years".
The text then states:
"Moore’s Law is a computing term which originated around 1970; the simplified version of this law states that processor speeds, or overall processing power for computers will double every two years. A quick check among technicians in different computer companies shows that the term is not very popular but the rule is still accepted.
"To break down the law even further, it specifically stated that the number of transistors on an affordable CPU would double every two years (which is essentially the same thing that was stated before) but ‘more transistors’ is more accurate."
So if you don't read very far you will be considerably misled. No wonder people are confused.
At least Wikipedia is more helpful.
He uses the word podule to refer to his translating device in this skit from the his live show in the late 80s. Of course, the word lends itself to his delivery.
http://www.firstpost.com/topic/person/rowan-atkinson-zak-the-alien-rowan-atkinson-video-r0LoLdjGy9o-3117-1.html
If you haven't seen it, it is highly recommended
How can anybody believe that this is useful. It can't possibly be in the interest of consumers having to remember different gestures to achieve the same result on different manufacturers' kit.
Say I have a certain brand of TV, and I've carefully learnt all the gestures required to control it. Maybe manufacturers will now see that I am locked in as I am less likely to want to relearn these skills if I switch to a different brand of TV. But the truth is that I (and vast numbers of people with less interest in learning arcane hand movements) will just not bother using these features, and will probably be more inclined to choose a brand that doesn't waste our time with them. All in all an own goal.
Imagine cars, if in the early days someone had patented 'using little sticks on the control column to operate the indicators' or 'moving the gear lever in an H pattern to move through a sequence of gears'. Nobody would have a clue what they were doing. Standardisation helps everybody, and rushing to patent obvious things just blocks the development of usable standards.
Why not split the charging, like with phone calls? Your monthly charge would then have two components.
First, the 'line rental' equivalent, i.e. your payment for the ISP making a connection available. It would be not unreasonable for this to be based on speed. When I signed up with BT I was told what speed to expect (6Mb vs advertised 8Mb) and this has proved a good estimate. Other posted above have clearly had similar experiences. So, this part could be based on a reasonable (and periodically reviewable in case of significant changes in infrastructure, contention levels etc) assessment of expected speed.
Second would be a usage-based fee, presumably charged in arrears, representing a fair share of the backhaul capcaity etc.
The balance between the two would be an additional area of differentation between ISPs. No doubt in practice most would offer a bundle including a certain amount of data, as now, but for fair pricing there would ideally be much less bundling.
After all, who should pay more: the rural-dweller with a 1Mb connection that keeps it busy 24/7 (monthly throughput 300GB+ unless my maths are out) or the urbanite with a 200Mb connection whose only use is checking email and a bit of shopping. I don't know the answer to that, as they're paying for different things: one is primarily paying for the exchange to premises connection and the other primarily for backhaul.
As has been pointed out, speed limits would be respected more if they showed more evidence of having been set carefully and with the aim of increasing road safety rather than, for example, satisfying some political objective or of saving effort for the person setting them.
Here is a fine example of illogic: http://maps.google.co.uk/?ll=52.997323,-1.524943
The main road - the B5023 - has a 50mph limit, as do all the main roads within a radius of several miles. I'm not necessarily disagreeing with this, incidentally: it is a lovely road but does have a number of field entrances and agricultural traffic, so perhaps it makes sense (at least for some stretches). The limit was reduced a few years ago (5ish).
But look at the short stretch of road that resembles a layby. It's about 150m long and serves three houses and a pub. No traffic will use that road unless to access one of those four destinations. Yet the limit there is 60mph (you can confirm that with Streetview)! Admittedly I'm not sure how fast a car would need to be to reach that speed in the space, but you may still ask why. Presumably it made the paperwork easier (only one road for which the limit was being changed), although it did mean that more signs had to be installed.
While we're at it, look at side the road opposite. Again, this carefully tells drivers that they may now drive at 60mph. But Streetview will also confirm that (a) this is a much smaller road than the main road, with a poorer surface; and (b) a few feet after the national speed limit sign are two warning signs indicating that the road narrows and is bumpy, because there are bridges over the river and railway.
Some limits should be higher. Some should be lower. But either way they should be properly thought out and relevant, so that 'my judgement was that I could safely drive above the limit' ceases to become a valid statement. Then, and only then, we can talk about proper enforcement (e.g. using GPS).
Final point: much though I hate them, I respect average speed cameras much more than spot ones. You still get numpties braking for them, but at least they mean that if, for instance, it becomes necessary to exceed the speed limit temporarily (e.g. because it is the quickest, and therefore safest, way of overtaking a slow-moving vehicle) then it can be done without necessarily falling foul of the machine.
So, which of these is it that raises matters above the vulgar brawl level? I thought it was artillery (blame Civilisation 4!) and Google supports this (214k hits for 'artillery "what would otherwise be a vulgar brawl"' vs 22k for 'cavalry "what would otherwise be a vulgar brawl"'. Does anyone happen to know the real story?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2026
