Posts by Loyal Commenter 5761 publicly visible posts • joined 20 Jul 2010
Quite frankly, if they're basing their principles on something they read in Knuth, they are probably doing it right. The fact that those principles are half a century old and counting only highlights the foolhardiness of those who choose to ignore them, the ignorance of those who never learned them, and the arrogance of those who never taught them.
I think the most valuable thing here is that the microservice architecture enforces encapsulation. Monolithic software doesn't, unless you do so by design. As someone who has to maintain software that began its life in the '80s in a language that doesn't even really have a concept of encapsulation, I can't understate how important encapsulation is for maintainability.
It also sounds like it makes scalability and resilience easier. Good to see also that Monzo recognise the adage that premature optimisation is the root of all evil.
It's not beyond the normal range of commission for auction houses. It's why auctioneers are often very well dressed...
In the UK, you'll pay 20% VAT on top of that as well. Things often look a lot less like bargains once you've added 45% on top of the price you just bid...
Unfortunately punitive damages are not permitted in such cases - they very rarely are in general, at least under English law, and class action (where permitted, but not in the UK) generally leads to compensation of fourpence per claimant. In reality, even the GDPR maximum fines are little more than an assumed cost of doing business to the large internet service providers.
The GDPR maximum fines are 2% or 4% of global annual turnover (I can't remember the exact criteria for whether it's 2% or 4%). The relevant word there is turnover, not profit. Now, I suspect VM's margins are more than 4%, but such a maximum fine really could be fairly described as punitive, and would certainly be hefty enough to wipe out any shareholder dividends and damage the share price. The trouble is, they'd probably just pass the cost onto the customers, unless the legislation explicitly prevents them from doing so.
To reach it they'd need to break into our network and then we'd be fscked anyway
You might be surprised by the number of attacks that come from within a corporate network, especially if the organisation is large and doesn't treat its techies well. Any organisation that takes security seriously practises proper data hygiene both externally and internally (oo-er).
RE: their call centre staff..
A few weeks ago, when we had all those winter storms coming over, we had an impromptu hail shower one evening, with one very loud and very nearby lightning strike. This somehow managed to induce enough current in the Virgin cable to cause the "V6" TV box to rapidly stop working with a loud bang and bright blue flash. Oddly it didn't blow up the cable modem which is on the same bit of co-ax but on the other branch of the filter, which probably says either something about either the filter or the cable modem being more robust than the TV box, but I digress.
When I called the Indian call-centre the next day and explained that it had been struck by lightning, the conversation went something like this:
Me: "Hello, we had a thunderstorm last night and the TV box was struck by lightning and stopped working"
Indian call centre worker: "What error code is it showing on the TV"
Me: "There is no error. It was struck by lightning. It's dead"
ICCW: "Are there any lights showing on the box"
Me: "No. It's dead. There was a bright flash and a bang. It was struck by lightning"
ICCW: "Is it plugged in
etc.
To their credit, they did send a new box out pretty quickly, but it goes to show that not even a so-called Act of God will get a call-centre operative to deviate from their script.
Well yes, I've actually been saying the same thing. I'm referring to the published mortality rates, which may be wildly inaccurate. Interesting to note that the communicability and mortality rates are both similar to those of the 1918 H1N1 flu pandemic. Of course, that happened in the middle of a time of global war, and medical treatment has progressed somewhat in the last century.
The problem with the Faraday cage idea is that it doesn't discriminate between signals from illicit mobile phones, and those from legitimate sources. I can't see how it would be hard to triangulate the origin of phone signals within the prison walls though, with a relatively small number of aerials.
addendum - the mortality rate by age group is, however, very informative. If you're under 50, you have pretty good odds. They get progressively worse as you get older; in the 80+ age bracket, the rate is around 15%. I'd link the figures here, but the top results I get from google are links to images in paywalled articles.
If you overlay the graph of mortality by age over one of incidence of COPD, there's an almost perfect match. Correlation isn't causation, of course, but it's not hard to posit a causal link.
It's almost as if the planet has taken a dislike to boomers and decided to do something about it.
In my experience, many companies over-tolerate the "super-type-A gladhanding salesy type without a knack for anything technical" types as well, despite the fact that they aren't "brilliant jerks" either.
It's probably because corporate boards seem to over-represent the type-A personalities (probably due to their knack for self-promotion) and ignore the tendency towards their affinity for white powder and misogyny.
If you have nothing, the barrier to work is pretty high. First you need to make yourself presentable. You'll need to have clean, smart clothes, which means you'll need to have somewhere to wash, which means you'll need to have somewhere to live, which means you'll need a job...
By all means, you can acknowledge that giving money to beggars isn't necessarily a helpful thing to do (for reasons I outlined above), but being a dick to them just makes you look like a dick.
It always struck me that you wouldn't have the screen of your laptop angled in such a way that the camera would record anything incriminating, unless your eyes happened to be in your balls.
Laptop cameras are, in my experience, without exception at the top of the screen angling towards the face of the person looking at the screen. The same goes for phone cameras.
Unless you're the sort of pervert who likes to toss themselves off over the laptop, then it's literally a case of "nothing to see here", even if the scam was half-way believable to start with.
And yet my wife still scolded me, when, sat in a cafe, a woman who was quite obviously a drug addict* from her demeanour, appearance and behaviour came up and started her spiel and I cut her off with a loud and forceful, "No".
Even though the scammer then proceeded to (successfully) beg money from the table next to us, and was trying it on people sat outside another place further down the street when I walked past five minutes later.
This sort of scam is as old as the hills, yet there still seem to be plenty of people taken in by it.
*The actual solution to drug addiction is proper treatment, and housing to get them off the street and somewhere safe. Decent social care for the most vulnerable was one of the first things to go under austerity, which is why our towns and cities are seeing an epidemic of homelessness, and of people with substance abuse issues. I'm not unaware of the problems such people face, but at the same time, I'm not giving them money to feed an addiction. I'll donate money to a shelter instead (and do).
This is another good point. I suspect the situation of profit-sharing is much worse now than it was some years ago.
I know at least one person who still gets patent dividends for work they did (IIRC for HP) in the '70s and '80s. The problem he has is trying to cash a US cheque in the UK; for the amounts involved, it's often just not worth it.
I don't disagree that they don't work well. The equation you have to consider is whether they work better than the alternatives we have, which as I understand it, are no patents system, or an attempt to reform it. The first is obviously worse, the second is not obviously achievable - although I'd not rule it out.
Indeed. Writing good regulations is very difficult, because people will almost always find a way around them.
This is why I prefer the other solution (publicly funded and publicly owned research).
It strikes me as odd that we acknowledge that public health is a common good, so have the publicly owned NHS*, but fail to recognise that the medicines the NHS uses are part of that public good.
*for now, just about - creeping privatisation of the NHS is a topic for another discussion.
Without the patent system, no company is going to sink hundreds of millions of quid into the drug discovery process, stage I-III clinical trials et al only for $generic_manufacturer to turn around and start knocking the pills out at 2% over cost, without any of the sunk costs. You can't even have the protections fo a "trade secret" because the testing, safety trials and approval process will have to clearly publish what is in those pills.
The patent system isn't great, but, to murder a Churchill quote, like democracy, it's worst system we have apart from all the others that have been tried.
This is a very good point.
I think it boils down to the fact that, unless you have large-scale state-funded pharmaceutical research, then it will be the remit of private enterprise.
Corporations have no moral compass, they exist merely to find the greatest profits and economic growth. Say what you like about capitalism, but this is its biggest flaw; the fact that it is fundamentally amoral, and the reason why you need regulation; also the reason why governments that want to get rid of "red tape" usually want to do so because they are in the pay of (or simply are) those who would profit from deregulation, and not for the public good.
So there are two pretty clear routes to solving the problems of big-pharma - publicly funded, public interest pharmaceutical research (and by all means, this can be for-profit, if that profit goes back into the public coffers), or more and more regulations to rein in the worst behaviours, such as punitive taxation on "lifestyle" drugs, and / or tax breaks for public-interest research.
I'd suggest the best place to look would be at pharmaceutical patents, since some of the world's biggest pharma companies are UK-based. These last 20 years, so I'd suggest that if the likes of GSK haven't been churning out new ones, they'd be losing money, which they most definitely are not.
IIRC, they are largely based on manufacturing processes for chemical compounds, and not the compound themselves, so do meet the criterion for "novel" as well.
Without getting into the discussion about price-gouging of on-patent pharmaceuticals, I will point out that the drug discovery process is difficult, expensive, and littered with blind-alleys, so I do think this sort of patent has merit.
I can only think of one nation that comes close to that "utopian ideal" - the DPRK...
The end-game seems to be to either become a pariah state, or the lackey of another (for the personal profit of those nominally in charge). I'm still trying to work out if it's Russia, or the US, although it's arguable that both are controlled by the same bunch of far-right nutters.
And I expect it's perfectly possible to create the equivalent of
do while (true) {} in both languages.
If a language is Turing-complete, it's not possible to specify it such that you can't create a bug, in much the same way it is not possible to determine if any program will complete.
Yeah , not pulling a non existant magic tool out of your ass.
Yeah, or doing a simple web search and finding such a "language" exists before displaying your ignorance of a well-known solution to a commonly set problem...
To answer your obvious straw-man analogy; if it was common practice in job interviews to ask about the kind of migration project plan you mention, then you can bet your bottom dollar that there would also be a well documented complete, well thought out way of doing it on the internet. My solution to the problem would be to start with that, and not make some bollocks up myself...
Wow.
My dad has an original copy of Pink Floyd's "Piper at the Gates of Dawn" (mono vinyl). Apparently that's for sale on eBay for £850, and his copy of Electric Ladyland is worth £400.
I know whose record collection I'll be nabbing when he goes, before the siblings see it...
Assuming you are testing for integer divisibility (as one poster above correctly stated), the number of division operations I'd be using would be zero. I'd use the modulo operator.
That, or state that my language of choice is one of my own invention. Lets call it splooge. Splooge has one valid operation, which is the keyword splug. When parsing this keyword, the output is to list the integers from 1 to 100, except for those exactly divisible by 3, or 5, for which it outputs "fizz", "buzz", or "fizzbuzz" accordingly. My program is as follows:
splug
Technically correct - the best sort of correct.
An important part of software development is knowing the correct tool for the job, and doing things in the best way possibly, after taking into account factors such as required performance, maintainability, and so on.
Technically, we still have "policing by consent" in this country, so unless you are being arrested, or stopped under a court order (e.g. for a legal stop-and-search*), you don't have to cooperate with the police. Whether this is sensible is another matter, but the Met do sound like several of their officers could do with a refresher course. See also: taking photographs of the police.
Compare and contrast with other countries where policing is not by consent, such as the US, where you have to do what the cop says, or risk getting shot, or at best cuffed and beaten.
*Let's not get into the subject of stop-and-search, and the legality of such under different circumstances, as once you start digging into that one, and into the issues it throws up like ethnic bias, you can get mired for quite some time...
It possibly also exposes a racial bias of the Met's finest in pursuing prosecutions of BAME ethnicity suspects
Top be fair to the Met here, it's the CPS that pursues prosecutions, nominally on the basis of whether they think there is sufficient evidence for a realistic prospect of conviction. The Met are just biased about who they stop and arrest in the first place. If you go after more BAME individuals, you're going to get more BAME convictions...
The former STASI headquarters in East Berlin is now home to a fascinating museum. It is to the credit of the German people that when the Berlin Wall fell and the building was stormed, they managed to preserve many of the very well-kept records that the people in the building were trying to destroy.
It's within walking distance of the Brandenburg Gate and I'd highly recommend it to anyone visiting Berlin as an insight into how far Bureaucracies can go.
Incidentally, it's also the filming location for the series Deutschland 83, where scenes set in the STASI headquarters were actually set in the STASI headquarters. The offices there are so well preserved, the only noticeably anachronistic thing is the modern lighting.
Never doubt how stupid people can be. Just think how dumb the average man in the street is, and then consider, that by definition, half of all people are dumber...
Just as there are individuals who reach the zenith of human intellect, there are their counterparts who reach the nadir.
Did it actually have a rocket engine in it, or was it a glorified "bottle-rocket"?
Whilst technically, this could be considered a rocket, more usually, you wouldn't call something a rocket unless it carried its own fuel / propellant. As far as I am aware, what he was launching was basically a tank full of live steam with some fins attached.
As for getting 500 feet off the ground, there are many, many ways you can manage that, without the means being a rocket in any way, such as using a balloon, or a helicopter, or jumping out of a window 2/3 of the way up the Burj Khalifa
I do get your point, but I'm going to continue to play Devil's Advocate with your analogy.
With a horse, you put oats in the front, and fertiliser comes out the back. You need to know how to make it go forward, how to stop, and how to turn.
When I took my driving test (and I acknowledge that this wasn't always the case), I was required to know how to open the bonnet, check the oil, washer fluid, brake fluid level, know about legal tyre tread levels, keep it fuelled, as well as having to demonstrate control of the vehicle, operate the gears (which I don't believe horses have), apply brakes, etc. etc. Of course, to anyone who has been driving for any length of time, most of the skills required to operate a vehicle become second nature, as, no doubt, those to drive a horse-drawn carriage would do. I'd posit that the set of skills for a car is more extensive than those for a horse.
As for the comparison between word processing software and a typewriter, it is worth noting that in evolutionary terms, there have been a series of progressions from typesetting, through manually operated typewriters, to electrical ones, to word-processing ones (as seen in the '80s and '90s), which allowed you to edit a line fo text before typing it, to "word processors" which were single-purpose computers, through text-editors and WYSIWYG word processors to what we have today with the likes of MS Word. The progression at times wasn't always in great leaps-and-bounds, but incremental. For instance the distinction between a typewriter and a "word processor" might not be immediately obvious to most.
The big difference between (any) word processing and (any) typewriter is the very unique capacity to edit text (rewrite parts, add stuff, copy/paste, in short change your mind till you're satisfied with the result), something no typewriter can do.
FWIW, I learned to type on an electric typewriter. Back in the day, there was such a thing as white-out paper (and probably still is), where you would put the sheet between the ribbon and the sheet you were typing on, and over-type your mistakes to white them out. Then type something else over the top.
It's not perfect, but it exists...
My parents got one of those as a wedding present in the '70s. It still works, although IIRC, it had a motor replacement some time in the '90s when the brushes wore out. The blender attachment (that sits on top of the mixer head, if you remove a metal panel from the top) still gets use from my dad.
It's going to be a long, long time before it suffers any sort of mechanical failure that is irreparable. You could probably drop it off a car-park onto concrete and it'd keep working (and make a good old dent in the concrete to boot.)
(protip: 3.5" floppy-USB converter won't do you any good)
What is the interface on it? If it is a standard cable (e.g. RS232 serial, or parallel), I'd get a USB-whatever converter and read up on how to hack together something that talks the same language. I might throw together something in C# to talk to it, or maybe dust off the Rasp-Pi. If it's something more proprietary, I'd wire it to the pin-outs on the Rasp-Pi.
Actually, I'd probably just google it and follow the instructions in a youtube video...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
