There's a lot to be said for that. A lot of the most well designed, perfectly
structured software ever written gets scrapped, yet much of the actual working
software is, as you say, full of GOTO statements, uses global variables etc.
If a thing works, there's a lot to be said for it. You can always get people
to write newer, better structured parts for it later, but most of all it has to work.
There are two types of doubt - doubt of innocence and doubt of guilt (as well as their inverses - certainty of innocence and guilt!) When the judge says "beyond reasonable doubt", which type of doubt does he mean? "Beyond reasonable doubt" may vary depending on whether you are straining every sinew to convict a wastrel, or whether you are doing all in your power to find a reason to exonerate the chap.
Perhaps it should be rephrased to say "plausible doubt". That way, "statistical convictions" would be eliminated. If circumstances suggest a chance of 1 in 6 billion, then it is almost certain to happen to one person at least. Yet using "beyond reasonable doubt" means you'll go to jail for something you haven't done, because 1 in 6 billion is pretty heavy.
Using the "plausible doubt" is different. If circumstances suggest a chance of 1 in 6 billion, then it is almost certain to happen to one person, and a defense could be based on "plausible doubt" that would get you off. It is, after all, completely plausible, as all debuggers can attest! The problem is that more crooks would get off too.
She was the most gullible Prime Monster they ever had. Reagan and Kohl outsmarted her at every turn. They all decided to cut back on subsidies to heavy industry, but only Thatcher went ahead. Reagan and Kohl did nothing, and the US and Germany still have their manufacturing base. Poor old Blightly lost everything but its accountants! Now all the kids want to be marketing managers, but they have nothing to market, bless them. Berners-Lee left for America a long time back.
Directly, it makes no odds whether there is a capacity problem - it's only about producers squeezing consumers, and consumers coughing up or doing without. That's business for you, and the idea is to control resources, and make people pay. A fair price is where both parties feel it's not fair. When you are exchanging bandwidth for money, the fair price is a compromise between the most bandwidth the consumer can get for his or her money, and the most money the provider can get for their bandwidth. That's how prices are set by markets. If a firm can get a lot out of a client, they'll do so - that's the system we have selected.
But I take your point that this system can be counter productive in new fields. Where do you draw the line?
I'm not so sure. In many other areas of life, we expect to pay more money for more of a thing, and to pay less for less of a thing - money is the standard way to manage scarcity since time immemorial. Should the Internet be immune from the usual laws of supply and demand, and why?
I think we have to accept that, in space flight, humans are not safe, else we spend the whole budget trying to work around "what if?" situations. That's brutal, but we all have short lives, and they all end the same way. It's good fun to send folks into space and see them on TV, but part of the fascination is to do with the isolation and danger of it.
> A goodly chunk of the population would surely
> steal to their heart's content if they knew they
> would never be caught and punished.
Some civilisations are stronger than that. During the Blitz, which lasted 8 months, London, Liverpool and other British cities lost around 50K lives, and a million houses - that was a catastrophe! Discipline did not break down, the British went on to defeat Rommel in north Africa and eventually triumphed over Hitler and his henchmen in Europe. Criminality during the war was rife, of course, but general standards remained in place.
We tend to be fearful about our society, mostly due to sensational rubbish in the media, in my opinion.
I know a control theorist who would disagree. He hates PID controllers with a passion.:)
What is it about these generic algorithms that your control theorist friend objects to? One of the good things about PID is that the methods can be simulated in a variety of computers, including mechanical ones that have no electronic parts, entirely eliminating fire risk. Are there other, better methods that can be implemented in pneumatic components?
Proportional action would seems reasonable to anybody - the size of the correction depends on the size of the error.
But it led to constant offsets (steady state error), so somebody dreamed up integral action (reset), where the rate of change of the correction is proportional to the size of the error. That seems fair to me - it boils down to "if it doesn't budge, give it bit more".
Then someone even smarter than that (!) dreamed up derivative action, where the controller output is proportional to the rate of change of the error. This can stop overshooting, I believe.
> I know a guy who used to work in computing and
> electrical engineering, in around the 60s, 70s, so
> pretty primitive stuff.
> Apparently at the time it was common to approximate integrals
> electronically by building up a charge on a capacitor over some
> time, representing the range of the integral, with the current
> behaving as the function to be integrated.
It's not all primitive - this is usually done in analog 3-term PID controllers. Where electonics cannot be used due to environment restrictions, pneumatic bellows (!) are often used to model integral action.
> He had to try and explain this concept to a member of
> senior management one day. The first question he was ask
> was "what the hell's an integral?".
I know what you mean - you'd have to teach him calculus before he'd understand.
What is the point of having sensors and computers when it is so easy
in my beat-up Toyota? Just back up until you feel a slight jolt from the car behind, or the alarm goes off, then pull forward a few inches.
Yes - we've got a 1/4 teraflop system down the hall (hundreds of nodes) and we light it up full scale for weeks at a stretch on some classes of computational chemistry problems.
A system like this one would bring the time span down to hours, not weeks, but it would still need to execute reliably flat out for that time.
The brits are desperately trying to beat the yanks to be the first to compute the square root of a negative number, and they are throwing everything they got at it.
The DeHavilland factory at Broughton, Wales, is still up and running, making the Airbus. Boeing did all that work on the 747 for (what's become of) DeHavilland to benefit.
Practically all brits have passports and frequently travel about in Europe. You might be able to get by in Canada or the states with no passport, but it would be a big deal in the UK.
Airbus has market dominance. They sell and deliver more jets than Boeing, so they must be doing something right, with or without EU assurances about Gallileo.
Yeah, but they'd have to be pretty smart ball bearings to tell the difference between US and Euro satellites. In any case, the Brits have put money into this, and they are the only friend left!
Fripp's a limey, like his King Crimson/Roxy Music chum, Eno, who has also created sound effects for Windows. Limies are very good at this type of thing. For example, check out http://www.delia-derbyshire.org/
Yeah, they graduated into FORTRAN Hyper-Enthusiasts instead. Pretty soon, they'll be Algol Hyper-Enthusiasts, then Pascal Hyper-Enthusiasts, then small-talk Hyper-Enthusiasts, then C++ Hyper-Enthusiasts, then finally they'll be Java Hyper-Enthusiasts again just in time to retire.
On the other hand, they might become C Hyper-Enthusiasts and stay like that...
> mouthful of broken teeth in the keyer's mouth
Yes, but violence can put you in gaol too. Keying SUVs now is far better than using nuclear weapons to knock out industrial polluters in 35 years time, when we are in survival mode.
Slashdot Mirror
This puts in the final nail, sorry - rivet - in the arguments!
> as connecting to a site next door.
Unless Russia IS next door, the cost must be more because of all the hops. The question is not whether the cost is higher, but who should pay it.
There's a lot to be said for that. A lot of the most well designed, perfectly structured software ever written gets scrapped, yet much of the actual working software is, as you say, full of GOTO statements, uses global variables etc. If a thing works, there's a lot to be said for it. You can always get people to write newer, better structured parts for it later, but most of all it has to work.
There are two types of doubt - doubt of innocence and doubt of guilt (as well as their inverses - certainty of innocence and guilt!) When the judge says "beyond reasonable doubt", which type of doubt does he mean? "Beyond reasonable doubt" may vary depending on whether you are straining every sinew to convict a wastrel, or whether you are doing all in your power to find a reason to exonerate the chap. Perhaps it should be rephrased to say "plausible doubt". That way, "statistical convictions" would be eliminated. If circumstances suggest a chance of 1 in 6 billion, then it is almost certain to happen to one person at least. Yet using "beyond reasonable doubt" means you'll go to jail for something you haven't done, because 1 in 6 billion is pretty heavy. Using the "plausible doubt" is different. If circumstances suggest a chance of 1 in 6 billion, then it is almost certain to happen to one person, and a defense could be based on "plausible doubt" that would get you off. It is, after all, completely plausible, as all debuggers can attest! The problem is that more crooks would get off too.
You appear to have missed a Conservative PM called John Major, who governed from 1990 to 1997. Are you sure you know much about British politics?
She was the most gullible Prime Monster they ever had. Reagan and Kohl outsmarted her at every turn. They all decided to cut back on subsidies to heavy industry, but only Thatcher went ahead. Reagan and Kohl did nothing, and the US and Germany still have their manufacturing base. Poor old Blightly lost everything but its accountants! Now all the kids want to be marketing managers, but they have nothing to market, bless them. Berners-Lee left for America a long time back.
Yeah - Winnie the Pooh is a limey!
Directly, it makes no odds whether there is a capacity problem - it's only about producers squeezing consumers, and consumers coughing up or doing without. That's business for you, and the idea is to control resources, and make people pay. A fair price is where both parties feel it's not fair. When you are exchanging bandwidth for money, the fair price is a compromise between the most bandwidth the consumer can get for his or her money, and the most money the provider can get for their bandwidth. That's how prices are set by markets. If a firm can get a lot out of a client, they'll do so - that's the system we have selected. But I take your point that this system can be counter productive in new fields. Where do you draw the line?
I'm not so sure. In many other areas of life, we expect to pay more money for more of a thing, and to pay less for less of a thing - money is the standard way to manage scarcity since time immemorial. Should the Internet be immune from the usual laws of supply and demand, and why?
I think we have to accept that, in space flight, humans are not safe, else we spend the whole budget trying to work around "what if?" situations. That's brutal, but we all have short lives, and they all end the same way. It's good fun to send folks into space and see them on TV, but part of the fascination is to do with the isolation and danger of it.
> steal to their heart's content if they knew they
> would never be caught and punished.
Some civilisations are stronger than that. During the Blitz, which lasted 8 months, London, Liverpool and other British cities lost around 50K lives, and a million houses - that was a catastrophe! Discipline did not break down, the British went on to defeat Rommel in north Africa and eventually triumphed over Hitler and his henchmen in Europe. Criminality during the war was rife, of course, but general standards remained in place. We tend to be fearful about our society, mostly due to sensational rubbish in the media, in my opinion.
> I know a guy who used to work in computing and
> electrical engineering, in around the 60s, 70s, so
> pretty primitive stuff.
> Apparently at the time it was common to approximate integrals
> electronically by building up a charge on a capacitor over some
> time, representing the range of the integral, with the current
> behaving as the function to be integrated.
It's not all primitive - this is usually done in analog 3-term PID controllers. Where electonics cannot be used due to environment restrictions, pneumatic bellows (!) are often used to model integral action. > He had to try and explain this concept to a member of > senior management one day. The first question he was ask > was "what the hell's an integral?". I know what you mean - you'd have to teach him calculus before he'd understand.
What is the point of having sensors and computers when it is so easy in my beat-up Toyota? Just back up until you feel a slight jolt from the car behind, or the alarm goes off, then pull forward a few inches.
Yes - we've got a 1/4 teraflop system down the hall (hundreds of nodes) and we light it up full scale for weeks at a stretch on some classes of computational chemistry problems. A system like this one would bring the time span down to hours, not weeks, but it would still need to execute reliably flat out for that time.
The brits are desperately trying to beat the yanks to be the first to compute the square root of a negative number, and they are throwing everything they got at it.
The DeHavilland factory at Broughton, Wales, is still up and running, making the Airbus. Boeing did all that work on the 747 for (what's become of) DeHavilland to benefit.
Practically all brits have passports and frequently travel about in Europe. You might be able to get by in Canada or the states with no passport, but it would be a big deal in the UK.
Airbus has market dominance. They sell and deliver more jets than Boeing, so they must be doing something right, with or without EU assurances about Gallileo.
I've never known a Canadian who would admit that in public!
Yeah, but they'd have to be pretty smart ball bearings to tell the difference between US and Euro satellites. In any case, the Brits have put money into this, and they are the only friend left!
Fripp's a limey, like his King Crimson/Roxy Music chum, Eno, who has also created sound effects for Windows. Limies are very good at this type of thing. For example, check out http://www.delia-derbyshire.org/
Yeah, they graduated into FORTRAN Hyper-Enthusiasts instead. Pretty soon, they'll be Algol Hyper-Enthusiasts, then Pascal Hyper-Enthusiasts, then small-talk Hyper-Enthusiasts, then C++ Hyper-Enthusiasts, then finally they'll be Java Hyper-Enthusiasts again just in time to retire. On the other hand, they might become C Hyper-Enthusiasts and stay like that...
> mouthful of broken teeth in the keyer's mouth
Yes, but violence can put you in gaol too. Keying SUVs now is far better than using nuclear weapons to knock out industrial polluters in 35 years time, when we are in survival mode.