Problem is, those extremely complex NURBS surfaces have all sorts of trims around the perimeter and in the surface itself to match bolt holes, and other components. It's not just a simple regular square or triangular patch.
Calculating the local tangent space of each point of a regular or N-sided patch, isn't too difficult (tangent, normal, binormal), it's all the trimming that takes up the time. Just a single bolt with a spiral thread is going to generate a whole bucketload of triangles per revolution of that thread. Another complication is that the CATIA file format isn't simply a geometry files, it's more of relational database entry, where everything is cross-referenced to the manufacturer, specification standard, measurement units, required sub-parts. That way, you just have one file, it pulls in everything else that you need to view that one part.
I'd say it's more that we just need practise to build up the connections to all the different areas of the brain. Much like when solving a mathematics problem, there are so many different ways of looking at the problem. Sometimes it helps to use algebra or to draw a picture or diagram. That makes use of the visual areas of the brain - around 30% in mammalian brains. For simple arithmetic we usually learn times-tables (1x1 = 1, 2x2=2, 3x2=6,... 12x12 = 144). Doing this for certain larger numbers is easy (350x200, 400x250), but the extra digits requires practise in remembering the intermediate results and performing the final addition. If you look as silicon design papers, they would have papers on how multiplication and division could be optimized with fewer transistors.
The brain has an architecture like a super-computer, with a data-flow design. If you look at the diffusion MRI/CAT scan images which look at the directional bias of the fibres in the white matter of the brain (the long-distance communication network), there is definitely a tree-like network. Then all the different areas of the brain have separate purposes, like object image->name, or object image->orientation, sound->object name, objects images->distance (stereoscopic vision). Rehabilitation clinics with stroke and brain injury patients have diagnostic tests that are to measure impairment in these areas. Some patients could tell that there was an apple in front of them, but not whether it was upside down or not. Others would know there was an object in front of them, but not know what is was, or even how far away it was.
Figuring out how to make supercomputers superscalar was one of the research areas in the past. That was solved by figuring out to rack, stack and pack multi-core CPU's on standard boards that fitted into scalable rack units. Then there was the problem of inter-node communication. All sorts of topologies were considered; grid, cube, torus, hyper-torus, hyper-cube, then they figured out having a dynamically configurable topology adapted to the algorithm being run worked best. Once you've got those two sorted out, it really does reduce down to a pissing match of who can politically afford the most hardware. Extend this concept to GPU's, and that is the situation now.
Figuring out how to pack multiple cores onto a single chip is the simplest way to reduce space, power and cost. Once you've got a GPU with 500+ stream processors, your only option is mathematics and algorithm research, but that goes under the scientific field those algorithms are being used for; CFD, climate modelling, biochemistry, protein folding.
Some time ago, parallel computing research labs (EPCC) were complaining that there was no real diversity in the CPU research field. Everyone just had to take whatever standard commodity hardware was on special offer at the time, as all the variety had gone - basic price/FLOPS performance governed the purchase sale. Back in the 1980's and 1990's, there was a real variety of systems for different applications; Pixel Planes, MIPS, Dec Alpha's, Sun SPARC, SGI MIPS, The Connection Machine, Cray supercomputers, Transputers. Each had their own ideas on compiler design/instruction set/CPU registers/inter-processor design; RISC vs. CISC, SIMD vs MIMD, bitwise data vs. floats vs. vectors. Whichever vendor managed to incorporate all those ideas into a single CPU became the winner.
The only other competitor in terms of performance to GPU's are custom ASIC's and FPGA design, because they remove the overhead of instruction interpretation.
Sadly, these lawsuits aren't new - some companies filed patents on multi-player network games in the 1980's, and proceeded to sue other game companies for using those techniques, despite the fact that similar games had been written and designed at universities and other research labs in the 60's and 70's. Even bedroom game programmers had worked on multi-player games using RS-232 ports.
Anything to do with paging and interrupts could be a security vulnerability - some kernel processing has to go on in order to update process states. Being able to interrupt the process at the point just after getting kernel ring permission and before interrupts are disabled would be a dream, but probably just a theoretical one. Usually, it would be an atomic process, you couldn't do one without the other, but with these instructions, who knows...
Our OS theory courses covered early OS's which had a messed-up login process - something like a custom system would start off in admin mode to read the password file, change to the selected UID for "safety", then decide whether the password was correct and start a shell in that UID or return to admin to wait for another login. Somehow, users could attach a debug process at the right moment (maybe when the system was running really slow), set a breakpoint at the password test, fudge the comparison result, and set the UID to whatever they wanted. Instantly, a superuser login....
It has practical uses as well... if you are running a combustion based rocket motor with a continuous flow of liquid fuel, you really don't want the ignition to travel back through the fuel lines. At the same time, you don't want so much fuel going out that it is unburnt. One way to solve this is to have a narrowed constriction along the fuel line which forces the pressure and velocity up, thus preventing any backflow.
I remember those days - reading his book on the "Fractal complexity of nature" was a real inspiration. It was strange to realize that snowflakes, ice crystallisation, mountain terrain, the outlines of coastlines, branching of trees and lightning, aggregation of soot particles, growth of coral and seashells, periodicity of landslides and earthquakes could all be modelled by fractals.
Some of those simulations could be done within seconds on an Atari(XL) or other home computer. Others took hours like the Mandelbrot set as well as others like John Conway's Game of Life - the 1D version was a bit faster. Spending three Summer evenings running a 6502 implementation of John Conway's "Life" program on a for all 1000+ generations on a 160x80 grid. I always remember the stars in the twilight sky at that time looked just like the cover of the 1978 BYTE magazine.
In the wild, mice live in tunnels under tree roots and in hills. House mice have adapted to living with humans and taken advantage of the warmth provided by human dwellings. Either way, they have to remember where food and water can be found, and the safest places to sleep.
Just about every creature with a hypothalamus (where route memories are stored, as well as being wired to the vision and audio pathways) will be able to remember all these things.
Sometimes those petroleum companies built their refineries in the middle of nowhere, only to have property developers choose to build land right next to them because the surrounding land was cheap and house prices were high elsewhere due to a local economic boom-time. Then people would be pleading for homes to be built anywhere; under the flight paths of airports, under high-voltage power lines, right next to chemical processing plants, in flood-plains or over disused mines.
75% of that surface is water, and is not inhabited. Of the remaining 25% which is land, only around 12% is suitable for farming and 4% has been urbanised into cities, roads, factories and housing. That would give each person a much smaller area of land to modify.
Creativity is not conducive to performing difficult manual tasks.
Most of the difficult and dangerous manual tasks seems to be done using programmble machinery these days. This leaves the actual design work done using a 3D modelling application.
From my experience of high-school woodwork and metalwork, both seemed to require mental visualization in three dimensions. With just a choice of two wood varnishes or metal coatings, all sorts of different appearances could be created. My next door neighbor from 20 years ago was a joiner/carpenter/electrician. He did whole building conversions from installing bathrooms and kitchens to room layouts. He would never waste a single piece of scrap wood, but always use to experiment with different polishing and varnishing techniques to show customers.
We can see infra-red - if you put an infra-red filter (Hoya) over your eyes, and let them adjust in a dark room, you will be able to see around in infra-red. But that frequency of light gets washed out by the stronger blue-green-yellow light from the sky, grass and sea. Infra-red is only useful if you are hunting in caves or at dusk/dawn like a snake.
That simply means, they hadn't practised that particular interview technique before. Much like those for-fun IQ tests that you can find online - some are simply word-matching (A is to B as C is to [D E F]? Others are tests of mental rotation/flipping of shapes and path prediction. More complex ones are deductive logic with arithmetic and remainders. Software development houses have their own tests which usually concentrate on list management, memory allocation as well as languages such as C, C++, C# and Visual Basic.
Between the time you do one and a handful, your score will improve. No different from the pop quiz's you would do in high school or college.
Some fish such as North Sea salmon seemed to prefer to swim against the ocean current. This was proved with ocean tank experiments. It wouldn't matter which direction the water was coming from, the time of day, the position of the Moon, Sun or stars, the fish would swim in the direction the water was being recirculated. If one fish changes direction, that would cause vortices and turbulence which in turn would create pressure changes that other fish would be able to detect.
You can do the same thing on the London underground - get off at any platform at a busy time and walk to the platform in the opposite direction. At least 10 or more tourists will follow you and each other, thinking you are heading to the street (the WAY OUT signs don't always point somewhere meaningful). After you stop, they figure you weren't going where they want to go, and start following someone else.
and the entitled college graduates throw around terms like "ERD" but have no actual skill and balk at Help Desk offers because they think it's beneath them.
The Registers B.O.F.H. series doesn't help either - and from my experience it is true. I am sure I heard the same comments by the drill instructors of the drafted college students during the Vietnam war about 40 years ago...
You could buy a PAYG mobile phone in cash in any store in the UK. But if you topped up using an ATM (just about every machine allow you to top up a mobile phone number these days), that anonymity would go.
Yes, you would want to get the Mobicarte - you could order one off any reputable shops in France which have an online presence on Ebay or Amazon. But they would still require you to send off a copy of your passport. It would save you money as you would only be paying for local network prices.
However, any modern mobile phone has some web browser functionality built in, and as long as the SIM card for that account was topped up, you would be able to surf the web using that phone and at international roaming prices. Many UK PAYG mobile phones have a USB connector for battery charging as well as data transfer from interal memory and for use as a GPRS/3G wireless modem.
France takes their mobile phones communications very seriously - you have to provide a photocopy of your ID just to get a Mobicarte (Pay-As-You-Go) SIM card. Even then it will be deactivated if you don't use it after three months.
I get the impression they really want to know the identity of anyone who surfs the web.
The worst-case scenario is that someone abroad has managed to set up their own command-and-control network inside the country and is using that to interfere with the normal operation of an online infra-structure system such as a traffic light system, electricity or water supply (makes me think someone's been watching too much Whiz Kids). They wouldn't know where the commands are coming from, or how they are getting to the site of attack, so the best option is to shut down the entire Internet. Much like the botnet's, the system would just switch from one service (http, ftp) to another in order to maintain communication.
If he was smart, he would have generated the encryption key automatically and randomly, and have it stored on a memory stick or mobile phone. Then he would never have a password that could be guessed through social engineering or psychology. The only thing would to keep secret would be where the encryption key was stored. Maybe that could be randomized as well.
Slashdot Mirror
Problem is, those extremely complex NURBS surfaces have all sorts of trims around the perimeter and in the surface itself to match bolt holes, and other components. It's not just a simple regular square or triangular patch.
Calculating the local tangent space of each point of a regular or N-sided patch, isn't too difficult (tangent, normal, binormal), it's all the trimming that takes up the time. Just a single bolt with a spiral thread is going to generate a whole bucketload of triangles per revolution of that thread.
Another complication is that the CATIA file format isn't simply a geometry files, it's more of relational database entry, where everything is cross-referenced to the manufacturer, specification standard, measurement units, required sub-parts. That way, you just have one file, it pulls in everything else that you need to view that one part.
I'd say it's more that we just need practise to build up the connections to all the different areas of the brain. Much like when solving a mathematics problem, there are so many different ways of looking at the problem. Sometimes it helps to use algebra or to draw a picture or diagram. That makes use of the visual areas of the brain - around 30% in mammalian brains. For simple arithmetic we usually learn times-tables (1x1 = 1, 2x2=2, 3x2=6, ... 12x12 = 144). Doing this for certain larger numbers is easy (350x200, 400x250), but the extra digits requires practise in remembering the intermediate results and performing the final addition. If you look as silicon design papers, they would have papers on how multiplication and division could be optimized with fewer transistors.
The brain has an architecture like a super-computer, with a data-flow design. If you look at the diffusion MRI/CAT scan images which look at the directional bias of the fibres in the white matter of the brain (the long-distance communication network), there is definitely a tree-like network. Then all the different areas of the brain have separate purposes, like object image->name, or object image->orientation, sound->object name, objects images->distance (stereoscopic vision). Rehabilitation clinics with stroke and brain injury patients have diagnostic tests that are to measure impairment in these areas. Some patients could tell that there was an apple in front of them, but not whether it was upside down or not. Others would know there was an object in front of them, but not know what is was, or even how far away it was.
Figuring out how to make supercomputers superscalar was one of the research areas in the past. That was solved by figuring out to rack, stack and pack multi-core CPU's on standard boards that fitted into scalable rack units. Then there was the problem of inter-node communication. All sorts of topologies were considered; grid, cube, torus, hyper-torus, hyper-cube, then they figured out having a dynamically configurable topology adapted to the algorithm being run worked best. Once you've got those two sorted out, it really does reduce down to a pissing match of who can politically afford the most hardware. Extend this concept to GPU's, and that is the situation now.
Figuring out how to pack multiple cores onto a single chip is the simplest way to reduce space, power and cost. Once you've got a GPU with 500+ stream processors, your only option is mathematics and algorithm research, but that goes under the scientific field those algorithms are being used for; CFD, climate modelling, biochemistry, protein folding.
Some time ago, parallel computing research labs (EPCC) were complaining that there was no real diversity in the CPU research field. Everyone just had to take whatever standard commodity hardware was on special offer at the time, as all the variety had gone - basic price/FLOPS performance governed the purchase sale. Back in the 1980's and 1990's, there was a real variety of systems for different applications; Pixel Planes, MIPS, Dec Alpha's, Sun SPARC, SGI MIPS, The Connection Machine, Cray supercomputers, Transputers. Each had their own ideas on compiler design/instruction set/CPU registers/inter-processor design; RISC vs. CISC, SIMD vs MIMD, bitwise data vs. floats vs. vectors. Whichever vendor managed to incorporate all those ideas into a single CPU became the winner.
The only other competitor in terms of performance to GPU's are custom ASIC's and FPGA design, because they remove the overhead of instruction interpretation.
Sadly, these lawsuits aren't new - some companies filed patents on multi-player network games in the 1980's, and proceeded to sue other game companies for using those techniques, despite the fact that similar games had been written and designed at universities and other research labs in the 60's and 70's. Even bedroom game programmers had worked on multi-player games using RS-232 ports.
Tales of Silicon Valley: Bruce Damer on Maze War
Anything to do with paging and interrupts could be a security vulnerability - some kernel processing has to go on in order to update process states. Being able to interrupt the process at the point just after getting kernel ring permission and before interrupts are disabled would be a dream, but probably just a theoretical one. Usually, it would be an atomic process, you couldn't do one without the other, but with these instructions, who knows...
Our OS theory courses covered early OS's which had a messed-up login process - something like a custom system would start off in admin mode to read the password file, change to the selected UID for "safety", then decide whether the password was correct and start a shell in that UID or return to admin to wait for another login. Somehow, users could attach a debug process at the right moment (maybe when the system was running really slow), set a breakpoint at the password test, fudge the comparison result, and set the UID to whatever they wanted. Instantly, a superuser login....
The Silicon Zoo website which shows what kind of artwork the layout engineers like to add to any unused area of silicon die..
There were also 24-bit/32-bit color paint systems like the Quantel paintbox, and Tempra as well.
It has practical uses as well... if you are running a combustion based rocket motor with a continuous flow of liquid fuel, you really don't want the ignition to travel back through the fuel lines. At the same time, you don't want so much fuel going out that it is unburnt. One way to solve this is to have a narrowed constriction along the fuel line which forces the pressure and velocity up, thus preventing any backflow.
I remember those days - reading his book on the "Fractal complexity of nature" was a real inspiration. It was strange to realize that snowflakes, ice crystallisation, mountain terrain, the outlines of coastlines, branching of trees and lightning, aggregation of soot particles, growth of coral and seashells, periodicity of landslides and earthquakes could all be modelled by fractals.
Some of those simulations could be done within seconds on an Atari(XL) or other home computer. Others took hours like the Mandelbrot set as well as others like John Conway's Game of Life - the 1D version was a bit faster. Spending three Summer evenings running a 6502 implementation of John Conway's "Life" program on a for all 1000+ generations on a 160x80 grid. I always remember the stars in the twilight sky at that time looked just like the cover of the 1978 BYTE magazine.
In the wild, mice live in tunnels under tree roots and in hills. House mice have adapted to living with humans and taken advantage of the warmth provided by human dwellings. Either way, they have to remember where food and water can be found, and the safest places to sleep.
Just about every creature with a hypothalamus (where route memories are stored, as well as being wired to the vision and audio pathways) will be able to remember all these things.
Sometimes those petroleum companies built their refineries in the middle of nowhere, only to have property developers choose to build land right next to them because the surrounding land was cheap and house prices were high elsewhere due to a local economic boom-time. Then people would be pleading for homes to be built anywhere; under the flight paths of airports, under high-voltage power lines, right next to chemical processing plants, in flood-plains or over disused mines.
75% of that surface is water, and is not inhabited. Of the remaining 25% which is land, only around 12% is suitable for farming and 4% has been urbanised into cities, roads, factories and housing. That would give each person a much smaller area of land to modify.
Creativity is not conducive to performing difficult manual tasks.
Most of the difficult and dangerous manual tasks seems to be done using programmble machinery these days.
This leaves the actual design work done using a 3D modelling application.
From my experience of high-school woodwork and metalwork, both seemed to require mental visualization in three dimensions.
With just a choice of two wood varnishes or metal coatings, all sorts of different appearances could be created. My next door neighbor from 20 years ago was a joiner/carpenter/electrician. He did whole building conversions from installing bathrooms and kitchens to room layouts. He would never waste a single piece of scrap wood, but always use to experiment with different polishing and varnishing techniques to show customers.
It really goes into conspiracy-theory territory here:
BP Subcontractor warned of incomplete design documents
BP cut corners while constructing well
Liberals claim Haliburton at fault
BP, Transocean, Halliburton will blame one another for the spill
It's going to be a real wham-dinger seeing them fight this one out ...
We can see infra-red - if you put an infra-red filter (Hoya) over your eyes, and let them adjust in a dark room, you will be able to see around in infra-red. But that frequency of light gets washed out by the stronger blue-green-yellow light from the sky, grass and sea. Infra-red is only useful if you are hunting in caves or at dusk/dawn like a snake.
That simply means, they hadn't practised that particular interview technique before. Much like those for-fun IQ tests that you can find online - some are simply word-matching (A is to B as C is to [D E F]? Others are tests of mental rotation/flipping of shapes and path prediction. More complex ones are deductive logic with arithmetic and remainders. Software development houses have their own tests which usually concentrate on list management, memory allocation as well as languages such as C, C++, C# and Visual Basic.
Between the time you do one and a handful, your score will improve. No different from the pop quiz's you would do in high school or college.
Some fish such as North Sea salmon seemed to prefer to swim against the ocean current. This was proved with ocean tank experiments. It wouldn't matter which direction the water was coming from, the time of day, the position of the Moon, Sun or stars, the fish would swim in the direction the water was being recirculated. If one fish changes direction, that would cause vortices and turbulence which in turn would create pressure changes that other fish would be able to detect.
You can do the same thing on the London underground - get off at any platform at a busy time and walk to the platform in the opposite direction. At least 10 or more tourists will follow you and each other, thinking you are heading to the street (the WAY OUT signs don't always point somewhere meaningful). After you stop, they figure you weren't going where they want to go, and start following someone else.
and the entitled college graduates throw around terms like "ERD" but have no actual skill and balk at Help Desk offers because they think it's beneath them.
The Registers B.O.F.H. series doesn't help either - and from my experience it is true. I am sure I heard the same comments by the drill instructors of the drafted college students during the Vietnam war about 40 years ago...
KLM used to sell or give away wall-posters of the flight-deck of their aircraft.
To think back in the 70's, pilots on international flights used to allow parents and children to visit the flight deck.
You could buy a PAYG mobile phone in cash in any store in the UK. But if you topped up using an ATM (just about every machine allow you to top up a mobile phone number these days), that anonymity would go.
Yes, you would want to get the Mobicarte - you could order one off any reputable shops in France which have an online presence on Ebay or Amazon. But they would still require you to send off a copy of your passport. It would save you money as you would only be paying for local network prices.
However, any modern mobile phone has some web browser functionality built in, and as long as the SIM card for that account was topped up, you would be able to surf the web using that phone and at international roaming prices. Many UK PAYG mobile phones have a USB connector for battery charging as well as data transfer from interal memory and for use as a GPRS/3G wireless modem.
France takes their mobile phones communications very seriously - you have to provide a photocopy of your ID just to get a Mobicarte (Pay-As-You-Go) SIM card. Even then it will be deactivated if you don't use it after three months.
I get the impression they really want to know the identity of anyone who surfs the web.
The worst-case scenario is that someone abroad has managed to set up their own command-and-control network inside the country and is using that to interfere with the normal operation of an online infra-structure system such as a traffic light system, electricity or water supply (makes me think someone's been watching too much Whiz Kids). They wouldn't know where the commands are coming from, or how they are getting to the site of attack, so the best option is to shut down the entire Internet. Much like the botnet's, the system would just switch from one service (http, ftp) to another in order to maintain communication.
If he was smart, he would have generated the encryption key automatically and randomly, and have it stored on a memory stick or mobile phone. Then he would never have a password that could be guessed through social engineering or psychology. The only thing would to keep secret would be where the encryption key was stored. Maybe that could be randomized as well.