I would like to see the filesystem of an OS partitioned into several levels: read-only disk drives where stuff never changes unless an update occurs (kernel, device drivers, configuration files), read-write disks where log files are update by the minute, hour or day, and local/user partition which is updated by the user.
Our university managed to do something similar by just having a ISO image that they overwrite the OS partition with, every time the PC was rebooted.
There was a guy at our university wanting to do some university psychology tests and figured the best way for the application to log the results was to send them as an E-mail to himself, where they could be timestamped independently. Only problem was that any application that wasn't on the PC's anti-virus whitelist was blocked from opening that port. So he just renamed his experiment application to "Agent.exe" and the anti-virus software allowed the message to be sent.
I am sure it would be possible for someone to create an Opera or Firefox plugin that filters downloads of slashdot beta articles and converts them back to slashdot classic.
Until I saw what slashdot beta looked like, I thought they were trolls too. But now that I see that half the browser window is going to be blank just so that the poll window can have more white space, I sympathize with them. It should be up to each slashdot reader to choose how they read slashdot. Other sites give their forum members a choice in font size, background and text colors, and theme.
Or it allows twice a many cases to conducted for half the price. The only limit on litigation in the past has been the expense of lawyers. In the UK, we have "ambulance chasers", "no win, no fee" lawyers who look for every opportunity to win a compensation payout.
Every drone could have it's own small radio transmitter. If it detected the signal of another drone (the inverse square distance law), it could adjust it's course automatically.
I've seen hardware engineers print out images of chip designs, but they weren't looking at track layouts, they were looking at transistor densities. The image looked like a a density map of a large American city, with empty space in white, low density in green, medium density in yellow, orange and red, and high density in gray and black. I saw what looked like a exclusive leafy-green street surrounded by green parks. Like a sadistic urban planner, the engineer said, "oh, those are multiplexers, don't worry, we'll get them to move, it's costing us money to have a block like that, we like everything to be high density".
Some of them most complex control systems for aircraft are on the order of five million lines of code plus; device drivers for every possible component and sensor, user interface systems for the pilot information system and some AI for the autopilot logic.
There are two places where I know that visual design tools are used; the first is QT Designer which is used to design application frames and dialog windows with radio buttons, sliders, text windows. However, this still requires that you connect all the input/output events between widgets manually. And those get really messy. That could be eliminated though, since if you add a push-button to a window, it is practically guaranteed that you want to know about the press/release event, so those event handlers could be added automatically. Another place is the composition editor of Blender where you can place and connect processing nodes to do image processing and lighting. Once again that still requires the data flow to be connected up manually.
While a visual programming environment could be extended to do mathematics, there would need to be some way of handling complexity - some image processing and geometry algorithms consist of 20 or 30 more different stages, with each stage consisting of a 100+ line algorithm. If this environment supported hierarchical designs like data-flow diagrams, it might work, but it would just become a complete mess otherwise.
I believe it is called a "stress test", one of the techniques that banks use to test the resilience of a trading system and how much liquidity it has. If they suddenly shutdown the exchange of currency, the punters are forced to make trades using other resources. How much of that other currency is used gives an idea of how is available. The Russians and EU did the same by restricting physical money currency transfers. The EU figured out there was black market the size of the GDP of the entire region.
There was the "Stop and Catch Fire Instruction" on 6809 chips. Designers decided to share the input/output ports using the same pin connects. These were selected between by a couple of bits. Set both bits at the same time and the silicon die went into meltdown. Some graphics card would do the same with a driver update.
I guess you could always have a compressed cylinder of nitrogen, release it suddenly onto the components to make them brittle and vibrate them at high frequency so they basically disintegrate into dust.
Well, at least having screws are better than having held together by glue, and the high-end models (PCG-GRT series) did have the hard disk drive on the edge of the laptop rather than some cage in the middle of the chassis.
I know what you mean about all those screws. Replacing the screen or even the cooling assembly would require several ice trays to put every different component. The best way to describe how Sony must have designed a laptop was to start off with a gaming PC with a GTX Titan GPU and motherboard, then slice up the motherboard until everything could fit inside the profile of a laptop.
Around 2004, the "high-end" models had Nvidia Geforce 5600Go's with about 64 Megabytes of memory, which was enough to run Bzflag. It was impressive to see a laptop do texture mapping, when previously, it would have taken a Silicon Graphics Extreme to do the same.
Sony did the same. But they usually made around several dozen different models around a single laptop chassis and priced each model separately, and had different sized screens, disk-drives and memory configurations and maybe an option TV tuner board. By far the LCD screen was the most expensive replacement, at around $1000 . Some companies actually accepted the old screen as a trade-in discount. So it was cheaper to buy the model with the highest resolution screen, but the lowest price, then replace the hard disk drive and memory. But the GPU's were always custom made. There was no way of upgrading as they had custom connectors.
I remember when Dos-Shell came out - it allowed you to view two directory listings side by side in a 120 x 50 text window split in half. That was considered revolutionary at the time.
The monitor and disk drives are the component most likely to break. The last thing I would want to do is to have to replace one simply because the other broke. The best design I can think of would be to go back to "podules" where the CPU/GPU/memory and disk drives were on separate blocks but slotted into the base of the monitor.
That happened in Edinburgh as well. The banks were offering starting salaries of £50K+ bonuses to those with the best qualifications. Those salaries exceeded what the professors with 20+ years experience and were leading research teams were earning. PhD students were lucky to get a £12K research stipend/year, and post-docs were lucky to get £30K/year.
Rent for an apartment anywhere central, on a bus-route or in leafy-green parts of the city were around £850, and food costs are £50/week per person.
The issue with Unity has come up many times - just about every forum article I had read about the Unity Lens was "How do I remove this feature?". These were everyone from sys-admins to scientific programmers and application developers who just wanted to do their work, and not find themselves browsing Amazon products, when trying to maintain their system.
The second issue which is more serious, involves network security. Just about every sys-admin will tell you that every TCP/IP port that is open to receive data is a potential security threat for network worms/viruses. Ubuntu loves to open the printer server (cups), a web server (apache2), network time protocol (ntp) and dhclient (dhcp), and mysteriously the control panel controls either don't work or say that the service has been closed when it hasn't, or that it is under control by some other OS process but that process doesn't support shutdown. Every time I have read articles about this, the developers seem to be fairly defensive, that "Well, it's your own fault if one day you try and connect a printer and you can't print". Then you'll find your web-browser (Opera) connects to Google (*1e100.net) and (backoo.canonical.com) when you haven't even visited another web page.
This attitude has led to the formation of a new version of Linux called "Mint".
I've been in that situation too. Head office management from 300 miles away had decided that every new programming recruit would cut their teeth on some back-end part of the application (and completed different from what they had been told what they were doing). The first result was that they had turned their workplaces into a coding gulag" where every programmer who had ever been put in that situation made "escape" their highest priority. Each and every one of those people had managed to escape by either leaving to set up their own company, move to a different part of the team or moving to another company.The second result was that this particular block of code ended up looking the results of one of the those drawing games where each person draws on part of a sheet of paper but never sees what anyone else has drawn. Consequently, it was never going to work.
It happened in the past with telephone exchanges. They had some self-maintenance code built in such that if one exchange detected a malfunction of some sort (accounts balance fail to match, line quality not good enough), it would send a fault message and a shutdown notice to it's neighboring exchanges. But there was a little bug. The message first hop was correct as it sent the ID of the originating exchange, the message relayed second and later hops was wrong because it sent the ID of the current exchange. Thus when one exchange went down, it shut down all the other exchanges.
Your classic desktop is a menu bar at the top with pull-down menus, mini icons for each desktop window, and the classic trash bin and folders/files on the desktop area itself. Every desktop window has maximize/iconize/minimize icons, scrollbars and other icons built into the window frame. The menu-bar includes dialogs for configuring printers, disabling/enabling network devices, network traffic sent/received per device and protocol, CPU/GPU temperature controls and power-on/restart/suspend controls.
Just about every desktop developer wants to try and change this in order to either look super-modern or conform to mobile devices. Sometimes they'll add fancy graphical effects like transparency, blurred transparency, smooth scaling windows, fade ins/fade outs. Others will eliminate the idea of resizable window frames and just try and force you to toggle between application windows, or they create a scrolling map desktop where all the windows click to each others edge. Others decide that you don't real want to have control over those settings, so just get rid of those dialogs. Others get rid of the pull-down menus requiring multi-language support and just have little color icons.
Canonical tried to get rid of the menu bar and just give you a search-window like google's web-page (The Unity Lens). Only problem was that *everything* you searched for was actually sent back to their servers, and for your efforts, they would send back a list of CD's, DVD's and computer games that you might want to play. Even the existing network control dialogs didn't work or actually disable network server processes they were supposed to (eg. network time, cups, apache web server, dhclient).
I like that idea. You could have modified commands like 'ls' that would display little thumbnails of each image, and a 'more'/ 'cat' commands that would just dump each of those images at full size or scaled to fit the terminal.
In the early day of internet research, they wanted to distinguish between corporations there to make a profit, non-profit organisations, educational groups and the military. So they had ".com" = corporations/companies, ".org" = non-profit organisation, and ".edu" for the educational research groups, ".mil" for the military", and ".net" for the companies that managed the continent wide networks built from fibre-optics and satellite communications.
That gets extended to giving each country it's own domain ".uk" = UK, ".fr" = France, ".de" = Germany. It was a sense of achievement for any country to get their own top-level domain.
Those smarter drives do insane things that having a pool of surplus disk blocks and having a virtual disk cylinder/sector map that can swap out old blocks that have become damaged and replace them with a new block. Just because you think you are writing on cylinder 32, sector 5, block 3, doesn't mean it's really at that location. Theoretically, it might be possible to fill up every possible block with data, but that's no guarantee.
So the only safe way is to destroy the hard disk drives.
Weirdo AMD stuff? The descriptors that DirectX and Mantle use are basically the registers that control your GPU. Traditional desktop GL has the large GL state (textures + framebuffer objects + blending + vertex buffers + transform feedback) and which is what the driver maintains and attempts to sanity check with all those GL calls (500+ of them) to make sure nothing inconsistent or fatal is allowed to run on any of the GPU cores. Setting a single parameter can involve cross-referencing dozens of other variables. If you look at some of the GL commands with the most parameters such as the texture image uploads, the driver has to sanity-checking ten or more parameters eg. the texture image functions have to check texture width, height, type, internal format, mip-map level, pixel data source type, pixel data pointer.
DirectX11 has the debug mode interface which does a similar thing, but can be bypassed once everything is working correctly. Those drivers don't just issue graphics commands, they must also compile and link shaders, run OpenCL, CUDA and interoperate with all the other API's that you can find at Khronos (www.khronos.org)
Slashdot Mirror
I would like to see the filesystem of an OS partitioned into several levels: read-only disk drives where stuff never changes unless an update occurs (kernel, device drivers, configuration files), read-write disks where log files are update by the minute, hour or day, and local/user partition which is updated by the user.
Our university managed to do something similar by just having a ISO image that they overwrite the OS partition with, every time the PC was rebooted.
There was a guy at our university wanting to do some university psychology tests and figured the best way for the application to log the results was to send them as an E-mail to himself, where they could be timestamped independently. Only problem was that any application that wasn't on the PC's anti-virus whitelist was blocked from opening that port. So he just renamed his experiment application to "Agent.exe" and the anti-virus software allowed the message to be sent.
I am sure it would be possible for someone to create an Opera or Firefox plugin that filters downloads of slashdot beta articles and converts them back to slashdot classic.
Until I saw what slashdot beta looked like, I thought they were trolls too. But now that I see that half the browser window is going to be blank just so that the poll
window can have more white space, I sympathize with them. It should be up to each slashdot reader to choose how they read slashdot. Other sites give their forum members a choice in font size, background and text colors, and theme.
Or it allows twice a many cases to conducted for half the price. The only limit on litigation in the past has been the expense of lawyers. In the UK, we have "ambulance chasers", "no win, no fee" lawyers who look for every opportunity to win a compensation payout.
Every drone could have it's own small radio transmitter. If it detected the signal of another drone (the inverse square distance law), it could adjust it's course automatically.
I've seen hardware engineers print out images of chip designs, but they weren't looking at track layouts, they were looking at transistor densities. The image looked like a a density map of a large American city, with empty space in white, low density in green, medium density in yellow, orange and red, and high density in gray and black. I saw what looked like a exclusive leafy-green street surrounded by green parks. Like a sadistic urban planner, the engineer said, "oh, those are multiplexers, don't worry, we'll get them to move, it's costing us money to have a block like that, we like everything to be high density".
Some of them most complex control systems for aircraft are on the order of five million lines of code plus; device drivers for every possible component and sensor, user interface systems for the pilot information system and some AI for the autopilot logic.
There are two places where I know that visual design tools are used; the first is QT Designer which is used to design application frames and dialog windows with radio buttons, sliders, text windows. However, this still requires that you connect all the input/output events between widgets manually. And those get really messy. That could be eliminated though, since if you add a push-button to a window, it is practically guaranteed that you want to know about the press/release event, so those event handlers could be added automatically. Another place is the composition editor of Blender where you can place and connect processing nodes to do image processing and lighting. Once again that still requires the data flow to be connected up manually.
While a visual programming environment could be extended to do mathematics, there would need to be some way of handling complexity - some image processing and geometry algorithms consist of 20 or 30 more different stages, with each stage consisting of a 100+ line algorithm. If this environment supported hierarchical designs like data-flow diagrams, it might work, but it would just become a complete mess otherwise.
I believe it is called a "stress test", one of the techniques that banks use to test the resilience of a trading system and how much liquidity it has. If they suddenly shutdown the exchange of currency, the punters are forced to make trades using other resources. How much of that other currency is used gives an idea of how is available. The Russians and EU did the same by restricting physical money currency transfers. The EU figured out there was black market the size of the GDP of the entire region.
There was the "Stop and Catch Fire Instruction" on 6809 chips. Designers decided to share the input/output ports using the same pin connects. These were selected between by a couple of bits. Set both bits at the same time and the silicon die went into meltdown. Some graphics card would do the same with a driver update.
I guess you could always have a compressed cylinder of nitrogen, release it suddenly onto the components to make them brittle and vibrate them at high frequency so they basically disintegrate into dust.
Well, at least having screws are better than having held together by glue, and the high-end models (PCG-GRT series) did have the hard disk drive on the edge of the laptop rather than some cage in the middle of the chassis.
I know what you mean about all those screws. Replacing the screen or even the cooling assembly would require several ice trays to put every different component. The best way to describe how Sony must have designed a laptop was to start off with a gaming PC with a GTX Titan GPU and motherboard, then slice up the motherboard until everything could fit inside the profile of a laptop.
Around 2004, the "high-end" models had Nvidia Geforce 5600Go's with about 64 Megabytes of memory, which was enough to run Bzflag. It was impressive to see a laptop do texture mapping, when previously, it would have taken a Silicon Graphics Extreme to do the same.
Sony did the same. But they usually made around several dozen different models around a single laptop chassis and priced each model separately, and had different sized screens, disk-drives and memory configurations and maybe an option TV tuner board. By far the LCD screen was the most expensive replacement, at around $1000 . Some companies actually accepted the old screen as a trade-in discount. So it was cheaper to buy the model with the highest resolution screen, but the lowest price, then replace the hard disk drive and memory. But the GPU's were always custom made. There was no way of upgrading as they had custom connectors.
I remember when Dos-Shell came out - it allowed you to view two directory listings side by side in a 120 x 50 text window split in half. That was considered revolutionary at the time.
The monitor and disk drives are the component most likely to break. The last thing I would want to do is to have to replace one simply because the other broke. The best design I can think of would be to go back to "podules" where the CPU/GPU/memory and disk drives were on separate blocks but slotted into the base of the monitor.
That happened in Edinburgh as well. The banks were offering starting salaries of £50K+ bonuses to those with the best qualifications. Those salaries exceeded what the professors with 20+ years experience and were leading research teams were earning. PhD students were lucky to get a £12K research stipend/year, and post-docs were lucky to get £30K/year.
Rent for an apartment anywhere central, on a bus-route or in leafy-green parts of the city were around £850, and food costs are £50/week per person.
In the UK, most such mechanical engineering companies pride themselves on having the goal of "turning their brightest engineers into managers".
The issue with Unity has come up many times - just about every forum article I had read about the Unity Lens was "How do I remove this feature?". These were everyone from sys-admins to scientific programmers and application developers who just wanted to do their work, and not find themselves browsing Amazon products, when trying to maintain their system.
The second issue which is more serious, involves network security. Just about every sys-admin will tell you that every TCP/IP port that is open to receive data is a potential security threat for network worms/viruses. Ubuntu loves to open the printer server (cups), a web server (apache2), network time protocol (ntp) and dhclient (dhcp), and mysteriously the control panel controls either don't work or say that the service has been closed when it hasn't, or that it is under control by some other OS process but that process doesn't support shutdown. Every time I have read articles about this, the developers seem to be fairly defensive, that "Well, it's your own fault if one day you try and connect a printer and you can't print". Then you'll find your web-browser (Opera) connects to Google (*1e100.net) and (backoo.canonical.com) when you haven't even visited another web page.
This attitude has led to the formation of a new version of Linux called "Mint".
I've been in that situation too. Head office management from 300 miles away had decided that every new programming recruit would cut their teeth on some back-end part of the application (and completed different from what they had been told what they were doing). The first result was that they had turned their workplaces into a coding gulag" where every programmer who had ever been put in that situation made "escape" their highest priority. Each and every one of those people had managed to escape by either leaving to set up their own company, move to a different part of the team or moving to another company.The second result was that this particular block of code ended up looking the results of one of the those drawing games where each person draws on part of a sheet of paper but never sees what anyone else has drawn. Consequently, it was never going to work.
The only option is "escape".
It happened in the past with telephone exchanges. They had some self-maintenance code built in such that if one exchange detected a malfunction of some sort (accounts balance fail to match, line quality not good enough), it would send a fault message and a shutdown notice to it's neighboring exchanges. But there was a little bug. The message first hop was correct as it sent the ID of the originating exchange, the message relayed second and later hops was wrong because it sent the ID of the current exchange. Thus when one exchange went down, it shut down all the other exchanges.
Your classic desktop is a menu bar at the top with pull-down menus, mini icons for each desktop window, and the classic trash bin and folders/files on the desktop area itself. Every desktop window has maximize/iconize/minimize icons, scrollbars and other icons built into the window frame. The menu-bar includes dialogs for configuring printers, disabling/enabling network devices, network traffic sent/received per device and protocol, CPU/GPU temperature controls and power-on/restart/suspend controls.
Just about every desktop developer wants to try and change this in order to either look super-modern or conform to mobile devices. Sometimes they'll add fancy graphical effects like transparency, blurred transparency, smooth scaling windows, fade ins/fade outs. Others will eliminate the idea of resizable window frames and just try and force you to toggle between application windows, or they create a scrolling map desktop where all the windows click to each others edge. Others decide that you don't real want to have control over those settings, so just get rid of those dialogs. Others get rid of the pull-down menus requiring multi-language support and just have little color icons.
Canonical tried to get rid of the menu bar and just give you a search-window like google's web-page (The Unity Lens). Only problem was that *everything* you searched for was actually sent back to their servers, and for your efforts, they would send back a list of CD's, DVD's and computer games that you might want to play. Even the existing network control dialogs didn't work or actually disable network server processes they were supposed to (eg. network time, cups, apache web server, dhclient).
I like that idea. You could have modified commands like 'ls' that would display little thumbnails of each image, and a 'more'/ 'cat' commands that would just dump each of those images at full size or scaled to fit the terminal.
In the early day of internet research, they wanted to distinguish between corporations there to make a profit, non-profit organisations, educational groups and the military. So they had ".com" = corporations/companies, ".org" = non-profit organisation, and ".edu" for the educational research groups, ".mil" for the military", and ".net" for the companies that managed the continent wide networks built from fibre-optics and satellite communications.
That gets extended to giving each country it's own domain ".uk" = UK, ".fr" = France, ".de" = Germany. It was a sense of achievement for any country to get their own top-level domain.
Now you can make your domain name look like a USENET discussion forum:
alt.fashion.goth.clothing
comp.languages.cobol.programmer.guru
Those smarter drives do insane things that having a pool of surplus disk blocks and having a virtual disk cylinder/sector map that can swap out old blocks that have become damaged and replace them with a new block. Just because you think you are writing on cylinder 32, sector 5, block 3, doesn't mean it's really at that location. Theoretically, it might be possible to fill up every possible block with data, but that's no guarantee.
So the only safe way is to destroy the hard disk drives.
Weirdo AMD stuff? The descriptors that DirectX and Mantle use are basically the registers that control your GPU. Traditional desktop GL has the large GL state (textures + framebuffer objects + blending + vertex buffers + transform feedback) and which is what the driver maintains and attempts to sanity check with all those GL calls (500+ of them) to make sure nothing inconsistent or fatal is allowed to run on any of the GPU cores. Setting a single parameter can involve cross-referencing dozens of other variables. If you look at some of the GL commands with the most parameters such as the texture image uploads, the driver has to sanity-checking ten or more parameters eg. the texture image functions have to check texture width, height, type, internal format, mip-map level, pixel data source type, pixel data pointer.
DirectX11 has the debug mode interface which does a similar thing, but can be bypassed once everything is working correctly. Those drivers don't just issue graphics commands, they must also compile and link shaders, run OpenCL, CUDA and interoperate with all the other API's that you can find at Khronos (www.khronos.org)