But the Xerox ID is not necessarily linked to the original document's origin so unless the copier extracts the dots from the original documents and somehow replicates them on the copy, the means to determine with reasonable proof the most likely origin through ID is lost.
If a compromising Bush document is printed on a White House colour laser printer then Xerox'd at a local pharmacy before mailing the copy or simply faxing, the exact origin is most likely lost.
Worst case, buy yellow transparent film if you do not want to take any chances. If the whole page is filtered to be yellow, it makes the equipment's job of picking it up much tougher, assuming said equipment actually tries to pick them up.
The reason why CDs are 70mins is to fit the whole of Beethoven's 7th or 9th symphony... I wonder if any download service is offering it as a single track at the regular track price. (Yeah, I know, they most likely give each movement its own track.)
Every combination of PC/Laptop+Windows ships by the thousands, Linux is isolated cases across the board.
One Windows image is probably usable across similar configurations... no need for images for every HDD size, CPU, display, etc. So Dell probably only has a dozen images of XP to accomodate IGPs, Radeon/GeForce and other major configuration discrepancies in their lineup such as XP Home/Pro/TabletPC/MediaCenter... but not all OS options are available for every model.
Well, this is only one way how Linux orders can get in the way of a major Windows OEM's workflow.
Drivers are rarely perfect and hardware often is not. At least, drivers can be fixed and be made to work around most non-critical hardware bugs.
Since most devices are targeted for Windows environments, it is only natural for Windows to be exposed to more questionable drivers than most other OSes.
They probably get all their HDDs pre-imaged with WinXP. This probably means the labour involved in specifically reconfiguring laptops for Linux use exceeds the OEM Windows' price.
(They have to unpack the laptop, wipe out the HDD, maybe swap the WiFi and WinModem, remove Windows stickers, install Linux then repackage, the whole process probably takes something like 2h.)
My guesses: 1- At startup, the CPU will check the BIOS image against a private key stored in the CPU. If this check is positive, the TC flag is set, otherwise it is cleared and the only way to set it is to flash a compilant BIOS and do a hard-reset. 2- The BIOS checks the boot image's certificate and clears the TC flag if the check is negative. 4- The boot image monitors the OS' boot sequence and clears the TC flag if any file/driver's certificate does not check out.
This way, until the root keys used to sign the BIOS or loader images is cracked, the TC flag gets cleared before the first non-compliant layer gains control, prohibiting the CPU from decrypting TC code. People still get to run non-secure BIOS/OS/Apps by forfeiting the ability to use TC/DRM stuff.
No need to have a BIOS option to turn TC off, it would turn itself off as soon as non-compliant components are encountered. (Unless you want it off on an otherwise compliant setup.)
I have no doubt that offshoring will eventually strenghten other currencies and increase per-capita income... but it will take a while (probably no less than 10 years but no more than 20) for this to nullify offshoring savings. Once China&all catch up with American/Japanese/European consumerism/capitalism, things will get tough for everyone.
I thought the USA culture was offshoring everything and cutting seats to cut costs and increase profits to make stock holders happier.
As for chip makers, Taiwan appears to be where everyone is headed... or would be if there was enough space to accomodate them. It's been a while since I last saw "Japan" printed on an IC. The majority come from either Taiwan, Malaysia or Korea... just like nearly everything else (once you add China) and only more so in the future.
With WEP, there were some major shortcuts thanks to implementation flaws and general weaknesses. WPA-AES has some weakness that could have been avoided but none that would allow for overnight cracking the way WEP does.
At the moment, WPA's only weakness is weak key phrase... but this has always been a weakness of any password-like thing. As long as a decently long and arbitrary key is used, cracking WPA will take weeks/months.
Unlike WEP where the key was hashed as-is and could be looked up in tables, WPA hashes the SSID with the key so any such table would end up SSID-specific and therefore not worth building... unless there are enough people who go through the trouble of setting up WPA yet leave the factory (or other common) SSID.
I think my SSID and key are arbitrary enough to belong in the tougher category. Given the number of open WAPs I have logged in a ~1km radius of where I live, freeloaders have many easier targets to pick from. (None appear to be in-range from my location though.)
After disabling SSID, using WPA-AES-PSK and MAC filter, there is not much left to do in terms of off-the-shelf measures other than changing the SSID and AES key every now and then.
This is the way I have setup my router and I am guessing it should be generally enough. Without SSID, wardrivers will have a harder time finding my network. With MAC filters, they will have to work a little more. With WPA-AES, they will have a hard time getting in and by the time they crack my key, I might have already changed it.
Try building an equivalent plant using solar panels...
1) The sun provides roughly 1kW/m^2 under ideal conditions 2) The most efficient solar cells have ~35% efficiency and cost over $5k per square meter (these are space-grade) 3) To produce 250MW, one would need 750 000 square meters of high-efficiency solar panels, at a cost exceeding $4.5 bilion. (Land + panels + infrastructure + other equipment + staff + maintenance + etc.) 4) A metropolitan area's power requirements are measured in GigaWatts so one or two solar plants like the above would make little to no difference other than consuming land... and if airborne residues are a problem, a solar plant would need full-time staff to clean the panels.
I never bought into "disposable" (1-year) drives... although the manufacturers did this while saying "drives destined to fail will fail in the first year", my prior experience says 100% of my dead drives failed between the 12th and 18th month.
I wonder if other manufacturers will start putting their warranty where their MTBF is... claiming a milion hours MTBF is meaningless when combined to a one or even three years warranty. And the "Component Design Life: 5 years" in many Maxtor (and others) docs seem to contradict their MTBF.
So... three years warranty, designed to last five years, 100 years MTBF. The first two make sense but the MTBF looks like a fantasy and false representation.
My 80GB WD drives peak around 60MB/s, my 120GB drive peaks around 65MB/s, all three being nearly two years old.
50MB/s is history, 70MB/s peak (outer cylinders) is probably common with current (2005) 7200RPM/3.5" drives... but I have not benched my two newest drives yet to verify that.
SATA-II is only the name of the new standards comittee for SATA devices. Devices are not guaranteed to support NCQ or 3Gbps unless the device manufacturer specifically says so.
As for why 3Gbps != 375MB/s, this is because 3Gbps is the wire speed but this wire uses 8B/10B code expansion to facilitate clock recovery and error detection, which is like having 10bits per byte. Add out-of-band signals, CRCs and the rest, the usable speed might drop to 250MB/s... full-duplex though.
RAM is cheap but HDD manufacturers ask for a relatively hefty premium for the privilege of this trivial upgrade: a 64Mbits (8MB) RAM chip costs around $2 while a 256Mbits (32MB) chip costs under $4... yet HDD manufacturers ask $10-$20 more for the 16MB version. For that kind of price difference, they could offer 32MB and still make some extra profit.
Then again, having huge on-board caches would increase the risk of losing data when power randomly goes out, which would seem like a valid reason to hold back on caches... let an auxiliary RAID/storage controller do extra caching for battery-backed-up external boxes.
3Gbps is the wire signaling speed. With transmissions using 8B/10B code expansion and out-of-band signals, the effective maximum speed will be around 220MB/s.
SATA supports that, it is just that no disk drive head/platter technology has that sort of bandwidth yet.
With SATA-attached RAID arrays and multiport expansion switches, maxing out 3Gbps with is very possible, just like it is possible to max out PATA's 133MB/s by using a pair of fast drives. With 3Gbps SATA, it will simply require four drives.
With a BSD license, the person would have to copy your car before doing anything with it. The person would never have to actually return it after doing tune-up and customizations on it while keeping the details to him/her-self.
With GPL, the person would still have to make a copy of the car before modifying/using it with the difference that any modifications have to be disclosed and the car shall not be used to tow non-GPL trailers nor use non-GPL wheels and accessories.
(Well, GPL can use BSDL code but not the other way around because BSDL allows people to fork the sources and go closed-source from there.)
And with outsourcing/offshoring, the proportion of non-technical/engineering/research employees at major corporations is increasing.
I guess the reduced likelyhood of downtime from viral infections using preconfigured off-the-shelf systems could justify the expense in a number of cases. Then again, they might simply be upgrading the graphists/designers' Macs and migrating the older ones to Mac-compatible seats... because 17-20% of employees use a Mac does not mean they have to be brand-new.
Intel has specifications for SATA PHY interfaces, search for "SAPIS". Most of the SATA PHYs comply to this interface, though all the implementations of this that I have found are in the form of IP cores.
SiliconImage had some (using their proprietary SATALite interface, press release here: http://www.siliconimage.com/news/press/detailpress release.aspx?id=134) but they disappeared from their product list earlier this year. Back then, the datasheets were only available on-demand. This particular chip was often used with older VIA chipsets and some PATA+SATA RAID boards.
I think Marvell might also have some but all their SATA products appear to be on-demand.
As for latency, it is only measurable when issuing read commands. In that case, the controller has to wait for the "slowest" drive and this is worst when using an array of disparate drives: active head positions and transfer speeds are uneven so data cannot be expected to be typically late by one rotation at most. I wonder how much of this an NCQ controller fitted with NCQ drives could hide.
With all the gigabit serial links being deployed now, maybe next-gen low-cost FPGAs will start featuring some MGTs... they will need at least PCIE-x1 compliance sooner than later but probably would not feature more than four MGTs for a while.
I wonder what the pricing on 4VFX20 is... it is the closest 90nm equivalent to the 130nm XC2VP12 and I would be curious to see how much above/below $200 it might be. (The fact that it is not yet available for open distribution screams hefty premium though.)
I was thinking about doing something of that sort... but the specs for SATA PHY chips (so low-cost FPGAs without fancy IOs can be used) are somewhat scarce and most FPGAs with multi-gigabit transceivers cost over $200USD. (The elderly Xilinx XC2VP7 with 8xMGT currently costs ~$200 while the XC3S1000 which has 60% more logic capacity without the fancy IOs and PPC core costs $50.)
A simple RAID solution can simply XOR the first two/four drive's data and store the result on the 3rd/5th drive. If any drive fails, XORing the remaining two/four drives will reproduce the failed drive's data so it can be returned or stored on the spare/replacement drive.
The next step after the XOR RAID5 and 5+1 (hot spare) is to use Reed-Solomon (or equivalent) coding. By using the two extra drives (XOR/parity and spare) to store RS codes instead, the array can survive up to two failures until the failed drives are changed and sync'd.
It would be a fun project if the SATA specs were more openly available and chips/PCBs more affordable. Well, I do not really have the spare change 5+ midrange SATA HDDs ATM anyway.
If they fight piracy by putting everyone in prison, how is that supposed to rake money? Putting someone in prison costs over $30k-$100k/year to the state, this quickly costs more in REAL money than the piracy's real (or even hypothetical) cost in lost sales.
And old-school piracy where people swapped disks or downloaded stuff from HTTP/FTP servers is somewhat far on the way out with all the well established P2P technologies that are replacing them today.
Next step, governments start requiring registration and annual audit of all computer equipment owned by a person to be sent along with the tax forms so additional taxes can be collected as necessary.
How much engineering (other than mechanical and architectural design) is done on Macs? In electrical engineering there are only four common choices for the software I have used or read about so far: x86/*nix, x86/Windows, Sparc/SunOS and PA-RISC/HP-UX, with some x86-64/*nix and Itanium/*nix. None of those I remember exist for PowerPC/*. I do not know much about the other engineering fields but at least for EE, PowerPC appears to be nowhere land, which hints that PPC are at least somewhat uncommon in engineering.
Many of the larger corporations are falling apart in part because their balance is heavily biased towards marketing. Macs historically fit pretty well with marketing (drawing, publishing, etc.) and I think the market share gains simply illustrate this increasing marketing-centric bias.
Yup... worst case, detonate as soon as capture becomes imminent, blowing up the security personnel with everything and everyone else within range. More security only means that somewhat fewer terrorists will manage to reach their targets but will only marginally reduce casualities.
If security is beefed up after each and every incident, it would not take long before public transportation is paralyzed... full searches before boarding a bus or entering a station wastes a fair amount of time and would be a major inconvenience. I hate driving but I definitely would get a car if this happened in my area.
Slashdot Mirror
But the Xerox ID is not necessarily linked to the original document's origin so unless the copier extracts the dots from the original documents and somehow replicates them on the copy, the means to determine with reasonable proof the most likely origin through ID is lost.
If a compromising Bush document is printed on a White House colour laser printer then Xerox'd at a local pharmacy before mailing the copy or simply faxing, the exact origin is most likely lost.
Worst case, buy yellow transparent film if you do not want to take any chances. If the whole page is filtered to be yellow, it makes the equipment's job of picking it up much tougher, assuming said equipment actually tries to pick them up.
The reason why CDs are 70mins is to fit the whole of Beethoven's 7th or 9th symphony... I wonder if any download service is offering it as a single track at the regular track price. (Yeah, I know, they most likely give each movement its own track.)
Every combination of PC/Laptop+Windows ships by the thousands, Linux is isolated cases across the board.
One Windows image is probably usable across similar configurations... no need for images for every HDD size, CPU, display, etc. So Dell probably only has a dozen images of XP to accomodate IGPs, Radeon/GeForce and other major configuration discrepancies in their lineup such as XP Home/Pro/TabletPC/MediaCenter... but not all OS options are available for every model.
Well, this is only one way how Linux orders can get in the way of a major Windows OEM's workflow.
I wonder how long a Mac using PC-style hardware and x86 CPUs will retain its factory hardware configuration.
I also wonder about what proportion of the general public will see the x86 Macs as simply overpriced mid-range PCs.
Well, this is all pointless as far as I am concerned since none of my major apps support OSX, x86 or otherwise.
Drivers are rarely perfect and hardware often is not. At least, drivers can be fixed and be made to work around most non-critical hardware bugs.
Since most devices are targeted for Windows environments, it is only natural for Windows to be exposed to more questionable drivers than most other OSes.
They probably get all their HDDs pre-imaged with WinXP. This probably means the labour involved in specifically reconfiguring laptops for Linux use exceeds the OEM Windows' price.
(They have to unpack the laptop, wipe out the HDD, maybe swap the WiFi and WinModem, remove Windows stickers, install Linux then repackage, the whole process probably takes something like 2h.)
My guesses:
1- At startup, the CPU will check the BIOS image against a private key stored in the CPU. If this check is positive, the TC flag is set, otherwise it is cleared and the only way to set it is to flash a compilant BIOS and do a hard-reset.
2- The BIOS checks the boot image's certificate and clears the TC flag if the check is negative.
4- The boot image monitors the OS' boot sequence and clears the TC flag if any file/driver's certificate does not check out.
This way, until the root keys used to sign the BIOS or loader images is cracked, the TC flag gets cleared before the first non-compliant layer gains control, prohibiting the CPU from decrypting TC code. People still get to run non-secure BIOS/OS/Apps by forfeiting the ability to use TC/DRM stuff.
No need to have a BIOS option to turn TC off, it would turn itself off as soon as non-compliant components are encountered. (Unless you want it off on an otherwise compliant setup.)
I have no doubt that offshoring will eventually strenghten other currencies and increase per-capita income... but it will take a while (probably no less than 10 years but no more than 20) for this to nullify offshoring savings. Once China&all catch up with American/Japanese/European consumerism/capitalism, things will get tough for everyone.
I thought the USA culture was offshoring everything and cutting seats to cut costs and increase profits to make stock holders happier.
As for chip makers, Taiwan appears to be where everyone is headed... or would be if there was enough space to accomodate them. It's been a while since I last saw "Japan" printed on an IC. The majority come from either Taiwan, Malaysia or Korea... just like nearly everything else (once you add China) and only more so in the future.
WEP yes... WPA-AES is a slightly different story.
With WEP, there were some major shortcuts thanks to implementation flaws and general weaknesses. WPA-AES has some weakness that could have been avoided but none that would allow for overnight cracking the way WEP does.
At the moment, WPA's only weakness is weak key phrase... but this has always been a weakness of any password-like thing. As long as a decently long and arbitrary key is used, cracking WPA will take weeks/months.
Unlike WEP where the key was hashed as-is and could be looked up in tables, WPA hashes the SSID with the key so any such table would end up SSID-specific and therefore not worth building... unless there are enough people who go through the trouble of setting up WPA yet leave the factory (or other common) SSID.
I think my SSID and key are arbitrary enough to belong in the tougher category. Given the number of open WAPs I have logged in a ~1km radius of where I live, freeloaders have many easier targets to pick from. (None appear to be in-range from my location though.)
After disabling SSID, using WPA-AES-PSK and MAC filter, there is not much left to do in terms of off-the-shelf measures other than changing the SSID and AES key every now and then.
This is the way I have setup my router and I am guessing it should be generally enough. Without SSID, wardrivers will have a harder time finding my network. With MAC filters, they will have to work a little more. With WPA-AES, they will have a hard time getting in and by the time they crack my key, I might have already changed it.
My quest for silence ended when I bought a laptop... practically silent from any distance while idle and ~2h of built-in backup power.
Try building an equivalent plant using solar panels...
1) The sun provides roughly 1kW/m^2 under ideal conditions
2) The most efficient solar cells have ~35% efficiency and cost over $5k per square meter (these are space-grade)
3) To produce 250MW, one would need 750 000 square meters of high-efficiency solar panels, at a cost exceeding $4.5 bilion. (Land + panels + infrastructure + other equipment + staff + maintenance + etc.)
4) A metropolitan area's power requirements are measured in GigaWatts so one or two solar plants like the above would make little to no difference other than consuming land... and if airborne residues are a problem, a solar plant would need full-time staff to clean the panels.
I never bought into "disposable" (1-year) drives... although the manufacturers did this while saying "drives destined to fail will fail in the first year", my prior experience says 100% of my dead drives failed between the 12th and 18th month.
I wonder if other manufacturers will start putting their warranty where their MTBF is... claiming a milion hours MTBF is meaningless when combined to a one or even three years warranty. And the "Component Design Life: 5 years" in many Maxtor (and others) docs seem to contradict their MTBF.
So... three years warranty, designed to last five years, 100 years MTBF. The first two make sense but the MTBF looks like a fantasy and false representation.
My 80GB WD drives peak around 60MB/s, my 120GB drive peaks around 65MB/s, all three being nearly two years old.
50MB/s is history, 70MB/s peak (outer cylinders) is probably common with current (2005) 7200RPM/3.5" drives... but I have not benched my two newest drives yet to verify that.
SATA-II != 3Gbps.
SATA-II is only the name of the new standards comittee for SATA devices. Devices are not guaranteed to support NCQ or 3Gbps unless the device manufacturer specifically says so.
As for why 3Gbps != 375MB/s, this is because 3Gbps is the wire speed but this wire uses 8B/10B code expansion to facilitate clock recovery and error detection, which is like having 10bits per byte. Add out-of-band signals, CRCs and the rest, the usable speed might drop to 250MB/s... full-duplex though.
RAM is cheap but HDD manufacturers ask for a relatively hefty premium for the privilege of this trivial upgrade: a 64Mbits (8MB) RAM chip costs around $2 while a 256Mbits (32MB) chip costs under $4... yet HDD manufacturers ask $10-$20 more for the 16MB version. For that kind of price difference, they could offer 32MB and still make some extra profit.
Then again, having huge on-board caches would increase the risk of losing data when power randomly goes out, which would seem like a valid reason to hold back on caches... let an auxiliary RAID/storage controller do extra caching for battery-backed-up external boxes.
3Gbps is the wire signaling speed. With transmissions using 8B/10B code expansion and out-of-band signals, the effective maximum speed will be around 220MB/s.
SATA supports that, it is just that no disk drive head/platter technology has that sort of bandwidth yet.
With SATA-attached RAID arrays and multiport expansion switches, maxing out 3Gbps with is very possible, just like it is possible to max out PATA's 133MB/s by using a pair of fast drives. With 3Gbps SATA, it will simply require four drives.
With a BSD license, the person would have to copy your car before doing anything with it. The person would never have to actually return it after doing tune-up and customizations on it while keeping the details to him/her-self.
With GPL, the person would still have to make a copy of the car before modifying/using it with the difference that any modifications have to be disclosed and the car shall not be used to tow non-GPL trailers nor use non-GPL wheels and accessories.
(Well, GPL can use BSDL code but not the other way around because BSDL allows people to fork the sources and go closed-source from there.)
And with outsourcing/offshoring, the proportion of non-technical/engineering/research employees at major corporations is increasing.
I guess the reduced likelyhood of downtime from viral infections using preconfigured off-the-shelf systems could justify the expense in a number of cases. Then again, they might simply be upgrading the graphists/designers' Macs and migrating the older ones to Mac-compatible seats... because 17-20% of employees use a Mac does not mean they have to be brand-new.
Intel has specifications for SATA PHY interfaces, search for "SAPIS". Most of the SATA PHYs comply to this interface, though all the implementations of this that I have found are in the form of IP cores.
s release.aspx?id=134) but they disappeared from their product list earlier this year. Back then, the datasheets were only available on-demand. This particular chip was often used with older VIA chipsets and some PATA+SATA RAID boards.
SiliconImage had some (using their proprietary SATALite interface, press release here: http://www.siliconimage.com/news/press/detailpres
I think Marvell might also have some but all their SATA products appear to be on-demand.
As for latency, it is only measurable when issuing read commands. In that case, the controller has to wait for the "slowest" drive and this is worst when using an array of disparate drives: active head positions and transfer speeds are uneven so data cannot be expected to be typically late by one rotation at most. I wonder how much of this an NCQ controller fitted with NCQ drives could hide.
With all the gigabit serial links being deployed now, maybe next-gen low-cost FPGAs will start featuring some MGTs... they will need at least PCIE-x1 compliance sooner than later but probably would not feature more than four MGTs for a while.
I wonder what the pricing on 4VFX20 is... it is the closest 90nm equivalent to the 130nm XC2VP12 and I would be curious to see how much above/below $200 it might be. (The fact that it is not yet available for open distribution screams hefty premium though.)
I was thinking about doing something of that sort... but the specs for SATA PHY chips (so low-cost FPGAs without fancy IOs can be used) are somewhat scarce and most FPGAs with multi-gigabit transceivers cost over $200USD. (The elderly Xilinx XC2VP7 with 8xMGT currently costs ~$200 while the XC3S1000 which has 60% more logic capacity without the fancy IOs and PPC core costs $50.)
A simple RAID solution can simply XOR the first two/four drive's data and store the result on the 3rd/5th drive. If any drive fails, XORing the remaining two/four drives will reproduce the failed drive's data so it can be returned or stored on the spare/replacement drive.
The next step after the XOR RAID5 and 5+1 (hot spare) is to use Reed-Solomon (or equivalent) coding. By using the two extra drives (XOR/parity and spare) to store RS codes instead, the array can survive up to two failures until the failed drives are changed and sync'd.
It would be a fun project if the SATA specs were more openly available and chips/PCBs more affordable. Well, I do not really have the spare change 5+ midrange SATA HDDs ATM anyway.
If they fight piracy by putting everyone in prison, how is that supposed to rake money? Putting someone in prison costs over $30k-$100k/year to the state, this quickly costs more in REAL money than the piracy's real (or even hypothetical) cost in lost sales.
And old-school piracy where people swapped disks or downloaded stuff from HTTP/FTP servers is somewhat far on the way out with all the well established P2P technologies that are replacing them today.
Next step, governments start requiring registration and annual audit of all computer equipment owned by a person to be sent along with the tax forms so additional taxes can be collected as necessary.
How much engineering (other than mechanical and architectural design) is done on Macs? In electrical engineering there are only four common choices for the software I have used or read about so far: x86/*nix, x86/Windows, Sparc/SunOS and PA-RISC/HP-UX, with some x86-64/*nix and Itanium/*nix. None of those I remember exist for PowerPC/*. I do not know much about the other engineering fields but at least for EE, PowerPC appears to be nowhere land, which hints that PPC are at least somewhat uncommon in engineering.
Many of the larger corporations are falling apart in part because their balance is heavily biased towards marketing. Macs historically fit pretty well with marketing (drawing, publishing, etc.) and I think the market share gains simply illustrate this increasing marketing-centric bias.
Yup... worst case, detonate as soon as capture becomes imminent, blowing up the security personnel with everything and everyone else within range. More security only means that somewhat fewer terrorists will manage to reach their targets but will only marginally reduce casualities.
If security is beefed up after each and every incident, it would not take long before public transportation is paralyzed... full searches before boarding a bus or entering a station wastes a fair amount of time and would be a major inconvenience. I hate driving but I definitely would get a car if this happened in my area.