Seriously, I know a guy who moonlights for the pentagon, he's one of the inventors of 'digital video storage', AFAIK.
Every few weeks a few burly men in a caravan of Ford SUVs pulls up and he's off for a few days. He just bought a house in southern france, 9/11 didn't hurt HIS economy:-)
Seriously though, great guy, It was an absolute joy to live next door to him.
I was in a dorm at Northeastern that was 10base-t hubbed, it was a total nightmare. There was so much packet loss on the LAN that NOBODY could get anything over the internet connection, web pages would take several minutes to show and most of the pictures would be 'broken.'
There's no question that everything MUST be switched in this day and age, upgrading to switches in the dorms might not cost that much, the labor is already paid for, right?
Leave the Cat3 in the walls and replace the 10bt hubs with new Cisco switches, turn the switches down to 10bt/full-duplex and make sure the uplink on the switch to the router/other switch is running at 100/full or better.
No attempt at managing the internet traffic will work until you fix the broken LAN traffic, if you can't get packets to the filter, they won't be filtered, they'll just soak up LAN resources and be dropped. Get the reliability and speed of the traffic from those dorm machines to the traffic shaper as good as you can and work from there.
On a side note, a lot of admins are under the belief that a group of clients on a 10bt/half hub will cause less load because they can't load the server-room as much as switched clients can, but they are wrong, a 'lossy' LAN is the worst enemy of the server room, because much of the work that the routers and servers push out gets dropped and must be retransmitted over and over again. Also, a faster LAN lets the servers and routers 'satisfy' client requests in much less time, so there are fewer open connections and service denials. I've seen client-end infrastructure upgrades that make our server hits go up but 'average simultaneous http connections' drop like a rock. Also, the stream of calls from students who couldn't get to the intranet disappeared.
Alright, an LED light will last through the lifetimes of several flourescents, produce MUCH less heat (though both kick incandescents' ass), and the total cost of ownership is probably lower since they use less juice and need less replacing.
I feel the same way, my rig (file server, network equipment, two PCs, and monitors) puts out about 1000 watts of heat, or 2/3 of a space heater. When I moved in our electric:gas bills went from $15:$50 to $50:$20, It only costs me $5 more to run all my computers in the winter because the heater kicks in a lot less. During the summer it costs me about $35/monthe to run the same rig. It would cost a LOT more to run in the summer if I had air conditioning, as I would be paying for energy to HEAT and COOL at once.
One thing I'm looking to do is build a 'chimney' for the cabinets I put all the computers in, one that I could have blow heat into the house during the winter and out a vent in the summer, all while keeping my PCs a bit cooler. Note that if the PCs are in a hot cabinet during the winter, it makes no difference to the temperature of the house if I blow that heat out or if it stays in the cabinet, as all that heat eventually finds it's way into the house anyway, it just gets there slower through the wood.
right, but in terms of heating the ROOM around the PC, all cooling solutions are the same, except for the small difference that active cooling ADDS to the wattage of the system.
A box with a regular heatsink/fan will actually generate less heat than one with a peltier cooler, because the peltier is less efficient, even if it gets your CPU down to near-freezing temperatures.
Hmm, at my work (big bank) 8 people getting paid $25,000/year cover 13 buildings with over 5000 desktops.
Then again, we run virus-protection at the gateways, and the whole LAN is proxied-out, and we can push software updates to the users with ZenWorks instead of running out to machines manually.
If you're paying techs $70K+/year, they should have deployed patches and virus updates WAY before anything broke out. Seriously, I could script up something to do that in 15 minutes, and I work for $19,500/year!
But you forget that most drug dealers don't really make much money at the trade. The profit margins are very low because there are so many millemen, often when purchasing quantities of one to five pounds of pot, you're the 10th person handling it. The marijuana business doesn't really generate that much profit for any one player because it's VERY distributed and there are a lot of people 'just doing it to pay rent' or 'just to smoke free'. I recall one summer when I knew people flipping two pounds a week, and their total profit was only $600/week, and they had to split that three ways, they'd be better off flipping burgers.
Some drugs have bigger margins, like heroin and coke, but that's because it's a lot harder to source, users are willing to front more money to someone they trust, and it relies more on 'trust' relationships; not to mention addictive qualities of the drugs themselves.
Overall I think we ought to legalize and regulate most drugs because the idea that they cause crime is self-fulfilling. The only crime I've seen related to drugs (several shootings and murders, MANY beatings, a lot of theft) is because someone either RIPPED someone off, or someone got busted and couldn't hold up their end (resulting in retribution). If the trade were moved off the street and into the storefront, there would be a lot more order to the whole thing. Sure, maybe there would be a small rise in holdups, but there'd be a HUGE drop in street crime.
all in all it doesn't put more heat into the room than any other case. There's conservation of energy here, if you take two PCs of equal footing, one in a CoolerMaster 6-fan case and one in this fanless thing the fanned case will actually heat up the room MORE, because the fans themselves generate (some) heat. I think overall you can measure the amount of heat that a device will produce by measuring how much it draws from the power source, sinc ein the end all that juice is being turned to heat anyway.
Because RedHat and other OSS is 'mostly secure' and they find issues and fix them right away, with 8000 software packages (and many building on each other) it's a lot of patches. Windows is basically starting from a 'mostly unsecure' point and they're keeping all the massive vulnerabilities secret, they release a fix when an exploit is discovered or they feel like it.
Thhink about the dedication someone writing an OSS software package has, when they discover a hole in their code they sit down and fix it. When you discover a vulnerability in Windows (as an MS employee) you have all these levels of management and PR who's interests are in covering it up, especially if it could break APIs or programs.
The fact that there is a greater NUMBER of patches for OSS and other GNU-ish systems is a sign of 'security through dedication' while on the MS side its 'security through management'.
Remember that if MS really started fixing their code there would be patches EVERY DAY.
I just go in circles a few times between kernel rebuilds and software reinstalls until it works, then I sacrifice an old hard disk to the gods of computing in hope they will not make me do it again.
In all honesty, once I upgraded to CardBus cards it was much easier, they just appeared as PCI devices. I don't need to pop them in and out much, so one they worked, I just left them in there.
i386 and 680x0 still supported! [bashes own skull]
on
GCC 3.3.1 Released
·
· Score: 1
As far as I can tell you can still use the latest GCC for your Mac, you just have to be running a recent version of NetBSD or Linux, not Libc1-based or a.out-formatted systems. Read the full line of what's being obsolesced, the i386 is only unsupported in Windows NT 3 now, if you care to read the entire line.
You'll be fine if you use Debian 3.0, or NetBSD-current. AFAIK, you could also bootstrap NetBSD-stable with ELF support.
Now what would be really swell is a DFA-description for the 680x0 series, shouldn't be too hard to whip up for someone who works withthem, right? It might make supporting these CPUs a lot easier in the future.
Don't worry, the 680x0 is old and slow, but it's part of the coldfire series, which is still in production and use in embedded appliances, networking, and scientific equipment where there's interest and demand.
Linux is going to start the march on the server end. Only after you build confidence using Linux for things like internal DNS and DHCP will management let you roll it out to things like file serving and domain control. After it's clear that Linux is reliable and secure you can push for Linux on the desktop within a small test group.
Right now (here on the east coast, at least) most managers and IT people will laugh you out of the room if you mention Linux seriously. Hell, most places I won' even mention that I 'do' Linux because people automatically think you're a neer-do-well or a commie, not to mention that those in IT who DO know are scared SHITLESS that their days are numbered.
A huge portion of the IT department where I work (a big bank) don't know ANYTHING about linux other than what they've read in 'Information Week'. I had a server admin ask me last week if she can 'run version 8 of Linux on Windows XP', this lady earns three times what I do as a server admin and all she knows is how to 'end task' and reboot, there's no chance an army of that kind of person is going to want to accept a new player on the network, she'd smell her job evaporating.
Seriously, I left college because I'm a learner. I got absolutely shunned for participating in class, everyone thought I was trying to make them look bad. I just felt that if the professor ASKS the class for a response we should offer one. Often in my economics class of 75 students there was NOBODY willing to raise their hand and give an answer. It had the dual-effect of totally disheartening the professor (whom I often chatted with after class) and making me look like some sort of smart-aleck.
I went to a small private school for most of my school-career (because I was nothing but trouble in public schools), and I was raised to believe that education is an interactive full-duplex operation. I was so depressed when I finally got to college and found out that it was just a big version of high-school with beer and laundry added, I thought there's be more to it.
I'm with you, while X isn't the simplest thing one could think of, it really does perform amazingly well. The frame buffer as a 'performance solution' is a total dead-end as writing a stream of pixels to a buffer is a LOT slower than using X to draw complete objects.
I do think X is a bit creaky though, maybe it is time to start a new one, one where major (and even compatability-breaking) changes can happen. Some things on my wishlist:
*A single, standard, simple font system.
*Integration of a more modern toolkit and WM, even if it has to borrow heavily from GTK+ or another project. This would be inclusive, it wouldn't prevent you from using other toolkits and WMs (think WindowMaker instead of TWM in the base set).
*Ability to run like Quartz Extreme (as an OpenGL-based system). Also, not as a requirement, just as an option.
*There's no excuse for not vectorizing this from the bottom-up, and we'll be thankful when the commercial OSs get this done and we've already got it. Think about running your monitor at 1600X1200 and telling the system it's 200 DPI so it zooms everything accordingly. Apple has this up their sleeve now, and Longhorn might unleash it on Windows.
*Transparency, which personally doesn't get me hot and bothered, but I guess people think it's cool.
*Ability to act as the 'console' layer for the OS, no more framebuffer-for-console, X for graphical. Have the thing run a full-screen native terminal, and have the OS work with it.
*extensive database of video cards and monitors for easier configuration, this should be integral to the graphics system. It took me a LONG time to find the specs on some of my monitors and I'd rather not do it ever again.
*Generally simpler/more elegant design. I'm pretty sure that a lot of what's in XFree86 today is there just to prop itself up, while a newer system might have a better chance of coming out with a clean design.
Why not just run Y and have an X-server program built for it for 'legacy' apps? You could also write a 'blank' window and use a foreign toolkit in a more 'raw' way. This isn't my field, so I might be wildly off-base.
alright, so it sounds like all the disk-end optimizations are tkaen care of. I am very aware that the drive has no idea how to intelligently prefetch data inside filesystems. My other thinking was much more aggressive CPU-end prefetch and caching, not this FIFO dumb-cache of most modern OSs. I want Linux to totally prefetch folder/file indexes that have been accessed most (up to 2/3 max disk cache size) in the background when it fires up. I wrote some scripts to recursively 'ls -l' my most commonly used folders and 'cat >/dev/null' my most commonly used files, and starting up apps is VERY much improved, even 'login' presents its 'password:' faster now, which was a major annoyance when I first booted the laptop.
You have dead pixels on the eMac? THE ONLY CRT-BASED COMPUTER APPLE SELLS? Smells like a faker to me.
The repair bill for a 3-week old laptop would be $0 unless your boss broke it himself or he took it to a shop that isn't really working with Apple (or is cheating Apple).
It's also a miefield that can be taken out by a sharpshooter from 1KM away. All you'd have to do is shoot these things. I'm pretty sure they're not invisible.
I think custom-camouflage would be good for this sort of thing. Just drop a bunch of these in individualized 'rock' shells that blend in in Afghanistan, etc.
true, 16MB is only half a second of full-blast transfer, but the VAST majority of desktop disk-access is random small accesses, 32K here and 192K there sort of stuff. A better prefetching mechanism would help greatly for typical use, maybe have the drive read the entire track and put it into the buffer (and move the head to the next track) so the next access doesn't have to wait for data on said track to 'line up' again.
As for using the system's main memory, I'm all for it, someone needs to figure out how to have to OS profile how 'hot' sections of the filesystem are and prefetch all the indexes, directory info, and metadata into a smarter kind of disk cache.
Right now if I load Mozilla, 'dd' a 250MB file and re-load mozilla it doesn't hold moz in the cache. My OS should know that Moz is very 'hot' with several loads/day and it should keep all the metadata in the cache. I was also thinking that if we got really fancy, it could prefetch all the data in/usr/lib/mozilla and write a raw 'cache set' to swappable memory, loading a pre-organized directory of files from semi-swapped memory is probably faster than recursing the directories again and again, right?
As for battery-backed persistent caches, it could be made easier, I think, with a capacitor on the drive and some leftover inaccessible sectors at the end of the drive. When power cuts out the head drops back to the 'recover area' and starts dumping whatever's in the write buffer to the disk, when power is restored the first thing the drive does is check the 'recover area' for any 'dumped data', which it commits to disk (this time where it belongs) before your BIOS can even query the drive for booting. The capacitor would just buy you enough time to do that dump, and the drive could remain unplugged indefinitely without the risk of losing that data, unlike a battery-backed cache.
I'm sorry, I meant to say that it was for low-end CONSUMER-level perhipherals. Nobody who actually needs to get work done sticks with USB for long, it's flaky and slow.
Simulating nuclear weapons is a heck of a lot better than finding out 'the old way'. If we're going to HAVE nukes (and there's no turning back now, I'm afraid) we may as well have the right ones for the job. I'd rather have them develop 'right-size' weapons for the post-cold war era than use giant 1970's warheads on third-world dictators. I find it foolish to demand reductions in 'collateral damage' while simultaneously demanding limits on harmless weapons simulations aimed at reducing said damage.
I'd rather have 20 very accurate, very small nukes in our arsenal than the thousands of ICBMs we have now, it would be just as effective a deterrent; getting there means money and (simulated) testing.
Slashdot Mirror
As a guy who's read the constitution of this country , I'm shocked that executives like this aren't assasinated after saying this sort of thing.
With millions of folks using P2P, you'd think someone would have the balls to remove jerks like this from society.
Seriously, I know a guy who moonlights for the pentagon, he's one of the inventors of 'digital video storage', AFAIK.
:-)
Every few weeks a few burly men in a caravan of Ford SUVs pulls up and he's off for a few days. He just bought a house in southern france, 9/11 didn't hurt HIS economy
Seriously though, great guy, It was an absolute joy to live next door to him.
I was in a dorm at Northeastern that was 10base-t hubbed, it was a total nightmare. There was so much packet loss on the LAN that NOBODY could get anything over the internet connection, web pages would take several minutes to show and most of the pictures would be 'broken.'
There's no question that everything MUST be switched in this day and age, upgrading to switches in the dorms might not cost that much, the labor is already paid for, right?
Leave the Cat3 in the walls and replace the 10bt hubs with new Cisco switches, turn the switches down to 10bt/full-duplex and make sure the uplink on the switch to the router/other switch is running at 100/full or better.
No attempt at managing the internet traffic will work until you fix the broken LAN traffic, if you can't get packets to the filter, they won't be filtered, they'll just soak up LAN resources and be dropped. Get the reliability and speed of the traffic from those dorm machines to the traffic shaper as good as you can and work from there.
On a side note, a lot of admins are under the belief that a group of clients on a 10bt/half hub will cause less load because they can't load the server-room as much as switched clients can, but they are wrong, a 'lossy' LAN is the worst enemy of the server room, because much of the work that the routers and servers push out gets dropped and must be retransmitted over and over again. Also, a faster LAN lets the servers and routers 'satisfy' client requests in much less time, so there are fewer open connections and service denials. I've seen client-end infrastructure upgrades that make our server hits go up but 'average simultaneous http connections' drop like a rock. Also, the stream of calls from students who couldn't get to the intranet disappeared.
Alright, an LED light will last through the lifetimes of several flourescents, produce MUCH less heat (though both kick incandescents' ass), and the total cost of ownership is probably lower since they use less juice and need less replacing.
I feel the same way, my rig (file server, network equipment, two PCs, and monitors) puts out about 1000 watts of heat, or 2/3 of a space heater. When I moved in our electric:gas bills went from $15:$50 to $50:$20, It only costs me $5 more to run all my computers in the winter because the heater kicks in a lot less. During the summer it costs me about $35/monthe to run the same rig. It would cost a LOT more to run in the summer if I had air conditioning, as I would be paying for energy to HEAT and COOL at once.
One thing I'm looking to do is build a 'chimney' for the cabinets I put all the computers in, one that I could have blow heat into the house during the winter and out a vent in the summer, all while keeping my PCs a bit cooler. Note that if the PCs are in a hot cabinet during the winter, it makes no difference to the temperature of the house if I blow that heat out or if it stays in the cabinet, as all that heat eventually finds it's way into the house anyway, it just gets there slower through the wood.
right, but in terms of heating the ROOM around the PC, all cooling solutions are the same, except for the small difference that active cooling ADDS to the wattage of the system.
A box with a regular heatsink/fan will actually generate less heat than one with a peltier cooler, because the peltier is less efficient, even if it gets your CPU down to near-freezing temperatures.
Hmm, at my work (big bank) 8 people getting paid $25,000/year cover 13 buildings with over 5000 desktops.
Then again, we run virus-protection at the gateways, and the whole LAN is proxied-out, and we can push software updates to the users with ZenWorks instead of running out to machines manually.
If you're paying techs $70K+/year, they should have deployed patches and virus updates WAY before anything broke out. Seriously, I could script up something to do that in 15 minutes, and I work for $19,500/year!
Wow! I haven't seen a PWEI quote EVER, I thought I was the only person in the country who'd ever heard them :-)
Good choice of quotes though. I had that line as my AIM profile for quite some time last year.
Trampled underfoot by the rise of the right
Right, how could I have forgotten? :-)
But technically, wouldn't the photons from the LEDs get converted to heat energy as they travel through the air and/or hit the walls?
It's not 100% accurate, but the draw of your devices in watts is roughly how much heat they will pump into the room, right?
But you forget that most drug dealers don't really make much money at the trade. The profit margins are very low because there are so many millemen, often when purchasing quantities of one to five pounds of pot, you're the 10th person handling it. The marijuana business doesn't really generate that much profit for any one player because it's VERY distributed and there are a lot of people 'just doing it to pay rent' or 'just to smoke free'. I recall one summer when I knew people flipping two pounds a week, and their total profit was only $600/week, and they had to split that three ways, they'd be better off flipping burgers.
Some drugs have bigger margins, like heroin and coke, but that's because it's a lot harder to source, users are willing to front more money to someone they trust, and it relies more on 'trust' relationships; not to mention addictive qualities of the drugs themselves.
Overall I think we ought to legalize and regulate most drugs because the idea that they cause crime is self-fulfilling. The only crime I've seen related to drugs (several shootings and murders, MANY beatings, a lot of theft) is because someone either RIPPED someone off, or someone got busted and couldn't hold up their end (resulting in retribution). If the trade were moved off the street and into the storefront, there would be a lot more order to the whole thing. Sure, maybe there would be a small rise in holdups, but there'd be a HUGE drop in street crime.
all in all it doesn't put more heat into the room than any other case. There's conservation of energy here, if you take two PCs of equal footing, one in a CoolerMaster 6-fan case and one in this fanless thing the fanned case will actually heat up the room MORE, because the fans themselves generate (some) heat. I think overall you can measure the amount of heat that a device will produce by measuring how much it draws from the power source, sinc ein the end all that juice is being turned to heat anyway.
Because RedHat and other OSS is 'mostly secure' and they find issues and fix them right away, with 8000 software packages (and many building on each other) it's a lot of patches. Windows is basically starting from a 'mostly unsecure' point and they're keeping all the massive vulnerabilities secret, they release a fix when an exploit is discovered or they feel like it.
Thhink about the dedication someone writing an OSS software package has, when they discover a hole in their code they sit down and fix it. When you discover a vulnerability in Windows (as an MS employee) you have all these levels of management and PR who's interests are in covering it up, especially if it could break APIs or programs.
The fact that there is a greater NUMBER of patches for OSS and other GNU-ish systems is a sign of 'security through dedication' while on the MS side its 'security through management'.
Remember that if MS really started fixing their code there would be patches EVERY DAY.
I just go in circles a few times between kernel rebuilds and software reinstalls until it works, then I sacrifice an old hard disk to the gods of computing in hope they will not make me do it again.
In all honesty, once I upgraded to CardBus cards it was much easier, they just appeared as PCI devices. I don't need to pop them in and out much, so one they worked, I just left them in there.
As far as I can tell you can still use the latest GCC for your Mac, you just have to be running a recent version of NetBSD or Linux, not Libc1-based or a.out-formatted systems. Read the full line of what's being obsolesced, the i386 is only unsupported in Windows NT 3 now, if you care to read the entire line.
You'll be fine if you use Debian 3.0, or NetBSD-current. AFAIK, you could also bootstrap NetBSD-stable with ELF support.
Now what would be really swell is a DFA-description for the 680x0 series, shouldn't be too hard to whip up for someone who works withthem, right? It might make supporting these CPUs a lot easier in the future.
Don't worry, the 680x0 is old and slow, but it's part of the coldfire series, which is still in production and use in embedded appliances, networking, and scientific equipment where there's interest and demand.
Linux is going to start the march on the server end. Only after you build confidence using Linux for things like internal DNS and DHCP will management let you roll it out to things like file serving and domain control. After it's clear that Linux is reliable and secure you can push for Linux on the desktop within a small test group.
Right now (here on the east coast, at least) most managers and IT people will laugh you out of the room if you mention Linux seriously. Hell, most places I won' even mention that I 'do' Linux because people automatically think you're a neer-do-well or a commie, not to mention that those in IT who DO know are scared SHITLESS that their days are numbered.
A huge portion of the IT department where I work (a big bank) don't know ANYTHING about linux other than what they've read in 'Information Week'. I had a server admin ask me last week if she can 'run version 8 of Linux on Windows XP', this lady earns three times what I do as a server admin and all she knows is how to 'end task' and reboot, there's no chance an army of that kind of person is going to want to accept a new player on the network, she'd smell her job evaporating.
Seriously, I left college because I'm a learner. I got absolutely shunned for participating in class, everyone thought I was trying to make them look bad. I just felt that if the professor ASKS the class for a response we should offer one. Often in my economics class of 75 students there was NOBODY willing to raise their hand and give an answer. It had the dual-effect of totally disheartening the professor (whom I often chatted with after class) and making me look like some sort of smart-aleck.
I went to a small private school for most of my school-career (because I was nothing but trouble in public schools), and I was raised to believe that education is an interactive full-duplex operation. I was so depressed when I finally got to college and found out that it was just a big version of high-school with beer and laundry added, I thought there's be more to it.
I'm with you, while X isn't the simplest thing one could think of, it really does perform amazingly well. The frame buffer as a 'performance solution' is a total dead-end as writing a stream of pixels to a buffer is a LOT slower than using X to draw complete objects.
I do think X is a bit creaky though, maybe it is time to start a new one, one where major (and even compatability-breaking) changes can happen. Some things on my wishlist:
*A single, standard, simple font system.
*Integration of a more modern toolkit and WM, even if it has to borrow heavily from GTK+ or another project. This would be inclusive, it wouldn't prevent you from using other toolkits and WMs (think WindowMaker instead of TWM in the base set).
*Ability to run like Quartz Extreme (as an OpenGL-based system). Also, not as a requirement, just as an option.
*There's no excuse for not vectorizing this from the bottom-up, and we'll be thankful when the commercial OSs get this done and we've already got it. Think about running your monitor at 1600X1200 and telling the system it's 200 DPI so it zooms everything accordingly. Apple has this up their sleeve now, and Longhorn might unleash it on Windows.
*Transparency, which personally doesn't get me hot and bothered, but I guess people think it's cool.
*Ability to act as the 'console' layer for the OS, no more framebuffer-for-console, X for graphical. Have the thing run a full-screen native terminal, and have the OS work with it.
*extensive database of video cards and monitors for easier configuration, this should be integral to the graphics system. It took me a LONG time to find the specs on some of my monitors and I'd rather not do it ever again.
*Generally simpler/more elegant design. I'm pretty sure that a lot of what's in XFree86 today is there just to prop itself up, while a newer system might have a better chance of coming out with a clean design.
Why not just run Y and have an X-server program built for it for 'legacy' apps? You could also write a 'blank' window and use a foreign toolkit in a more 'raw' way. This isn't my field, so I might be wildly off-base.
alright, so it sounds like all the disk-end optimizations are tkaen care of. I am very aware that the drive has no idea how to intelligently prefetch data inside filesystems. My other thinking was much more aggressive CPU-end prefetch and caching, not this FIFO dumb-cache of most modern OSs. I want Linux to totally prefetch folder/file indexes that have been accessed most (up to 2/3 max disk cache size) in the background when it fires up. I wrote some scripts to recursively 'ls -l' my most commonly used folders and 'cat > /dev/null' my most commonly used files, and starting up apps is VERY much improved, even 'login' presents its 'password:' faster now, which was a major annoyance when I first booted the laptop.
You have dead pixels on the eMac? THE ONLY CRT-BASED COMPUTER APPLE SELLS? Smells like a faker to me.
The repair bill for a 3-week old laptop would be $0 unless your boss broke it himself or he took it to a shop that isn't really working with Apple (or is cheating Apple).
It's also a miefield that can be taken out by a sharpshooter from 1KM away. All you'd have to do is shoot these things. I'm pretty sure they're not invisible.
I think custom-camouflage would be good for this sort of thing. Just drop a bunch of these in individualized 'rock' shells that blend in in Afghanistan, etc.
prefetching is by nature read-only, so we can't get hurt there, writing while losing power is an easy fix that nobody's implemented. Really.
true, 16MB is only half a second of full-blast transfer, but the VAST majority of desktop disk-access is random small accesses, 32K here and 192K there sort of stuff. A better prefetching mechanism would help greatly for typical use, maybe have the drive read the entire track and put it into the buffer (and move the head to the next track) so the next access doesn't have to wait for data on said track to 'line up' again.
/usr/lib/mozilla and write a raw 'cache set' to swappable memory, loading a pre-organized directory of files from semi-swapped memory is probably faster than recursing the directories again and again, right?
As for using the system's main memory, I'm all for it, someone needs to figure out how to have to OS profile how 'hot' sections of the filesystem are and prefetch all the indexes, directory info, and metadata into a smarter kind of disk cache.
Right now if I load Mozilla, 'dd' a 250MB file and re-load mozilla it doesn't hold moz in the cache. My OS should know that Moz is very 'hot' with several loads/day and it should keep all the metadata in the cache. I was also thinking that if we got really fancy, it could prefetch all the data in
As for battery-backed persistent caches, it could be made easier, I think, with a capacitor on the drive and some leftover inaccessible sectors at the end of the drive. When power cuts out the head drops back to the 'recover area' and starts dumping whatever's in the write buffer to the disk, when power is restored the first thing the drive does is check the 'recover area' for any 'dumped data', which it commits to disk (this time where it belongs) before your BIOS can even query the drive for booting. The capacitor would just buy you enough time to do that dump, and the drive could remain unplugged indefinitely without the risk of losing that data, unlike a battery-backed cache.
I'm sorry, I meant to say that it was for low-end CONSUMER-level perhipherals. Nobody who actually needs to get work done sticks with USB for long, it's flaky and slow.
Simulating nuclear weapons is a heck of a lot better than finding out 'the old way'. If we're going to HAVE nukes (and there's no turning back now, I'm afraid) we may as well have the right ones for the job. I'd rather have them develop 'right-size' weapons for the post-cold war era than use giant 1970's warheads on third-world dictators. I find it foolish to demand reductions in 'collateral damage' while simultaneously demanding limits on harmless weapons simulations aimed at reducing said damage.
I'd rather have 20 very accurate, very small nukes in our arsenal than the thousands of ICBMs we have now, it would be just as effective a deterrent; getting there means money and (simulated) testing.