Actually, very little in 'mainstream' sci-fi is original, the way the Startrek federation works and the conflicts between it and scarcity driven cultures have a very long history in fiction and some in real life. A great many people have though about the effects that various forms of 'magic' would have on people (such as the 'replicator') and have explored both the good and the bad sides in depth and with great clarity. What's more history has shown, especially recent history, that much of this magic can become reality. (startrek communicator anyone?)
In a very real way science fiction is a simulation of how the world might develop, both how we hope it should and how we fear it will. The startrek universe is definitely on nicer side of that coin. (I haven't heard of them having a "bioweapons division" for example.)
Exactly, it's such a tiny cost and to paints your company as caring about the stability and security of the consumers machines.
It means your salesmen can point to these stats, just like you have, and say "we spend our hard earned money to fix your problems".
Of course you must never even hint that every patch is actually a failure in the system that supposed to exist to prevent broken software from ever being released.
A Linux box being used to scan for stupidly weak SSH passwords, you don't say how they broke in.
So the obvious answer is that you used a stupidly weak password too.
More like carrying fewer weapons, so leave the hydrofluoric acid and nerve toxin behind. Then think seriously if you really need the molotov cocktails.
Even a thirty pound sledge might be a good weapon, but if you have to drop it before you can move all you've accomplished is to make it available to the other guy.
The TCP/IP checksums are really weak, only 16bits and rather a poor algorithm anyway. So more than one in 65 thousand errors will be undetected by a TCP/IP checksum. And that's not including buggy network adaptors and drivers that 'fix' or ignore the checksums.
If you're transferring gigabytes of data you really need something a lot better.
Still that's probably not the most common source of errors. You see the same problem exists when data is transferred across an IDE or SCSI bus if there's a checksum at all it's very weak and the amounts of data transferred across a disk bus are scary.
We have the source, it is possible to make a Linux distribution that
lets a normal user defend themselves.
The first look would probably be something like
Puppy Linux.
This boots off of secure storage (a CDR) to which it adds another session
just before you power off. This very simple technique gives the user the
ability to throw away a session by just turning off the machine. Puppy
also keeps the sessions independent on the disk so you can go back to a
known good session but still collect user data from later sessions.
Think carefully, the problem isn't that the administrator is "unqualified"
the only qualification they need is the ability to notice that something
is wrong. A that point Puppy linux would allow them to go back or start
with a clean CDR. It's still a bit of a pain to recover later changes
from the old setup though.
That's the key.
It's called the factory reset button, it puts the machine back to
the state it was when you bought it. Complete with all the extras you
bought for it. It leaves the machine ready to continue working with your
stuff. Don't miss that bit; it must only delete code that's made itself
part of the OS not your passive documents.
That's actually it; the core is that simple, you need to have a trusted
boot and protect that boot from any 'untrusted' code that runs later
and give the user the ability to stop that untrusted code from running.
This trusted 'zone' also has one other job; protect the user data,
make sure it cannot be damaged by anything once saved. That bit's called
a versioning filesystem.
Microsoft are trying to do this, but they are severely hampered by years
of crap code, bad installation management tools and horrible APIs. Linux
OTOH has the Unix history of solid multiuser protection, it's got a lot
less further to go.
Wouldn't stop downloading anymore because the machines at each end are much more powerful and so they can use much better compression formats than way back then.
On top of that, the transfer protocols (eg: bittorrent) are seriously effective at using any bandwidth they are given, the only reason they seem slow sometimes is that they're only given a tiny upload bandwidth compared to most websites.
With a reasonable quality you can have video playback at about 10Mbyte/minute, that makes the numbers easy.
It sounds like a lot for a modem, but the trickle through a modem is like a dripping tap, little by little it fills any bucket.
1 hour is 15Mbyte.
1 day is 370MByte
1 week is 2.5 Gigabytes
1 month is 11 Gigabytes, what was your broadband usage limit again?
And with bittorrent you don't need a server; this is the modem to modem rate with any number of peers.
I haven't looked a 'high end' calculators in years, I use computers all the time. I kind of knew they were still around but really, these machines are terrible!
My expectations were, a modern cheap processor... like the arm, possibly underclocked for power consumption; well looks like the HPs have that. An infinite amount of memory; well probably 64M each of RAM and flash. That's infinite for a calculator. And a small, but usable screen, probably 320x200x16(4) grayscale, (colour's supposed to consume a lot more power). And a pair of USB cables that allow you to connect to a PC or an external flash drive. The PC software would let you copy the entire calculator and run and program it on the PC (emulator) or the actual hardware.
Well, These TI's with a z80 processor, sorry you only use a z80 mask nowadays if you're a complete skinflint, "high end" gear uses processors that are easier to program. The 68k sounds reasonable; but it's probably a powerhog compared to the Arm (most 32bit+ processors are).
Probably the thing I'm most shocked about is the screen, those 132x64x2 displays are at least 15 years old and have never been big enough for a reasonable graph. But here we are stuck in the 90's or even the 80's.
Quite simply these machines should be two chips, a screen, a load of buttons, usb connector and a battery.
They should no longer be expensive; but are being sold for about the same prices as the smaller netbooks. Or this: http://en.wikipedia.org/wiki/GP2X_Wiz
I wouldn't go with never, but yes it is pretty rare.
You see a modern hard drive has two levels of error correction. There's the 'on the fly' correction, it takes very little CPU and happens fast. Then there's the 'advanced' mode, not only does it use the ECC to it's full extent but it uses the block CRCs so it can check guesses and lots of other techniques like merging multiple passes.
So yes there's a pretty good chance that the hard drive will be able to read a sector even though it's bad enough that it won't use it again.
This is because the really bright white LEDs are actually monochrome blue, they have a phosphor that converts some of that blue light into other colours, but not normally enough for a nice (sun like) colour.
There are other techniques that seem to convert the frequencies better; or they could use the old trick of putting different colour LEDs in one bulb. But for the moment if you want highest efficiency you're stuck with lots of blue in the light and a "cold" feel.
One point though, white LEDs are normally closer to the spectrum of the sun than incandescents, it's just that the blue spike is in the opposite direction to the very reduced blues you get from a incandescent. This is a known problem, so the conversions will continue to get better.
If you're gonna attack the curmudgeon at least use something real, C-# is the copy Microsoft made of Java when Sun said they had to follow the standard as written and not do their normal E.E.E. trick.
Rather than fixing msjava they took they're marbles home and painted them blue.
Why are they surprised at all the blue ball jokes?
It doesn't matter what FTP client is used, if the client is infected the malware can monitor the network and record the packets that any FTP client uses to login exactly as if it was monitoring the wire.
The same for any unencrypted well know protocol.
Also if the machine is infected a root certificate could be added to the system certificate store and this could be used to intercept encrypted communications which could then be decrypted (then re-encrypted) and monitored in the same way.
You may be missed if the software is unknown and it uses something like SSH which doesn't use SSL certificates. But really, if the client is compromised you're screwed.
Not really, either you have to use an email for the trial serial numbers or the software generates a trial number itself.
If it's the latter it's usually so simple to defeat that anyone with some smarts can do it.
If it's the former it's ignored and there's a small trade in 'Serialz'.
Online activation of the serial numbers can work, but that obviously has it's own host of troubles. Not least being that it's very difficult to stop a determined attacker from ripping out the serial number code completely.
What you want seems to be like the "all_squash" option on nfs exports (with the anonuid and anongid options). You can use these with a loopback nfs mount directly.
It's possible to do something similar with some fuse filesystems too. (Perhaps encfs?)
Personally, I don't think it's work the effort and use vfat with the "fmask" option set to disable the execute bits.
The bits haven't been moved, they are exactly where they always were, on the hard drive.
In fact in a way you've hit the nail on the head. The only way DRM begins to work is if it's associated with a physical object like a hard drive or a TPM chip. Of course it falls over as soon as anybody wants to listen to the music because the bits cannot be moved to the playback device, they have to be copied.
It used to be possible to fake the movement, the CD was the physical device and the playback devices were under complete control of the publisher because of their cost to build. The playback device (CD player) required the presence of the CD because it was impossible to copy all the bits at once without expensive kit. So it mostly worked, good enough.
Not any more. Even the kit to re-digitise the analog rendition is bargain bin cheap.
Because they make the disk with a sector size of 512 bytes (likely 4096 bytes inside the drive)
With modern drives and most especially flash drives, the CHS values normally are physically meaningless.
Except, with a flash drive the erase block size is likely to be 2^19 or 2^20 bytes. It's easy to set the drive so that the cylinders are 1048576 bytes, just set the heads to 64 and the sectors to 32. Each cylinder is then 1Mbyte, one real megabyte and one or two erase blocks.
Then 2^20 bytes is a reasonable size for an allocation unit too.
The smallest power of 10 that has 512 as a factor is 10^9. That is far too large for a cylinder or an allocation unit, even on a terabyte drive.
To put it bluntly, they use powers of two unless it's needed to con the consumer.
Slashdot Mirror
Actually, very little in 'mainstream' sci-fi is original, the way the Startrek federation works and the conflicts between it and scarcity driven cultures have a very long history in fiction and some in real life. A great many people have though about the effects that various forms of 'magic' would have on people (such as the 'replicator') and have explored both the good and the bad sides in depth and with great clarity. What's more history has shown, especially recent history, that much of this magic can become reality. (startrek communicator anyone?)
In a very real way science fiction is a simulation of how the world might develop, both how we hope it should and how we fear it will. The startrek universe is definitely on nicer side of that coin. (I haven't heard of them having a "bioweapons division" for example.)
Exactly, it's such a tiny cost and to paints your company as caring about the stability and security of the consumers machines.
It means your salesmen can point to these stats, just like you have, and say "we spend our hard earned money to fix your problems".
Of course you must never even hint that every patch is actually a failure in the system that supposed to exist to prevent broken software from ever being released.
A Linux box being used to scan for stupidly weak SSH passwords, you don't say how they broke in.
So the obvious answer is that you used a stupidly weak password too.
More like carrying fewer weapons, so leave the hydrofluoric acid and nerve toxin behind. Then think seriously if you really need the molotov cocktails.
Even a thirty pound sledge might be a good weapon, but if you have to drop it before you can move all you've accomplished is to make it available to the other guy.
The TCP/IP checksums are really weak, only 16bits and rather a poor algorithm anyway. So more than one in 65 thousand errors will be undetected by a TCP/IP checksum. And that's not including buggy network adaptors and drivers that 'fix' or ignore the checksums.
If you're transferring gigabytes of data you really need something a lot better.
Still that's probably not the most common source of errors. You see the same problem exists when data is transferred across an IDE or SCSI bus if there's a checksum at all it's very weak and the amounts of data transferred across a disk bus are scary.
We have the source, it is possible to make a Linux distribution that lets a normal user defend themselves.
The first look would probably be something like Puppy Linux. This boots off of secure storage (a CDR) to which it adds another session just before you power off. This very simple technique gives the user the ability to throw away a session by just turning off the machine. Puppy also keeps the sessions independent on the disk so you can go back to a known good session but still collect user data from later sessions.
Think carefully, the problem isn't that the administrator is "unqualified" the only qualification they need is the ability to notice that something is wrong. A that point Puppy linux would allow them to go back or start with a clean CDR. It's still a bit of a pain to recover later changes from the old setup though.
That's the key. It's called the factory reset button, it puts the machine back to the state it was when you bought it. Complete with all the extras you bought for it. It leaves the machine ready to continue working with your stuff. Don't miss that bit; it must only delete code that's made itself part of the OS not your passive documents.
That's actually it; the core is that simple, you need to have a trusted boot and protect that boot from any 'untrusted' code that runs later and give the user the ability to stop that untrusted code from running. This trusted 'zone' also has one other job; protect the user data, make sure it cannot be damaged by anything once saved. That bit's called a versioning filesystem.
Microsoft are trying to do this, but they are severely hampered by years of crap code, bad installation management tools and horrible APIs. Linux OTOH has the Unix history of solid multiuser protection, it's got a lot less further to go.
Ho-ho-hold on, hold on one second.
It was three or four years late.
Let's go back to dialup.
Wouldn't stop downloading anymore because the machines at each end are much more powerful and so they can use much better compression formats than way back then.
On top of that, the transfer protocols (eg: bittorrent) are seriously effective at using any bandwidth they are given, the only reason they seem slow sometimes is that they're only given a tiny upload bandwidth compared to most websites.
With a reasonable quality you can have video playback at about 10Mbyte/minute, that makes the numbers easy. It sounds like a lot for a modem, but the trickle through a modem is like a dripping tap, little by little it fills any bucket.
And with bittorrent you don't need a server; this is the modem to modem rate with any number of peers.
I haven't looked a 'high end' calculators in years, I use computers all the time. I kind of knew they were still around but really, these machines are terrible!
My expectations were, a modern cheap processor ... like the arm, possibly underclocked for power consumption; well looks like the HPs have that. An infinite amount of memory; well probably 64M each of RAM and flash. That's infinite for a calculator. And a small, but usable screen, probably 320x200x16(4) grayscale, (colour's supposed to consume a lot more power). And a pair of USB cables that allow you to connect to a PC or an external flash drive. The PC software would let you copy the entire calculator and run and program it on the PC (emulator) or the actual hardware.
Well, These TI's with a z80 processor, sorry you only use a z80 mask nowadays if you're a complete skinflint, "high end" gear uses processors that are easier to program. The 68k sounds reasonable; but it's probably a powerhog compared to the Arm (most 32bit+ processors are).
Probably the thing I'm most shocked about is the screen, those 132x64x2 displays are at least 15 years old and have never been big enough for a reasonable graph. But here we are stuck in the 90's or even the 80's.
Quite simply these machines should be two chips, a screen, a load of buttons, usb connector and a battery.
They should no longer be expensive; but are being sold for about the same prices as the smaller netbooks. Or this: http://en.wikipedia.org/wiki/GP2X_Wiz
They are all so very disappointing.
I wouldn't go with never, but yes it is pretty rare.
You see a modern hard drive has two levels of error correction. There's the 'on the fly' correction, it takes very little CPU and happens fast. Then there's the 'advanced' mode, not only does it use the ECC to it's full extent but it uses the block CRCs so it can check guesses and lots of other techniques like merging multiple passes.
So yes there's a pretty good chance that the hard drive will be able to read a sector even though it's bad enough that it won't use it again.
This is because the really bright white LEDs are actually monochrome blue, they have a phosphor that converts some of that blue light into other colours, but not normally enough for a nice (sun like) colour.
There are other techniques that seem to convert the frequencies better; or they could use the old trick of putting different colour LEDs in one bulb. But for the moment if you want highest efficiency you're stuck with lots of blue in the light and a "cold" feel.
One point though, white LEDs are normally closer to the spectrum of the sun than incandescents, it's just that the blue spike is in the opposite direction to the very reduced blues you get from a incandescent. This is a known problem, so the conversions will continue to get better.
Come on! He's a curmudgeon of course; it's right there on the tin!
It's not as if Microsoft isn't well known for instilling this attitude in both current and former users.
If you're gonna attack the curmudgeon at least use something real, C-# is the copy Microsoft made of Java when Sun said they had to follow the standard as written and not do their normal E.E.E. trick.
Rather than fixing msjava they took they're marbles home and painted them blue.
Why are they surprised at all the blue ball jokes?
OMG, trying to navigate on a 17 foot touchscreen.
Excel likes to mangle leading zeros.
Oh shit, tell me about it.
It ain't just leading zeros either just try typing, pasting or importing the codes "MARA01", "MARB02" and "MARC01"
Schools can usually get MS Office for free, the discount code is "Linux".
It doesn't matter what FTP client is used, if the client is infected the malware can monitor the network and record the packets that any FTP client uses to login exactly as if it was monitoring the wire.
The same for any unencrypted well know protocol.
Also if the machine is infected a root certificate could be added to the system certificate store and this could be used to intercept encrypted communications which could then be decrypted (then re-encrypted) and monitored in the same way.
You may be missed if the software is unknown and it uses something like SSH which doesn't use SSL certificates. But really, if the client is compromised you're screwed.
Not really, either you have to use an email for the trial serial numbers or the software generates a trial number itself.
If it's the latter it's usually so simple to defeat that anyone with some smarts can do it.
If it's the former it's ignored and there's a small trade in 'Serialz'.
Online activation of the serial numbers can work, but that obviously has it's own host of troubles. Not least being that it's very difficult to stop a determined attacker from ripping out the serial number code completely.
What you want seems to be like the "all_squash" option on nfs exports (with the anonuid and anongid options). You can use these with a loopback nfs mount directly.
It's possible to do something similar with some fuse filesystems too. (Perhaps encfs?)
Personally, I don't think it's work the effort and use vfat with the "fmask" option set to disable the execute bits.
You're right, it should be a Ballmer borg icon now.
The bits haven't been moved, they are exactly where they always were, on the hard drive.
In fact in a way you've hit the nail on the head. The only way DRM begins to work is if it's associated with a physical object like a hard drive or a TPM chip. Of course it falls over as soon as anybody wants to listen to the music because the bits cannot be moved to the playback device, they have to be copied.
It used to be possible to fake the movement, the CD was the physical device and the playback devices were under complete control of the publisher because of their cost to build. The playback device (CD player) required the presence of the CD because it was impossible to copy all the bits at once without expensive kit. So it mostly worked, good enough.
Not any more. Even the kit to re-digitise the analog rendition is bargain bin cheap.
The playkey, unlike the title folder, can't be copied--but it can be moved.
It's a very simple rule, it applies to every digital device sold.
Bits cannot be moved.
They can be copied. They can be erased. But you can only simulate a move.
If it looks like a move, either the bits were copied and deleted or the link to the bits was.
Bits cannot be moved.
Wrong, most people don't care because they don't even know what a gigabyte is.
A great many of them really can't get their brain around the difference between RAM and DISK, or GB and MB, let alone GB and GiB.
Because they make the disk with a sector size of 512 bytes (likely 4096 bytes inside the drive)
With modern drives and most especially flash drives, the CHS values normally are physically meaningless.
Except, with a flash drive the erase block size is likely to be 2^19 or 2^20 bytes. It's easy to set the drive so that the cylinders are 1048576 bytes, just set the heads to 64 and the sectors to 32. Each cylinder is then 1Mbyte, one real megabyte and one or two erase blocks.
Then 2^20 bytes is a reasonable size for an allocation unit too.
The smallest power of 10 that has 512 as a factor is 10^9. That is far too large for a cylinder or an allocation unit, even on a terabyte drive.
To put it bluntly, they use powers of two unless it's needed to con the consumer.
Windows 2000 had (has) everything needed to get past 4Gb.
Windows 2000 only came in 32bit version
If Windows 2000 (PRO) is installed on a machine with 4Gb of RAM it will use it all, even the memory with a physical address above the 4Gb line.
Windows 2000 datacenter can use 32Gb of RAM before it hits it's license limit.