It seems to me that you need to breed them, not in the traditional sense, but you need to find people with an incredibily broad skillset that want to use it for evil.
It may not be so hard in a country that's had a decade of internet access, but i just dont see how north korea will have that many people to choose from.
Sure anyone can download script kiddie exploits and run them, but i've always imagined that industrial espionage was quite a bit removed from that.
I've never tried to argue that this would be a good thing for apple to do.
I just don't think it would be impossible for apple to get their heads together with dell and produce a line of 'certified x86' machines.
I'd imagine apple need a contingency plan in case PowerPC development dries up and they become less competetive with intel cpu's, and the obvious solution is something x86 based. I'd expect though that it'd be x86 based but not necessarily pc compatible.
Also apple already do support a fairly wide selection of hardware when you consider the number of peripherals that work. I've never had any trouble connecting a digital camcorder or camera to macosx - and i wish the same could be said for windows or linux.
What hardware do u have in mind as not working:
1 ) Network cards - most are rock solid in linux/bsd, and most are based on a handful of chipsets
2 ) Graphics cards - there aren't a whole lot of different families of these around anymore. Most macs use either ati or nvidia cards..
3 ) Hard drives - usb, scsi, ide and firewire are all standardized
4 ) Peripherals - anything usb or firewire that has existing mac drivers should be able to recompile and run on macosx86
5 ) Sound Cards - these could be trickier
Still it doesn't look *that* bad to me
Businesswise it doesn't make a lot of sense - but note that plenty of people do buy MacOS. I know more people that have paid for upgrades to OS 10 than have paid to upgrade from win2k to xp.
I know darwin is open source - i should have used sarcasm tags around it.
My point was that if apple were to release their GUI for x86 and their associated framework then i think it's quite likely that people would do things like port drivers for specific hardware. Right now there is no incentive for me to use darwin over linux or freebsd, but if darwin had the mac's user interface (even if it cost money) then i'd jump to it.
I'm fairly sure most of Apple's code is fairly portable. I'm sure they could take care of the GUI, iTunes, iPhoto etc.. and leave support for esoteric hardware to the OS community.
I know this isn't a perfect suggestion, but if they open sourced the underlying operating system then they could rely on community support to port drivers from linux/bsd architectures.
Then they'd be in a redhatish way supplying a certified OS base with a proprietory graphical interface.
Linuxes out of the box support for hardware on modern systems is pretty reasonable now, and i'm sure Apple would be able to have a "Certified for MacOS X86".
Having said that - apple's current stategy makes perfect sense so long as PowerPC's are available and keep getting faster.
In the US we generally pay for both outgoing and incoming messages (though my us message rate is less than half my uk rate, even though both are tmobile).
Google may have been able to cut a deal with the providers since they make money from the subscriber when the messages come each way.
When the screen on my 4 year old fujitsu finally dies it's going to become a server.
I'll get some external firewire enclosures so i strap a bunch of storage to it. It's got wireless networking and enough cpu power to handle most tasks very capabably.
Plus it's quiet, low power and since it's wireless i can hide it somewhere to avoid it counting towards the wife's computer limit.
I've never written a linux chat script but my box does need a secureid password to authenticate to the VPN. I just wrote a little perl wrapper that asks for the code then runs the vpn connect scripts.
I'm not even sure the exploit listed is valid. IIRC secureid now does use alpha characters and i'm not sure it allows you to be simultaneously logging in twice using the same credentials.
I can't find the source now, but i think Jonathan Schwartz mentioned that outside of a couple of universities Sun had more people on OS than anyone else.
Given OpenOffice, Netbeans, Evolution, Gnome and now Solaris that's not too surprising.
There's little point in Sun fighting with intel and amd to produce the highest Mhz chip. They can buy opteron's, stick them in boxes and provide a good low to mid-end system.
OTOH Solaris is already REALLY good at multitasking. The system i'm typing this on has almost 5,000 threads, it's at 80% utilization and it's still very responsive.
As you put more tasks onto a single CPU it'll have to burn more and more cycles doing context switches and suffer from register starvation.
Plus large boxes benefit from economies of scale and can have features that aren't practical in smaller ones:
When a CPU fails the system can take that motherboard out of circulation, then the admin can replace it at their convenience. Same for memory and psu's. Usually no downtime.
Plus we already know that it takes less resources to admin a unix machine than a windows box. Now consider a 144 CPU x 32 Core machine. Even IF it could only handle the workload of 500 windows servers the admin costs are slashed further.
Also consider that the cache might be shared, but then consider that all those cores will most likely be running the same application. I'm sure there's lots of code within oracle or java that gets reused frequently by all the processors. An eightcore chip with 16MB of cache will naturally be able to cache much more of the shared resources than 8 cpu's with 2MB cache.
I've got gentoo on my Fujitsu Lifebook and suspend-to-ram has always worked.
Maybe I enabled something in the kernel, but 'apm -s' always does the trick for me.
It's also not at all flakey. On average i suspend the machine twice a day and it's never failed to come back up - except when the battery has run out. I've hit up times of a month or so, and usually break it when i need windows for something.
The CPU isn't on the main datapath and wont be able to do any sort of high speed processing. In fact, the CPU is a 8032 with 4k of RAM - i had a more powerful system in '84. You are going to be REALLY struggling to do any sort of encryption in such a limited environment.
Now i'm not saying your suggestion is technically impossible, i'm just saying that it's not very condusive to the way current off-the-shelf componets are designed.
Doing a custom silicon design for this camera would probably add hundreds of thousands to the development costs, and it's hard to recoup that through increased sales.
It's true that they could place a key in a ROM area on the CPU, but that would require a CPU with enough space cycles to perform encryption and decryption.
You'd either have to store the images unencrypted on the flash and encrypt the transfers, or you'd have to encrypt them on capture and decrypt them each time you try to review one.
I know nothing about that particular smal chip, but i'm sure it's general purpose CPU isn't all that fast. Most cheap USB controllers use some sort of FPGA arrangement to do bulk transfers, which of course limits your encryption choices further.
Of course if there is an encryption key in there then you have to somehow stop the employees at every Big Print Central location in the US from copying it off the hard disk or leaking their latest software update CD....
Well obviously they could lock it down - they could devise their own interface spec, build a custom chip to put in the camera...
But in reality they decided to use off the shelf components to slash development costs and it's coming back to bite them.
The problem with anything like that is, that the reverse engineering will always be more trouble than it's worth to the person who does the engineering. If i put 10 hours of my time into it then i could have bought a few cheap digicams with the money i've lost in earnings.
But it's a challenge and if i had the time and didn't hate debugging usb so much i'd take a shot at it. And if one person suceeds then everybody can reap the benifits.
Slashdot Mirror
How do you hire hackers?
It seems to me that you need to breed them, not in the traditional sense, but you need to find people with an incredibily broad skillset that want to use it for evil.
It may not be so hard in a country that's had a decade of internet access, but i just dont see how north korea will have that many people to choose from.
Sure anyone can download script kiddie exploits and run them, but i've always imagined that industrial espionage was quite a bit removed from that.
I've never tried to argue that this would be a good thing for apple to do.
I just don't think it would be impossible for apple to get their heads together with dell and produce a line of 'certified x86' machines.
I'd imagine apple need a contingency plan in case PowerPC development dries up and they become less competetive with intel cpu's, and the obvious solution is something x86 based. I'd expect though that it'd be x86 based but not necessarily pc compatible.
Also apple already do support a fairly wide selection of hardware when you consider the number of peripherals that work. I've never had any trouble connecting a digital camcorder or camera to macosx - and i wish the same could be said for windows or linux.
What hardware do u have in mind as not working:
1 ) Network cards - most are rock solid in linux/bsd, and most are based on a handful of chipsets
2 ) Graphics cards - there aren't a whole lot of different families of these around anymore. Most macs use either ati or nvidia cards..
3 ) Hard drives - usb, scsi, ide and firewire are all standardized
4 ) Peripherals - anything usb or firewire that has existing mac drivers should be able to recompile and run on macosx86
5 ) Sound Cards - these could be trickier
Still it doesn't look *that* bad to me
Businesswise it doesn't make a lot of sense - but note that plenty of people do buy MacOS. I know more people that have paid for upgrades to OS 10 than have paid to upgrade from win2k to xp.
I know darwin is open source - i should have used sarcasm tags around it.
My point was that if apple were to release their GUI for x86 and their associated framework then i think it's quite likely that people would do things like port drivers for specific hardware. Right now there is no incentive for me to use darwin over linux or freebsd, but if darwin had the mac's user interface (even if it cost money) then i'd jump to it.
I'm fairly sure most of Apple's code is fairly portable. I'm sure they could take care of the GUI, iTunes, iPhoto etc.. and leave support for esoteric hardware to the OS community.
I know this isn't a perfect suggestion, but if they open sourced the underlying operating system then they could rely on community support to port drivers from linux/bsd architectures.
Then they'd be in a redhatish way supplying a certified OS base with a proprietory graphical interface.
Linuxes out of the box support for hardware on modern systems is pretty reasonable now, and i'm sure Apple would be able to have a "Certified for MacOS X86".
Having said that - apple's current stategy makes perfect sense so long as PowerPC's are available and keep getting faster.
In the US we generally pay for both outgoing and incoming messages (though my us message rate is less than half my uk rate, even though both are tmobile).
Google may have been able to cut a deal with the providers since they make money from the subscriber when the messages come each way.
They'd have to go into a FUQ
When the screen on my 4 year old fujitsu finally dies it's going to become a server.
I'll get some external firewire enclosures so i strap a bunch of storage to it. It's got wireless networking and enough cpu power to handle most tasks very capabably.
Plus it's quiet, low power and since it's wireless i can hide it somewhere to avoid it counting towards the wife's computer limit.
this shouldn't have any impact on the open source world
I've never written a linux chat script but my box does need a secureid password to authenticate to the VPN. I just wrote a little perl wrapper that asks for the code then runs the vpn connect scripts.
I'm not even sure the exploit listed is valid. IIRC secureid now does use alpha characters and i'm not sure it allows you to be simultaneously logging in twice using the same credentials.
I think most of Sun's stuff is dual licensed - GPL + Sun Public License.
Essentially contributors agree to license their code to sun as well as under the gpl.
I can't find the source now, but i think Jonathan Schwartz mentioned that outside of a couple of universities Sun had more people on OS than anyone else.
Given OpenOffice, Netbeans, Evolution, Gnome and now Solaris that's not too surprising.
There's little point in Sun fighting with intel and amd to produce the highest Mhz chip. They can buy opteron's, stick them in boxes and provide a good low to mid-end system.
OTOH Solaris is already REALLY good at multitasking. The system i'm typing this on has almost 5,000 threads, it's at 80% utilization and it's still very responsive.
As you put more tasks onto a single CPU it'll have to burn more and more cycles doing context switches and suffer from register starvation.
Plus large boxes benefit from economies of scale and can have features that aren't practical in smaller ones:
When a CPU fails the system can take that motherboard out of circulation, then the admin can replace it at their convenience. Same for memory and psu's. Usually no downtime.
Plus we already know that it takes less resources to admin a unix machine than a windows box. Now consider a 144 CPU x 32 Core machine. Even IF it could only handle the workload of 500 windows servers the admin costs are slashed further.
Also consider that the cache might be shared, but then consider that all those cores will most likely be running the same application. I'm sure there's lots of code within oracle or java that gets reused frequently by all the processors. An eightcore chip with 16MB of cache will naturally be able to cache much more of the shared resources than 8 cpu's with 2MB cache.
Let me see here...
Sun provide OpenOffice and Netbeans as open source, and have a proven track record of investing money it it.
IIRC Sun have more people working on OS software than any other company in the world.
The grandparent hasn't really grasped this.
It's not like i can download a BIOS form abit and make garbage which matches the hash.
It allows me to make two pieces of garbage which have the same hash
Once home users realise that they have to pay every month just to keep windows working - there will be a lot more drive to get linux into homes.
Anyway, what works in the enterprise is a lot different from what works in the home.
I've got gentoo on my Fujitsu Lifebook and suspend-to-ram has always worked.
Maybe I enabled something in the kernel, but 'apm -s' always does the trick for me.
It's also not at all flakey. On average i suspend the machine twice a day and it's never failed to come back up - except when the battery has run out. I've hit up times of a month or so, and usually break it when i need windows for something.
The proccessor doesn't normally do JPEG encoding.. there will be a custom DSP which will be FAR more capable.
In fact look at the block diagram for the original Dakota Disposable...
http://files.gentlenews.com/pdf/spca504bv01.pdf (Figure 4.1)
The CPU isn't on the main datapath and wont be able to do any sort of high speed processing. In fact, the CPU is a 8032 with 4k of RAM - i had a more powerful system in '84. You are going to be REALLY struggling to do any sort of encryption in such a limited environment.
Now i'm not saying your suggestion is technically impossible, i'm just saying that it's not very condusive to the way current off-the-shelf componets are designed.
Doing a custom silicon design for this camera would probably add hundreds of thousands to the development costs, and it's hard to recoup that through increased sales.
Sweeeet
If the key is generated when the camera is turned on then how will it be exchanged with the host?
Anwyay the first 128 bits of a jpeg file should be guessable - possibly more. Once you've got that, a known plaintext attack is trivial.
It's true that they could place a key in a ROM area on the CPU, but that would require a CPU with enough space cycles to perform encryption and decryption.
You'd either have to store the images unencrypted on the flash and encrypt the transfers, or you'd have to encrypt them on capture and decrypt them each time you try to review one.
I know nothing about that particular smal chip, but i'm sure it's general purpose CPU isn't all that fast. Most cheap USB controllers use some sort of FPGA arrangement to do bulk transfers, which of course limits your encryption choices further.
Of course if there is an encryption key in there then you have to somehow stop the employees at every Big Print Central location in the US from copying it off the hard disk or leaking their latest software update CD....
Well obviously they could lock it down - they could devise their own interface spec, build a custom chip to put in the camera...
But in reality they decided to use off the shelf components to slash development costs and it's coming back to bite them.
The problem with anything like that is, that the reverse engineering will always be more trouble than it's worth to the person who does the engineering. If i put 10 hours of my time into it then i could have bought a few cheap digicams with the money i've lost in earnings.
But it's a challenge and if i had the time and didn't hate debugging usb so much i'd take a shot at it. And if one person suceeds then everybody can reap the benifits.
Someone could license the tivo architecture and merge it with a dvd recorder.
Maybe pioneer could do it...
http://www.pioneerburner.com/
That's security through obscurity. If mutt became as popular as outlook then works would look at your address book.
It doesn't seem like it'd be hard to have a worm write an arbitrary address into your address book.
Then LOAF would propogate that address to your friends, and then spammers could use the address programmed into the worm as the from address.
On the whole though this seems like a really nice addition to existing spam blocking systems.
Unfortunately the cases where i recieve email from a friend of a friend are relatively rare - but that's just me.
It also does have some privacy issues - since it'd essentially enable me to check if one of my friends happens to have my wife in his address book...
That doesn't seem right.
Right now mine has 150 users on a 12CPU/12Gb machine and Mozilla runs complex websites fine. Flash is a little sluggish but that's not too suprising.