Read the complaint carefully - it tries to get around the "it's a trade secret once it's out you're SOL" argumnet by claiming that the defendants were under a license from Xing (presumably a shrink-wrap) that included a clause prohibiting reverse engineering - thus they attempt toextend the trade secret to at least one of the named defendants - presumably all the rest are 'criminal conspirators'
Now's the time to play the Scandinavian card (like in all the Scientology cases) - quick someone have the DeCSS source read into the Swedish parliamentary record......
This is not a criminal trial it's a civil trial - the rules are different - the right to a jury is only limited to the deciding of matters of fact - if the judge decides that all the facts are agreed to then you don't get a jury
entering a temporary restraining order and preliminary and permanent injunctions, enjoining and restraining Defendants, their officers, directors, principals, agents, servants,employees, attorneys, successors and assigns, and all those acting in concert, combination or participation with any of them either directly or indirectly, singly or together, from making any further use or otherwise disclosing or distributing, on their web sites or elsewhere, or "linking" to other web sites which disclose, distribute, or "link" to any proprietary property or trade secrets relating to the CSS technology... (my emphasis)
IANAL - but they want to try an nip any widespread mirroring in the bud before they lose control (I think they have already) - however this pretty much includes anyone in the US who might want to set up a mirror. I suggest that offshore mirrors pop up in as many different places as possible - this forces their lawyers to have to work on many many fronts (very expensive) - and whan/if they come after you raise a public stink and quietly shut your site - if for every site they shut 10 more rise up in protest they can't win. I've watched the same techniques used successfully with the scientology vs. the net brush war over the past few years.
I nominate the broom from the sorcerer's apprentice as the official team mascot.
Reading the complaint I think that it's going to be VERY important to impress on the judge at the very beginning that NONE of the defendants were bound by the trade secret because they were not lisencees of it in fact you need to pull into court the people who did license it and put it out in such a format that it was easy to crack.
I beleive the thing will play out (my guess): an attempt for a temporary restraining order this week (something the judge is probably inclined to grant if he sees probably cause - raising 1st ammendment concerns might derail this) followed by an attempt at a judgement that the alleged trade secret theft occured (might take a year or more and could result in the temporary injunction being made permanent) followed by a penalty phase to assess any damages. A jury trial in Silicon Valley on this issue could be a real hoot (to get a jury the defendants will have to prove that there is some matter of fact in the complaint that must be decided - if it's just arguing about the law then the judge will decide)
As a (somewhat silly) aside - given that info about DeCSS was posted to slashdot this injunction might be read as prohibiting anyone from linking to slashdot at all.... depends on how vindictive the lawyers on the other side are.
I would love to see a test case for linking as a 1st amendment right - this might make a wonderfull test case (or maybe not given that the other side probably has way too many lawyers).
I bet the NSA does this twice before breakfast.... seriously though It's decidedly doable with enough die to work from - just etch your way down, take photos - keep at it 'till you think you have all the layers then apply some form of pattern matching to extract the transistors
(much harder for something where the keys are stored in some form of charge storage BTW)
yeah - I agree - this is really my point I think that there are classes of applications that can't be addressed by Open Source at all (even though I'm a big OS advocate),in this case because their security depends (however weakly) on transformations that take place in a compiled binary that is hard to reverse engineer
I still remember accidentally wandering into a room full of people 10 years back with photo-micrographs spread out on a table trying to find hidden registers in a VGA controller....
You are just making it more obscure.... and raising the bar for people who want to break it.
BTW you should read up on 'skipjack' proposed by the NSA a while back - it also had 'tamper proof' packaging - which to my mind really means 'costs more to break'
The even easier solution is not build it into the CAD program, just have people mail files securly with GnuPG/PGP. Why make CAD programmers deal with encryption, its not their baliwick!
Ah - but remember the premise - the end-customer must not be able see the secret information - it's purely for use with a tool that sees the encrypted file and performs some action on it that in itself does not reveal the secret to the end customer
People want to be able to distribute encrypted IP (chip designs) to customers for simulation. There are a couple of reasons for this - IP providers want to protect their product - and they want to protect themselves from liability should the customer change the source and then blow a lot of money on building a bogus chip.
Traditionally this problem has been handled in a closed source world with a public key/private key sort of setup with the private key (for decryption) and encryption algorithm embedded in the compiler binary somewhere.
This leaves the IP provider's product at the mercy of the vendor of the CAD tool
A few years back this all fell apart for Verilog, a popular simulation compiler, for various reasons the language runtime is extensible, it was also interpreted - this left a version of the compiler which contained symbols. An anonymous poster to comp.lang.verilog pointed out how to write a gdb script that set a breakpoint in 'yylex' and extracted the decrypted token from IP.
All hell broke loose.... eveyone who'd ever sent encrypted IP to customers was now open to the world...
So - back to the topic - does anyone have any idea how one can do this sort of thing in an Open Source world - send people secret stuff to be used by an OS program without giving away the secret? Given that the only schemes I've seen to do this rely on security-by-obscurity (as above) I suspect it just can't be done.
no! we can't leave them in the clutches of the evil empire.... home users are people too.... if you want windows to go away (or go open source) you have to starve it's creator....
well I sort of had that in mind... I figured a dollar bill that could track everyone who'd ever used it would be just the sort of thing that big-brother would find attractive
I wouldn't go so far as to say "remember a time when money was directly convertable into gold AND silver".... some of the largest US political battles of about 100 years back were arguments over whether there should be a gold or a silver standard - this had more to do with the liquidity of the currency (how much cash was actually in circulation) and how much money is available for capitalising new investments is directly related to this. IANAE but I suspect that the abandonment of either standard (in a strict 'there must be actuall gold/silver in ft knox' sense) in favor of a more liquid but carefully govt. controlled currency (by the Fed) probably has a lot to do with the economic boom times we've seen in the latter half of this century.
As an aside "the Wizard of Oz" is suposedly an allegory about the gold (big biz./yellow brick) vs silver (populist farmers/silver slippers[ruby was a hollywood invention]).
Think about how we all type - we have to align our fingers with the keyboard - touch typists find the home positions tactilly on the keyboard and type relative to them, people like me who never learned to touch type have to look occasionally to orient their fingers every so often. It's also a mostly 2-d thing - you are typing onto a 2-d surface.
Typing in the air has no frames of reference (unless you have some VR keyboard and goggles etc) and it's a 3-d sort of thing - no hard 2-d thing to stop your fingers at the end of very stroke.
Instead I suspect it's probably getting close to the time when we can come up with a new typing metaphor - hopefully something a little easier on my wrists - maybe 'typing' with my arms relaxed in my lap or something. With something like this a form of virtual chord keyboard might work well too meaning we could get away from the positional locations of keys on a keyboard which might be more suited for virtual keyboards.
Has anyone out there become proficient with a chord keyboard of some sort? can you type as fast or are you limited more by the time between chords?
Of course with cool MEMS technology like this just think of the interesting musical instruments we can create!
Electromigration .... and metastability ....
on
Intel using FreeBSD
·
· Score: 1
As your tempature rises (every 10 degrees increases the reaction rate 2x times), and we approach.1 micron widths, (18 atoms wide!) you have faster migration of the chip chemistry out of where you want, to where you don't want. Even with old TTL, the projected life is 50 years. The newer chips will have less life. (I don't remember the projected life of the newest.18 micron chips)
You're talking about 'electromigration'... basicly it's something like the 'tide of electrons down a wire nudge the atoms a bit in their direction, the narrower the path the higher the current density and a stronger 'nudge'
This has a runnaway failure case where as atoms get nudged away from a spot the wire gets narrower, the current density goes up, more atoms move etc etc
The rate of electromigration has a lot to do with the width of the wires (how many atoms), the shape of them (sharp corners can be a problem) and the current density - which scales down as features go down - but from memory it's PEAK current density that's the issue so you have to be carefull about wires with lots of capacitance on them.
As important as this is there's another problem that effects up-time - metastability - basicly in every computer there are places where signals cross from one clock domain to another and modern flip-flops go a bit bizarre if the signals they are storing change just at the moment that the act of storage is occuring - usually this is avoided by good synchronous design - but where signals cross clock domains this can't be avoided. Instead we design special flops that are less likely to 'go metastable' and put multiple flops one after the other all to reduce the chances of metastabile failures. But that's all you can do - reduce the chance - you can't avoid it - all you can do is calculate the chance of failure for a particular clock crossing signal (say 1 per 100 years) and multiply it by the number of such crossings (say 100 in a system giving in this case a chance of 1/year).
All of this is a long way of pointing out that there's no way Intel has any idea which chips in their systems (at the least the ones that they didn't design) are subject to metastable failures - it's unlikely they are spec'd with data on such failure rates and I doubt anyone bothered to discover all the potentially metastable flops in a whole computer system and add up the chances - a MTBF specd by disk drive motor MTBFs may not include other failures that are transitory.
Finally - a quick note about clock chipping.... what happens when you have a clock-chipped processor running on the hairy edge is that you are forcing flops into operating in a metastable region - just because things don't fail right away doesn't mean they wont next week, or month or....
Am I the only one who is suprised that someone is shipping something worth "a million dollars" by UPS? Either it's not really worth that much (in which case M$ probably misrepresented themselves to the NYPD) or the guy who sent it to the mail room rather than buying it a seat on a plane and escorting it personally should be fired forthwith.
That's fine for a $50,000 commercial color copier, but how is that going to work for the $93 inkjet that I buy using paper money in WalMart?
Ah.... but how do you know that the printer driver didn't snarf your P3 ID, your IP address, your email address (and saved password) and send them off along with the printer ID to the FBI the first time you plugged your PC into the net.... if it's not open source you don't.... it probably wont ask first
back in the mid 70s me and my friends were writing a compiler for a 6800 - it fit in 2k bytes - we named ourselves 'uSoft' (with a greek mu) pronounced... well you guessed it - looking back it was kind or the generic software company name of it's time when every thing had a mu in it. Anyway we were pissed when we found some guy in the US using our name - and he had a lame-o basic interpreter....I still have some card decks somewhere labelet '(C) USOFT'
Sadly we were a bunch of comp-sci-geeks in NZ who didn't realise that we had no way to market our compiler (no platforms, no contact with the then expanding San Jose micro community.... etc) we were pretty naive.
Now I wish we'd incorporated - we could have sold that name for big-bux:-) oh well that's life.
After WW2 the Allies left the Germans with a constitutions that bans organisation which plan world domination.... which just happens to be one of $cientology's goals (well that and chasing away all the space aliens that are haunting them).
This is one of the main reasons that Germany is down on Scientology (that and some real estate scams by its members)
It used to be that we had 'copyright laws' that 'worked' because it was hard to copy things.... if you wanted to copy a book the cost of typsetting it was very high up-front NRE - if you got sued you lost all that money you put upfront - only a small number of people had the means (access to typesetting hardware and a press) so it was easy to find someone who'd copied your work.
Then the xerox machine came along and everything changed - suddenly you could copy anything on paper for pennies, you're no longer looking for someone who copied your work and did a big press run - the usefullness of copyright went way down.
Then came digital media and the cost of copying went to an effective 0 - worse yet thanks to the 'net the cost of DISTRIBUTION has gone down to pennies - there are whole industries based on copying things and getting them from one place to another - these guys are in trouble - whether they realise it or not.
I think that what we're seeing for audio and video is the equivalent of the the arrival of the xerox machine - pressing vinyl or CDs used to be beyond the reach of us mere mortals - now we can all do it on our computers.
And MP3's are the next step - no media, no cost of distribution - things like publishers, record labels and moveie distribution houses are dinosaurs - they just don't know it yet
However now we've got a problem.... we still need a way for the original content producers (authors, musicians, movie studios) to make money.... this is what we should be trying to solve - not how do we get the distribution people to get their cut - they are history.
Finally there's a fundamental problem with encrypting digital media - at some point - in the machine where it's being played - out of the control of the media's author(s) - the encrypted data has to be rendered into some form that is usefull - bits in a frame-buffer, bits into an audio DAC. So long as that's true people will find ways to get their hands on the digital bits. Sadly the only way that's likely to work is to integrate these functionalitys (sound out dacs, whole frame buffers) out of the reach of mere mortals like the writers of Linux drivers - this can only be bad for a platform like linux that has small market share - will people write drivers for complex media chips for us? will they give us enough information to do it ourselves? (without giving away how to get at the precious media bits) - we still can't play our DVD disks under Linux - pissing off a whole bunch of geeks by not supporting your hardware is just going to get them pissed enough to reverse engineer your technology so they can use what they paid for..... remember there are alot more of us geeks than the people trying to hide the bits
Heat disipation in rams has normally been in the sense circuitry and pads - and while hopefully accesses are spread randomly across a die dram manufacturers have to be carefull about hotspots (like when you loop reading the same location for ever).
In fact reducing the number of off-die memory accesses may reduce the power (no need to source/sink to those external bus signal's caps)
The more complex anything gets - the lower the yield (basic rule of nature - I suspect it applies to life too:-)
Slashdot Mirror
I figure that without good coffee I'm toast
yup - I did too (got to sneak that tax deduction in under the wire too :-) - thanks guys for your great work
Read the complaint carefully - it tries to get around the "it's a trade secret once it's out you're SOL" argumnet by claiming that the defendants were under a license from Xing (presumably a shrink-wrap) that included a clause prohibiting reverse engineering - thus they attempt toextend the trade secret to at least one of the named defendants - presumably all the rest are 'criminal conspirators'
How about this one this one where santa is taken away by security guards - although I can't say I blame him - I think the mom was evil ....
Now's the time to play the Scandinavian card (like in all the Scientology cases) - quick someone have the DeCSS source read into the Swedish parliamentary record ......
This is not a criminal trial it's a civil trial - the rules are different - the right to a jury is only limited to the deciding of matters of fact - if the judge decides that all the facts are agreed to then you don't get a jury
entering a temporary restraining order and preliminary and permanent injunctions, enjoining and restraining Defendants, their officers, directors, principals, agents, servants,employees, attorneys, successors and assigns, and all those acting in concert, combination or participation with any of them either directly or indirectly, singly or together, from making any further use or otherwise disclosing or distributing, on their web sites or elsewhere, or "linking" to other web sites which disclose, distribute, or "link" to any proprietary property or trade secrets relating to the CSS technology ... (my emphasis)
IANAL - but they want to try an nip any widespread mirroring in the bud before they lose control (I think they have already) - however this pretty much includes anyone in the US who might want to set up a mirror. I suggest that offshore mirrors pop up in as many different places as possible - this forces their lawyers to have to work on many many fronts (very expensive) - and whan/if they come after you raise a public stink and quietly shut your site - if for every site they shut 10 more rise up in protest they can't win. I've watched the same techniques used successfully with the scientology vs. the net brush war over the past few years.
I nominate the broom from the sorcerer's apprentice as the official team mascot.
Reading the complaint I think that it's going to be VERY important to impress on the judge at the very beginning that NONE of the defendants were bound by the trade secret because they were not lisencees of it in fact you need to pull into court the people who did license it and put it out in such a format that it was easy to crack.
I beleive the thing will play out (my guess): an attempt for a temporary restraining order this week (something the judge is probably inclined to grant if he sees probably cause - raising 1st ammendment concerns might derail this) followed by an attempt at a judgement that the alleged trade secret theft occured (might take a year or more and could result in the temporary injunction being made permanent) followed by a penalty phase to assess any damages. A jury trial in Silicon Valley on this issue could be a real hoot (to get a jury the defendants will have to prove that there is some matter of fact in the complaint that must be decided - if it's just arguing about the law then the judge will decide)
As a (somewhat silly) aside - given that info about DeCSS was posted to slashdot this injunction might be read as prohibiting anyone from linking to slashdot at all .... depends on how vindictive the lawyers on the other side are.
I would love to see a test case for linking as a 1st amendment right - this might make a wonderfull test case (or maybe not given that the other side probably has way too many lawyers).
(much harder for something where the keys are stored in some form of charge storage BTW)
You are just making it more obscure .... and raising the bar for people who want to break it.
BTW you should read up on 'skipjack' proposed by the NSA a while back - it also had 'tamper proof' packaging - which to my mind really means 'costs more to break'
Ah - but remember the premise - the end-customer must not be able see the secret information - it's purely for use with a tool that sees the encrypted file and performs some action on it that in itself does not reveal the secret to the end customer
Traditionally this problem has been handled in a closed source world with a public key/private key sort of setup with the private key (for decryption) and encryption algorithm embedded in the compiler binary somewhere.
This leaves the IP provider's product at the mercy of the vendor of the CAD tool
A few years back this all fell apart for Verilog, a popular simulation compiler, for various reasons the language runtime is extensible, it was also interpreted - this left a version of the compiler which contained symbols. An anonymous poster to comp.lang.verilog pointed out how to write a gdb script that set a breakpoint in 'yylex' and extracted the decrypted token from IP.
All hell broke loose .... eveyone who'd ever sent encrypted IP to customers was now open to the world...
So - back to the topic - does anyone have any idea how one can do this sort of thing in an Open Source world - send people secret stuff to be used by an OS program without giving away the secret? Given that the only schemes I've seen to do this rely on security-by-obscurity (as above) I suspect it just can't be done.
no! we can't leave them in the clutches of the evil empire .... home users are people too .... if you want windows to go away (or go open source) you have to starve it's creator ....
I wouldn't go so far as to say "remember a time when money was directly convertable into gold AND silver" .... some of the largest US political battles of about 100 years back were arguments over whether there should be a gold or a silver standard - this had more to do with the liquidity of the currency (how much cash was actually in circulation) and how much money is available for capitalising new investments is directly related to this. IANAE but I suspect that the abandonment of either standard (in a strict 'there must be actuall gold/silver in ft knox' sense) in favor of a more liquid but carefully govt. controlled currency (by the Fed) probably has a lot to do with the economic boom times we've seen in the latter half of this century.
As an aside "the Wizard of Oz" is suposedly an allegory about the gold (big biz./yellow brick) vs silver (populist farmers/silver slippers[ruby was a hollywood invention]).
Typing in the air has no frames of reference (unless you have some VR keyboard and goggles etc) and it's a 3-d sort of thing - no hard 2-d thing to stop your fingers at the end of very stroke.
Instead I suspect it's probably getting close to the time when we can come up with a new typing metaphor - hopefully something a little easier on my wrists - maybe 'typing' with my arms relaxed in my lap or something. With something like this a form of virtual chord keyboard might work well too meaning we could get away from the positional locations of keys on a keyboard which might be more suited for virtual keyboards.
Has anyone out there become proficient with a chord keyboard of some sort? can you type as fast or are you limited more by the time between chords?
Of course with cool MEMS technology like this just think of the interesting musical instruments we can create!
You're talking about 'electromigration' ... basicly it's something like the 'tide of electrons down a wire nudge the atoms a bit in their direction, the narrower the path the higher the current density and a stronger 'nudge'
This has a runnaway failure case where as atoms get nudged away from a spot the wire gets narrower, the current density goes up, more atoms move etc etc
The rate of electromigration has a lot to do with the width of the wires (how many atoms), the shape of them (sharp corners can be a problem) and the current density - which scales down as features go down - but from memory it's PEAK current density that's the issue so you have to be carefull about wires with lots of capacitance on them.
As important as this is there's another problem that effects up-time - metastability - basicly in every computer there are places where signals cross from one clock domain to another and modern flip-flops go a bit bizarre if the signals they are storing change just at the moment that the act of storage is occuring - usually this is avoided by good synchronous design - but where signals cross clock domains this can't be avoided. Instead we design special flops that are less likely to 'go metastable' and put multiple flops one after the other all to reduce the chances of metastabile failures. But that's all you can do - reduce the chance - you can't avoid it - all you can do is calculate the chance of failure for a particular clock crossing signal (say 1 per 100 years) and multiply it by the number of such crossings (say 100 in a system giving in this case a chance of 1/year).
All of this is a long way of pointing out that there's no way Intel has any idea which chips in their systems (at the least the ones that they didn't design) are subject to metastable failures - it's unlikely they are spec'd with data on such failure rates and I doubt anyone bothered to discover all the potentially metastable flops in a whole computer system and add up the chances - a MTBF specd by disk drive motor MTBFs may not include other failures that are transitory.
Finally - a quick note about clock chipping .... what happens when you have a clock-chipped processor running on the hairy edge is that you are forcing flops into operating in a metastable region - just because things don't fail right away doesn't mean they wont next week, or month or ....
Am I the only one who is suprised that someone is shipping something worth "a million dollars" by UPS? Either it's not really worth that much (in which case M$ probably misrepresented themselves to the NYPD) or the guy who sent it to the mail room rather than buying it a seat on a plane and escorting it personally should be fired forthwith.
Ah .... but how do you know that the printer driver didn't snarf your P3 ID, your IP address, your email address (and saved password) and send them off along with the printer ID to the FBI the first time you plugged your PC into the net .... if it's not open source you don't .... it probably wont ask first
Sadly we were a bunch of comp-sci-geeks in NZ who didn't realise that we had no way to market our compiler (no platforms, no contact with the then expanding San Jose micro community .... etc) we were pretty naive.
Now I wish we'd incorporated - we could have sold that name for big-bux :-) oh well that's life.
This is one of the main reasons that Germany is down on Scientology (that and some real estate scams by its members)
I saw John wandering around at the last Linuxworld in San Jose .... he's still with us
Then the xerox machine came along and everything changed - suddenly you could copy anything on paper for pennies, you're no longer looking for someone who copied your work and did a big press run - the usefullness of copyright went way down.
Then came digital media and the cost of copying went to an effective 0 - worse yet thanks to the 'net the cost of DISTRIBUTION has gone down to pennies - there are whole industries based on copying things and getting them from one place to another - these guys are in trouble - whether they realise it or not.
I think that what we're seeing for audio and video is the equivalent of the the arrival of the xerox machine - pressing vinyl or CDs used to be beyond the reach of us mere mortals - now we can all do it on our computers.
And MP3's are the next step - no media, no cost of distribution - things like publishers, record labels and moveie distribution houses are dinosaurs - they just don't know it yet
However now we've got a problem .... we still need a way for the original content producers (authors, musicians, movie studios) to make money .... this is what we should be trying to solve - not how do we get the distribution people to get their cut - they are history.
Finally there's a fundamental problem with encrypting digital media - at some point - in the machine where it's being played - out of the control of the media's author(s) - the encrypted data has to be rendered into some form that is usefull - bits in a frame-buffer, bits into an audio DAC. So long as that's true people will find ways to get their hands on the digital bits. Sadly the only way that's likely to work is to integrate these functionalitys (sound out dacs, whole frame buffers) out of the reach of mere mortals like the writers of Linux drivers - this can only be bad for a platform like linux that has small market share - will people write drivers for complex media chips for us? will they give us enough information to do it ourselves? (without giving away how to get at the precious media bits) - we still can't play our DVD disks under Linux - pissing off a whole bunch of geeks by not supporting your hardware is just going to get them pissed enough to reverse engineer your technology so they can use what they paid for ..... remember there are alot more of us geeks than the people trying to hide the bits
In fact reducing the number of off-die memory accesses may reduce the power (no need to source/sink to those external bus signal's caps)
The more complex anything gets - the lower the yield (basic rule of nature - I suspect it applies to life too :-)
leave it out in the sun ....