You can still get 68k family processors from Freescale Semiconductor (who split off from Motorola 2 years ago). www.freescale.com -> 32-bit Microcontrollers -> 68k/coldfire.
There's no reason why a 68k couldn't be an application processor (rather than modem processor, in a dual-chip design) in a mobile phone. There's also no fundamental reason why AmigaOS couldn't be the GUI/AFX too. (The anciliary chips (SID? et al.) might be a problem, perhaps they could be emulated?)
That's something I would certainly like to see. I know you were only joking, but it's a nice image.
FatPhil, not speaking for anyone apart from himself.
I think that actual trend is that webpages that are full of useful information and aren't full of crap tend to produce better search engine results. And that webpages that contain useful information and aren't full of crap are sometimes designed by people who care a little about what they present. So some correlation certainly makes sense.
However, gamers twist this by filling their pages with what google thinks isn't crap, but the viewers might do, so even the above isn't a reliable correlation.
And there's no cause and effect between the two, they are both potentially effects of a similar cause though.
Without the pain of personal loss, lusers will not be so inclined to tighten up their system. So what if I'm part of a botnet? I'm not using the machine overnight anyway...
They bounce photons off each other twice. Knowing the speed the photon travels at they calculate the two distances to the other ship, and hence calculate the ships speed.
The result will be <c.
Draw a diagram, bounce the photons, find the time taken, do the subtraction, do the division - you'll get the same result.
Note that this _explicitly_ has assumed that photons always travel at c, but there's no other useful way to define time or distance unless you make such an assumption.
My instructions to anyone cooking steak for me is "throw it so hard at the hotplate that it bounces onto the other side, and then pick it up".
I now know "I'd like my steak raw, please" in as many languages as I know "hello".
And it's not just steaks. Steak tartare (which is not a steak) with onion, olives and spices is absolutely the perfect starter. Note that I live in the country with probably the tastiest and healthiest cows in the world - there's no shortage of space, and their foodchain contains no short loops.
Don't get me wrong - I absolutely adore almost anything done with chick-peas too!
What proportion of Europe do you think _doesn't_ have a bank holiday on 1st May? To save you time, you don't need to check Italy, Germany, France, Spain, Poland, Sweden, Finland and Switzerland, as I know for a fact that they do -- I have live(d) or work(ed) there, or consorted on May 1st with nationals therefrom.
For spammers, I've always promoted the "Death by a Million Paper Cuts" method. Everyone wronged merely has to contribute one paper cut, but they will all add up.
You seem to have a better insight than most! There is a technique used for both (a) recovering from noise and (b) assuring that if the noise is caused by evesdroppers then the information they can recover tells them nothing about the original payload.
Basically you have to make the message larger, like any error correcting scheme, but such that if you assume no more than a certain proportion of bits are leaked, there are equiprobable encodings for every possible symbol that could contain such leaked bits.
Hmmm, here's a very clumsy example which doesn't handle the SNR part: encode 0 as any of { 00, 11 }, and 1 as { 01, 10 } Let's say the source stream is '0', encoded to '00' Let's also assume that none of the signal is lost to noise or evesdropping. However, let's also assume the evesdropper did grab one bit. If the evesdropper knows the message is 'x0' then it could have been '00' or '10', so the evesdropper has gained no info. Likewise if he knows the transmitted message is 'x1' or '0x' or '1x', he gains no information about the plaintext.
Traditionally you would compress the plaintext before expanding it using these kinds of schemes.
Such techniques are typically called "privacy amplification". I would hope that Helger Lipmaa has some stuff on his website, as he seems to have lots of useful things to refer to. (I think he may be at ut.ee now, rather than cs.hut.fi)
It is _impossible_ to distinguish an evesdropper causing a -X dB perturbation on the transmitted signal from line noise that causes a -X dB perturbation on it.
It's stormy - you're getting a lower SNR on your fibre line today - are you sure that a small fraction of your bits are not being evesdropped.
Of course, that's why you use privacy amplification techniques to make small leaks of the signal to an adversary give no useful information about the real payload.
From a laws-of-physics and mathematical standpoint it's perfect. However, we're neither wave equations nor formulae - we're humans in the real world, using high power electronic devices that broadcast their state and their changes of state to their environs. Nothing about OPTs and QKE helps preserve the secrecy of the data once it's been decrypted and, say, stuck on a hard disk, or displayed as text on screen.
Oh dear - you've obviously never heard of padding.
Message length may be part of the information in the message. (If length is constant, then it's not; if it's variable, then it is.) That too must be removed from pain sight.
If Eve is _already_ pretending to be Bob to Alice, or if Bob and Alice have no way of authenticating each other.
However, on the whole, if you're building a site-site fibre link, then you'd expect the two parties to already be able to authenticate each other.
At some point there needs to exist a side channel. But after that, assuming you use the provably secure _primitive_ using an accepted protocol, then you will have provably unbreakable and un-MITM-able crypto.
I find nothing explains such complicated issues better than just simply providing a program with its source code that can be examined and tinkered with:
Slashdot Mirror
Why not just buy a 3-axis accelerometer demo board and development kit and actually do something innovative with it instead?
m ary.jsp?code=MMA7260Q&nodeId=01126911184209&tid=ts fp
e.g. http://www.freescale.com/webapp/sps/site/prod_sum
Woh! Who put a third image on that page.
m ember_releases/2005/Infineon_Release_053105.pdf """
:-)
I'm now convinced. It's this:
https://www.trustedcomputinggroup.org/news/press/
Infineon Announces Trusted Platform Module to Enhance PC Security;
Technical details on Infineon's TPM (SLB 9635 TT 1.2).
"""
Which is the serial number in that lower photo.
I said I could easily be persuaded, didn't I?
It's obviously a related chip, but definitely not the same model as the clear picture at the top of the article. The logo's in a different place, and the middle line has a different number of characters in it. However, it's clearly an Infineon chip with the same package. (looks like http://www.infineon.com/cgi-bin/ifx/portal/ep/cont entView.do?channelId=-71967&contentId=100820&progr amId=36188&programPage=%2Fep%2Fprogram%2Fpackage.j sp&pageTypeId=17099&contentType=PACKAGE ) That really doesn't narrow it down.
I notice that Infineon make _no_ mention of Apple, Macintosh, OS-X, or any other related terms on their TPM pages. It's just MS this, Windows that.
I too am not convinced, but could easily be persuaded with proper evidence.
The dell will have an ethernet link light.
The mac will have a glowing apple, but no ethernet link light.
You can't ping your fileserver - which would you rather have?
(as seen on alt.sysadmin.recovery)
http://www.google.com/trends?q=OSX%2C+ubuntu&ctab= 0&geo=all&date=all
Aren't statistics great!
You can still get 68k family processors from Freescale Semiconductor (who split off from Motorola 2 years ago). www.freescale.com -> 32-bit Microcontrollers -> 68k/coldfire.
There's no reason why a 68k couldn't be an application processor (rather than modem processor, in a dual-chip design) in a mobile phone. There's also no fundamental reason why AmigaOS couldn't be the GUI/AFX too. (The anciliary chips (SID? et al.) might be a problem, perhaps they could be emulated?)
That's something I would certainly like to see. I know you were only joking, but it's a nice image.
FatPhil, not speaking for anyone apart from himself.
I think that actual trend is that webpages that are full of useful information and aren't full of crap tend to produce better search engine results. And that webpages that contain useful information and aren't full of crap are sometimes designed by people who care a little about what they present. So some correlation certainly makes sense.
However, gamers twist this by filling their pages with what google thinks isn't crap, but the viewers might do, so even the above isn't a reliable correlation.
And there's no cause and effect between the two, they are both potentially effects of a similar cause though.
FatPhil
Without the pain of personal loss, lusers will not be so inclined to tighten up their system. So what if I'm part of a botnet? I'm not using the machine overnight anyway...
Happy LARTing,
FatPhil
Look at its parent post.
Are you absolutely sure that the central limit theorem leads to the conclusion you state?
Firstly you seem to be treating power as a vector - or at least signed.
Secondly you have ignored the variance of the summed signals, which grows with the number of cables.
I think you'll find the power is proportional to the variance.
It's been 15 years since I did this stuff though.
They bounce photons off each other twice.
Knowing the speed the photon travels at they calculate the two distances to the other ship, and hence calculate the ships speed.
The result will be <c.
Draw a diagram, bounce the photons, find the time taken, do the subtraction, do the division - you'll get the same result.
Note that this _explicitly_ has assumed that photons always travel at c, but there's no other useful way to define time or distance unless you make such an assumption.
What are you on?
I eat raw meat quite frequently.
My instructions to anyone cooking steak for me is "throw it so hard at the hotplate that it bounces onto the other side, and then pick it up".
I now know "I'd like my steak raw, please" in as many languages as I know "hello".
And it's not just steaks. Steak tartare (which is not a steak) with onion, olives and spices is absolutely the perfect starter. Note that I live in the country with probably the tastiest and healthiest cows in the world - there's no shortage of space, and their foodchain contains no short loops.
Don't get me wrong - I absolutely adore almost anything done with chick-peas too!
FatPhil
What proportion of Europe do you think _doesn't_ have a bank holiday on 1st May?
To save you time, you don't need to check Italy, Germany, France, Spain, Poland, Sweden, Finland and Switzerland, as I know for a fact that they do -- I have live(d) or work(ed) there, or consorted on May 1st with nationals therefrom.
FatPhil
methane's odourless.
OT, but fun:
. php?mission_id=57>
<URL:http://www.improveverywhere.com/mission_view
Do us a favour -- either get a clue and RTFA, or don't spout crap. Preferably both.
For spammers, I've always promoted the "Death by a Million Paper Cuts" method. Everyone wronged merely has to contribute one paper cut, but they will all add up.
FatPhil
So does target's CEO say they don't need customers?
Gerald Ratner, head of the Ratner chain of jewellers, said
things pretty much as stupid back in the 90s.
It didn't do him much good in the short term at all though.
FatPhil
You seem to have a better insight than most!
There is a technique used for both (a) recovering from noise and (b) assuring that if the noise is caused by evesdroppers then the information they can recover tells them nothing about the original payload.
Basically you have to make the message larger, like any error correcting scheme, but such that if you assume no more than a certain proportion of bits are leaked, there are equiprobable encodings for every possible symbol that could contain such leaked bits.
Hmmm, here's a very clumsy example which doesn't handle the SNR part:
encode 0 as any of { 00, 11 },
and 1 as { 01, 10 }
Let's say the source stream is '0', encoded to '00'
Let's also assume that none of the signal is lost to noise or evesdropping.
However, let's also assume the evesdropper did grab one bit.
If the evesdropper knows the message is 'x0' then it could have been '00' or '10', so the evesdropper has gained no info.
Likewise if he knows the transmitted message is 'x1' or '0x' or '1x', he gains no information about the plaintext.
Traditionally you would compress the plaintext before expanding it using these kinds of schemes.
Such techniques are typically called "privacy amplification". I would hope that Helger Lipmaa has some stuff on his website, as he seems to have lots of useful things to refer to. (I think he may be at ut.ee now, rather than cs.hut.fi)
FatPhil
Alas it's not as easy as that.
It is _impossible_ to distinguish an evesdropper causing a -X dB perturbation on the transmitted signal from line noise that causes a -X dB perturbation on it.
It's stormy - you're getting a lower SNR on your fibre line today - are you sure that a small fraction of your bits are not being evesdropped.
Of course, that's why you use privacy amplification techniques to make small leaks of the signal to an adversary give no useful information about the real payload.
FatPhil
From a laws-of-physics and mathematical standpoint it's perfect.
However, we're neither wave equations nor formulae - we're humans in the real world, using high power electronic devices that broadcast their state and their changes of state to their environs. Nothing about OPTs and QKE helps preserve the secrecy of the data once it's been decrypted and, say, stuck on a hard disk, or displayed as text on screen.
FatPhil
"""
Because the pad is pure randomness, it is possible (using existing methods) to send the pad by public key encryption
"""
You would never do this. It's "just plain silly".
If you trust PKC, then you don't need OTP.
FatPhil
Oh dear - you've obviously never heard of padding.
Message length may be part of the information in the message. (If length is constant, then it's not; if it's variable, then it is.) That too must be removed from pain sight.
FatPhil
If Eve is _already_ pretending to be Bob to Alice, or
if Bob and Alice have no way of authenticating each other.
However, on the whole, if you're building a site-site fibre link, then
you'd expect the two parties to already be able to authenticate each
other.
At some point there needs to exist a side channel. But after that,
assuming you use the provably secure _primitive_ using an accepted protocol, then you will have provably unbreakable and un-MITM-able crypto.
Phil
I find nothing explains such complicated issues better than just simply providing a program with its source code that can be examined and tinkered with:
;-)
<URL:http://fatphil.org/crypto/QKE.html>
Heheheh
FatPhil