> The problem with Itanium is that Intel bet they could not only get everyone to abandon literally billions of lines of already paid for X86 code, but that they could build a compiler able to keep it fed and do all the heavy lifting and in the end they just weren't able to deliver compared to X86-64.
The problem with VLIW (and RISC) is that the compiler has to optimise the code for the CPU architecture, but the CPU architecture changes from generation to generation. You optimise for one chip, and it will run slow on the next generation, and not at all on the previous. SPARC had that problem, but they managed to produce an excellent compiler with lots of switches for tuning the compromise. But software developers don't like this kind of inevitable incompatibility, and they much prefer the much more gentle progress of x86.
It seems to me that the rate at which they fix bugs is ever increasing, maybe even exponentially.
That means that either a) they introduce new bugs faster than the fix old ones (also exponentially growing), b) there is an infinite number of bugs in Flash, c) most of the fix do not actually fix the bug in question, or d) they will run out of bugs to fix soon!
Now you can venture a guess as to what is actually happening.
"In conclusion, the main thing we did wrong when designing ATM security systems in the early to mid-1980s was to worry about criminals being clever
And you didn't do a good job there either: it does not take a lot of clever to copy a magnetic strip, and even skimming devices are actually quite a simple technology.
Sometimes you have a picture that you want to know more about, and so far the only option was to take a picture of your screen with a smart phone. A desktop based image search should be much easier to use.
> Banks create money because lending creates money.
No they don't. Banks get money from someone, and lend it to someone else. They do not create money.
In most societies only central banks can create money, and only in defined circumstances (usually interest on deposits). Of course in recent years "creating money" has become fashionable again, and they call it "quantitative easing" and stuff.
Heck if I could print money, I would apply a "quantitative ease" to my "deficit", too.
Paypal could be a safer option - with online verification, a secret code in your mobile, and a password to enter. I think they may be on to something - something that certainly involves a good cut for them:-)
The funny thing is that Linux just works - it picks up IPv6 from the DHCP server, and Firefox and Google Chrome just chose the right protocol (although I still think that IPv4 should be the default - at least for now). At home I need to use a tunnel - neither the router nor the ISP provide native support for IPv6, but again it just works.
Windows XP (we skipped Vista, and 7 is still being rolled out) however is unable to deal with IPv6 correctly. In a year that problem should be history for us, but I am sure many many companies have the same issue with Windows. Until Windows XP is truly dead, IPv6 will remain a niche protocol.
See, *ARM* is the key word here. It is not in the summary, and that makes it an absolute terrible summary. The fact that all the ARM fan boys don't realise that is quite illuminating in a way.
Those are two different questions, hm? As far as I can see nobody seems to seriously propose that this thing is actually useful. The discussion is about whether it is just a very bad half digital half analog computer with a lot of noise, or whether quantum effects have to be used to explain its behaviour. That behaviour would be a correlation of the noise beyond what classical theory predicts.
So even if you can, after long hours of observation, observe a tiny quantum effect in this device, does that mean it is useful? No, most like you just have noise that is slightly different from the classical model.
> If this attack is like previous timing attacks I've read about, then no, adding a fixed delay doesn't help, and adding a random delay doesn't help enough.
We already have a random delay, it is called "the internet". And of course it only makes it more difficult to use this side channel attack, but not impossible.
However, you could also respond a fixed delay after the *start* of the calculation. That should eliminate any timing difference in the algorithm itself.
You still have access to the timing side channel if you are logged on, but a) that is a rare situation, and b) the data you get from top is typically pretty bad.
True, although they were usually not taught. Of course once you start using pointers, Pascal is no longer safe... not as bad as zero terminated strings in C, but a security issue non the less.
> Actually not true - Vista/Win7 support Transactional NTFS which gives you integrated filesystem transactions. Changes to a file or folder in the transaction aren't visible until the transaction is committed and you can specify that your read transaction doesn't see changes once it begins so you get a consistent view of the FS.
So why is it still impossible for an installer to update an open file? Is it because most installers are written by incompetent interns (which is quite possible...)?
Under UNIX the very same operation is no problem at all, and it is all down to the concept of an inode.
Get a USB to serial cable, and the problem on the laptop side is solved. Of course there is still an issue with embedded devices, but most of them have a JTAG connector of sorts somewhere on the board.
Dropbox, subversion, time machine, snapshots - it is just that we have more solutions now than back in the days, and you have to pick one.
> Creating a new empty file and then writing it did not make the file visible to other processes until the file was closed and committed.
Semantics - UNIX still does it that way (if you want to), Windows does not, mostly because it does not have inodes.
> Rings of protection
Called sandboxes nowadays. Yes, a number of systems had a more elegant implementation, including OS/360 and I believe Plan 9. The x86 architecture just does not make this easy, and that is indeed something worth lamenting.
> Safe, fast languages
Pascal was and still is safe, as is LISP. But those languages are safe because they miss important element such as pointers. Java is also reasonably safe, because it only has references, no pointers.
> Capability machines
You can still do that, but most people just want to get stuff done, and not worry all day long about fine grain security features. If you have a team of specialists to maintain a single computer, you can do it.
> I miss the subtle chirping sounds that my old MiniScribe MFM 20MB hard drive made.
Wasn't that just caused by bad bearings? Modern fluid bearings and electronically commutated DC motors make most hard disks rather quiet in comparison.
> the greatest work station epiphany i recently had involved turning my 9:16 monitor 90 degrees
I agree - I want more height first of all, and more width second. Turning the screen around is good way to address that, but it has two problems: a) the driver does not always like it, and b) it breaks sub-pixel rendering.
My perfect screen would be 5:4 portrait with the usual lateral RGB sub-pixel arrangement. Anybody cares to produce one for a reasonable price?
Two portrait screens side by side work nicely, although arguable one 16:10 screen that is twice a big is even better. Sadly even 16:10 is dying out in favour of 16:9. What's next? 2.35:1 cinemascope?
What a whiner. All those things were around in the 80s, and all those things are still around. Most are a lot cheaper than they were in the 80s (spending a few grand on a PC was considered the minimum). But most of them have not made it into the bargain low cost PCs that dominate the market now for less than grand total.
You can still have the control key in the middle (where caps-lock usually sits), it is just a question of software configuration. You can still get clickedy-clack keyboards, and for less money than in the days of old. Most people prefer soft touch keyboards, because they do not have the same tendency to cause RSI, but keyboards have always been and will always be a matter of taste.
Height adjustable screens are still around, but they cost 5 bucks more. That being said, there is just no comparison even between the best CRTs in the 80s and even a cheap modern TFT screen. Geometry distortion, convergence issues, masks - all gone and replaced by perfectly spaced, much higher resolving, and mostly flicker free sub-pixel.
All the advanced GUI features are still available today, but most of them have always been and will always be difficult to implement on Windows. Windows caters for the masses, not for the IT expert. But you can have as many views of the same document open in Word as you like - you just need to use separate windows. (Excel is a different matter, and it could have evolved more than it actually has in some ways.)
And the flat memory model of the C64 (which actually was not flat, it was banked and quite complicated!) is still available. It is called kernel mode. But you are not man enough to use it... Remember that real programmer just need a hex keyboard and a system monitor.
That being said, there have been some real setbacks. TFT monitors have a delay now, and colour rendition was universally better on a colour CRT than a TFT. Programs seem to take longer to respond than they need to. Modal pop-ups seem to be on the rise, especially on Windows and MS programs. OLE never reached its potential, and the implementation seems to get worse and worse with new programs. And still on the whole, productivity of IT systems is up, and continuing to increase.
> Investing for the future. If bitcoin ever really takes off the deflation will be massive.
Fair enough, but what is it backed by? Currencies are strong if they are either backed by some other value (gold standard or more likely another currency basket now), or they are used by an economy with decent inflation control.
I can see neither for the bitcoin. And scarcity alone does not create value.
Some even say that Windows 7 was Vista done right, and that it should have been a free update.
But of course the logic of the market was different: people didn't like Vista, so they wanted a new OS. Especially the corporate buying behaved exactly like sheep.
Slashdot Mirror
> The problem with Itanium is that Intel bet they could not only get everyone to abandon literally billions of lines of already paid for X86 code, but that they could build a compiler able to keep it fed and do all the heavy lifting and in the end they just weren't able to deliver compared to X86-64.
The problem with VLIW (and RISC) is that the compiler has to optimise the code for the CPU architecture, but the CPU architecture changes from generation to generation. You optimise for one chip, and it will run slow on the next generation, and not at all on the previous. SPARC had that problem, but they managed to produce an excellent compiler with lots of switches for tuning the compromise. But software developers don't like this kind of inevitable incompatibility, and they much prefer the much more gentle progress of x86.
> It is a free product and they don't really want to spend a lot of money on it ?
That would be a good reason. Unfortunately Acrobat (the $$$ product) has even more security issues.
It seems to me that the rate at which they fix bugs is ever increasing, maybe even exponentially.
That means that either
a) they introduce new bugs faster than the fix old ones (also exponentially growing),
b) there is an infinite number of bugs in Flash,
c) most of the fix do not actually fix the bug in question, or
d) they will run out of bugs to fix soon!
Now you can venture a guess as to what is actually happening.
"In conclusion, the main thing we did wrong when designing ATM security systems in the early to mid-1980s was to worry about criminals being clever
And you didn't do a good job there either: it does not take a lot of clever to copy a magnetic strip, and even skimming devices are actually quite a simple technology.
Never ever trust the browser!
Epic fail, but of course we would not expect any less from Citibank.
Sometimes you have a picture that you want to know more about, and so far the only option was to take a picture of your screen with a smart phone. A desktop based image search should be much easier to use.
> Banks create money because lending creates money.
No they don't. Banks get money from someone, and lend it to someone else. They do not create money.
In most societies only central banks can create money, and only in defined circumstances (usually interest on deposits). Of course in recent years "creating money" has become fashionable again, and they call it "quantitative easing" and stuff.
Heck if I could print money, I would apply a "quantitative ease" to my "deficit", too.
Actually they are a bank - registered in Luxembourg - since 2007.
Paypal could be a safer option - with online verification, a secret code in your mobile, and a password to enter. I think they may be on to something - something that certainly involves a good cut for them :-)
The funny thing is that Linux just works - it picks up IPv6 from the DHCP server, and Firefox and Google Chrome just chose the right protocol (although I still think that IPv4 should be the default - at least for now). At home I need to use a tunnel - neither the router nor the ISP provide native support for IPv6, but again it just works.
Windows XP (we skipped Vista, and 7 is still being rolled out) however is unable to deal with IPv6 correctly. In a year that problem should be history for us, but I am sure many many companies have the same issue with Windows. Until Windows XP is truly dead, IPv6 will remain a niche protocol.
http://xkcd.com/892/
> Origen = an ARM based motherboard
See, *ARM* is the key word here. It is not in the summary, and that makes it an absolute terrible summary. The fact that all the ARM fan boys don't realise that is quite illuminating in a way.
The article is not even getting the technical facts right, much less the grammar. Shouldn't we expect better from PCPro? No, I guess not.
> It sure seems to be a Quantum Computer to me.
> It either works or it doesn't.
Those are two different questions, hm? As far as I can see nobody seems to seriously propose that this thing is actually useful. The discussion is about whether it is just a very bad half digital half analog computer with a lot of noise, or whether quantum effects have to be used to explain its behaviour. That behaviour would be a correlation of the noise beyond what classical theory predicts.
So even if you can, after long hours of observation, observe a tiny quantum effect in this device, does that mean it is useful? No, most like you just have noise that is slightly different from the classical model.
> If this attack is like previous timing attacks I've read about, then no, adding a fixed delay doesn't help, and adding a random delay doesn't help enough.
We already have a random delay, it is called "the internet". And of course it only makes it more difficult to use this side channel attack, but not impossible.
However, you could also respond a fixed delay after the *start* of the calculation. That should eliminate any timing difference in the algorithm itself.
You still have access to the timing side channel if you are logged on, but a) that is a rare situation, and b) the data you get from top is typically pretty bad.
There were 1600x1200 screens, and I was tempted. But 1920x1200 is certainly better, especially if you work with two windows side by side.
Sadly even 1920x1200 is getting rare. I bought one recently, but it could not display native resolution, WTH?
> Pascal has pointers
True, although they were usually not taught. Of course once you start using pointers, Pascal is no longer safe... not as bad as zero terminated strings in
C, but a security issue non the less.
> Actually not true - Vista/Win7 support Transactional NTFS which gives you integrated filesystem transactions. Changes to a file or folder in the transaction aren't visible until the transaction is committed and you can specify that your read transaction doesn't see changes once it begins so you get a consistent view of the FS.
So why is it still impossible for an installer to update an open file? Is it because most installers are written by incompetent interns (which is quite possible...)?
Under UNIX the very same operation is no problem at all, and it is all down to the concept of an inode.
Get a USB to serial cable, and the problem on the laptop side is solved. Of course there is still an issue with embedded devices, but most of them have a JTAG connector of sorts somewhere on the board.
> File revisions
Dropbox, subversion, time machine, snapshots - it is just that we have more solutions now than back in the days, and you have to pick one.
> Creating a new empty file and then writing it did not make the file visible to other processes until the file was closed and committed.
Semantics - UNIX still does it that way (if you want to), Windows does not, mostly because it does not have inodes.
> Rings of protection
Called sandboxes nowadays. Yes, a number of systems had a more elegant implementation, including OS/360 and I believe Plan 9. The x86 architecture just does not make this easy, and that is indeed something worth lamenting.
> Safe, fast languages
Pascal was and still is safe, as is LISP. But those languages are safe because they miss important element such as pointers. Java is also reasonably safe, because it only has references, no pointers.
> Capability machines
You can still do that, but most people just want to get stuff done, and not worry all day long about fine grain security features. If you have a team of specialists to maintain a single computer, you can do it.
> I miss the subtle chirping sounds that my old MiniScribe MFM 20MB hard drive made.
Wasn't that just caused by bad bearings? Modern fluid bearings and electronically commutated DC motors make most hard disks rather quiet in comparison.
> the greatest work station epiphany i recently had involved turning my 9:16 monitor 90 degrees
I agree - I want more height first of all, and more width second. Turning the screen around is good way to address that, but it has two problems: a) the driver does not always like it, and b) it breaks sub-pixel rendering.
My perfect screen would be 5:4 portrait with the usual lateral RGB sub-pixel arrangement. Anybody cares to produce one for a reasonable price?
Two portrait screens side by side work nicely, although arguable one 16:10 screen that is twice a big is even better. Sadly even 16:10 is dying out in favour of 16:9. What's next? 2.35:1 cinemascope?
What a whiner. All those things were around in the 80s, and all those things are still around. Most are a lot cheaper than they were in the 80s (spending a few grand on a PC was considered the minimum). But most of them have not made it into the bargain low cost PCs that dominate the market now for less than grand total.
You can still have the control key in the middle (where caps-lock usually sits), it is just a question of software configuration. You can still get clickedy-clack keyboards, and for less money than in the days of old. Most people prefer soft touch keyboards, because they do not have the same tendency to cause RSI, but keyboards have always been and will always be a matter of taste.
Height adjustable screens are still around, but they cost 5 bucks more. That being said, there is just no comparison even between the best CRTs in the 80s and even a cheap modern TFT screen. Geometry distortion, convergence issues, masks - all gone and replaced by perfectly spaced, much higher resolving, and mostly flicker free sub-pixel.
All the advanced GUI features are still available today, but most of them have always been and will always be difficult to implement on Windows. Windows caters for the masses, not for the IT expert. But you can have as many views of the same document open in Word as you like - you just need to use separate windows. (Excel is a different matter, and it could have evolved more than it actually has in some ways.)
And the flat memory model of the C64 (which actually was not flat, it was banked and quite complicated!) is still available. It is called kernel mode. But you are not man enough to use it... Remember that real programmer just need a hex keyboard and a system monitor.
That being said, there have been some real setbacks. TFT monitors have a delay now, and colour rendition was universally better on a colour CRT than a TFT. Programs seem to take longer to respond than they need to. Modal pop-ups seem to be on the rise, especially on Windows and MS programs. OLE never reached its potential, and the implementation seems to get worse and worse with new programs. And still on the whole, productivity of IT systems is up, and continuing to increase.
> Investing for the future. If bitcoin ever really takes off the deflation will be massive.
Fair enough, but what is it backed by? Currencies are strong if they are either backed by some other value (gold standard or more likely another currency basket now), or they are used by an economy with decent inflation control.
I can see neither for the bitcoin. And scarcity alone does not create value.
Some even say that Windows 7 was Vista done right, and that it should have been a free update.
But of course the logic of the market was different: people didn't like Vista, so they wanted a new OS. Especially the corporate buying behaved exactly like sheep.