No. XOR gates cannot be used to build anything but linear transforms (or, since you can use the power rails, affine transforms) on your data, represented as a vector of Z_2^n.
In other words, each output of any network of XOR gates is the XOR of some set of the inputs, or its complement.
NAND or NOR, as well as some messier gates like (A AND NOT B), can build any 2-value boolean logic.
Just to educate the masses further, a quantum CNOT(controlled NOT)/XOR gate is not equivalent to a classical XOR gate. A CNOT gate includes a control bit, and retain the information of this bit at the end of the gate.
Of course, in a classical system, you're generally feeding more than one particle into each gate lead, which means you can return that "control bit" simply by tapping the appropriate gate input.
What is fun in Quantum Computing is that you do not need a lot of basic gates(AND, OR, XOR, NOT, etc.), you only need a small number of basic gates to make up the Universal gate.
In classical computing, you only need the NAND gate to make all the other gates. However, you usually make most real circuits out of NAND, NOR, NOT, AOI/OAI (AND-OR-INVERT) and Transmit gates (building blocks for MUXes).
Furthermore, ALL the elementary gates in QC are reversible!! Unlike classical gates, like XOR, the quantum CNOT, for example, is fully reversible.
Um... correct me if I'm wrong... but this would imply that it is impossible to construct an AND gate, right? Because it is impossible to extend a conventional AND gate (or OR, or NOR, or NAND) so that it is reversible. That is, you can't make a reversible gate that takes n inputs to n outputs, such that one of the outputs is the AND of all the inputs.
In fact, any reversible network of gates would have to effect an invertible affine transform on Z_2^n. That sort of limits what you can do with the data, doesn't it (also implies that CNOT is the only basic gate)? Or is there some special quantumy-thing that I'm missing here?
No, that's a maximum rating. It doesn't draw nearly 400W if you aren't going to use all that, otherwise it would have to dissipate 200W on its own, which it certainly cannot.
Well, if someone really dumb had . before/bin in their path, and you dropped this into a public directory they used (say/tmp), then creating this modified cat script would change the command
cat foo bar baz
into
rm -f cat foo bar baz
thereby removing whatever they wanted to see, along with the script itself. This would be very puzzling, as several files would have disappeared for no apparent reason. Of course, this bit of deviltry doesn't work unless . is first, but you can change it from cat to cta and hope they make a typo.
If you approach me you had better be able to prove that you have not altered your grades.
Um... depending on the exploit, proving that he did not alter his grades wouldn't be trivial. If the security hole leaves no traces in the logs, it's pretty much impossible to prove that you didn't change something. If you want to prove that you *did* change it, you can get a trusted source to make and sign a "before and after" shot of it (tougher on something you do not have legitimate access to), but if you want to prove that you *didn't* change it, you basically must diff with the backup tapes.
I have dealt with students bringing up exploits to me that they have found work in our system. First I have to verify their claim, second I have to consider the damage they may have done (purposefully or not). If this means a call to security then I am obligated to do that. After that I have to consider fixing my system and damage control.
Oh, come on man, you can do better than that! You're the sysadmin. You have the logs. If anyone can check whether a student damaged something, it's you. And given that the university may take disciplinary action if they find out the student's identity, you shouldn't disclose that identity. They're doing you a service, aren't they?
Including a PGP key is sort of overkill. Just include the hash of some random number, concatenated with your name. Your knowledge of that value proves your hand in the exploit. A key has basically no advantages over a hash in this case, as either could be changed by some party wishing to deny your involvement.
If you really wanted to make sure you could prove your involvement (IMHO there is little point in this), you could mail it through a timestamping service (eg stamper@itconsult.co.uk); they will publish (and mail to you, if you specify an account; maybe Hotmail?) a signature that they remailed it on that date.
Come on, you must be kidding me. It's a college. There will be a public terminal SOMEWHERE, and there's not much they can do if you send from there, especially if it's via forged SMTP or anonymous mailer. It doesn't have to hide you IP if you use a public terminal accessable to any student. And no, most keyboards do not leave fingerprints.
At worst, you can boot a terminal in single-user or with Knoppix and
telnet smtp.yourcollege.edu 25 HELO anonymouscoward MAIL FROM: grades@exploit.com RCPT TO: dean DATA Subject: Grades Exploit I've discovered an exploit in the grades system... [details] If you need more information, post and sign a notice that you won't bust me [state more legally], and I'll reveal my identity.
Sincerely, [MD5(your name + long random password)] .
4. GNC also sells soy-protein. On the protein utilization scale, soy has the lowest value. ie. just 30-40% of soy can be utilized by body, the rest is excreted. Besides, soy protein intake leads to man-boobs.
Uh-oh... I can see a new advertising campaign coming...
ENLARGE YOUR BREASTS! INCREASE YOUR CUP BY 1-3 SIZES WITH OUR NEW HERBAL SOY BREAST ENLARGEMENT PILLS!!!
This is more true than you might guess. Hollerith, the founder of IBM, is famous for building an electronic census-analyzing machine for the goverment.
Re:What's the power curve on that?
on
Replacing SMTP?
·
· Score: 1
Well, having a spambox is ok, but then you still have to check through 100 spam messages a day... remember these people's email addresses are posted on their website...
What's the power curve on that?
on
Replacing SMTP?
·
· Score: 3, Insightful
A combination of white lists/black lists, and Baysian filtering stops so close to 100% of spam that it's really silly for anyone to be bitching about spam these days. I don't GET any spam anymore - 0. Not 0.001%, 0 - the integer 0, as in none.
Have you done the power-curve analysis on that? My mother works at a law firm, and they once tried to install a spam filter. It was state-of-the-art, with Bayesian filtering, and white/black lists, and additional whitefilters on top. It blocked most (not all) of their spam. But it also blocked some tiny fraction of legitimate messages.
Even if you have the (extremely impressive) power curves of Paul Graham's Plan for Spam -- and that was on a very well-trained Bayesian filter written by a coding genious -- it is Not Good Enough when missing a legit email could get you sued for millions. Either you risk blocking legit email, or else you have to wade through a pile of spam bigger than that legit email... either way, another protocol would be nice.
I think that you are forgetting the buffer cache in your assessment of the efficiency of swapping. As well as keeping programs in memory, you also want to keep commonly-accessed disk blocks in memory for as long as possible. With a large hard drive, this can take up quite a lot of memory. Swapping application memory that hasn't been used in a long time for cache space that will be reused soon causes a substantial increase in speed.
I run Linux at home without a swapfile (I dual boot and the Linux partition is 1GB), and I've found that when the computer has been running for a while without too many applications, many commands can complete almost instantly and without accessing the disk.
But when several large apps are sleeping (say I have 2 X sessions open, one with my sister's apps while she's afk), commands take much longer to complete. This doesn't happen on other computers, with a swapfile, because the cache stays up-to-date.
It is worth noting that some OS researchers these days are regarding memory as a unified object, consisting of disks, RAM and CPU cache, and attempting to maximize the performance of this single object under standard workloads. This view is nonsensical without swapping capabilities.
I haven't used Eclipse or netbeans. I'll have to check them out, especially if they are as good as Visual Studio.
And no, Safari is not Konqueror ported to OSX. Safari's Webcore is KHTML ported to OSX, and *very much improved.* I use Konqueror at home (no OSX there, only Linux), and Safari renders many, many more pages correctly. Hopefully the next version of Konq will use Apple's changes to KHTML.
And that's just the core. Safari's interface is much superior to Konqueror's, at least for web pages (Konq handles other things besides the web; it's also KDE's default local file browser). Safari's googlebar, tabbed browsing, status bar, and bookmarks are all just great. I take issue with the brushed metal, but hey, there are hacks around that.
I've used Konqueror, Mozilla, Chimera/Camino, MSIE (Mac and Windows), Phoenix/Firebird, and Safari extensively, and occasionally Opera and Galeon, and Safari is by far the best from what I've seen. YMMV.
Disclaimer: I use BBEdit and emacs (learned it before I knew about vi, so I have no idea which one's better) to program at school, as I have a Mac.
I would say that what the OSS programming community needs is a better IDE. (This is distinct from what the OSS community as a whole needs, which I would say is a better desktop environment, to attract computer newbs.) While emacs / vim are pretty good programs, they are nothing to Visual Studio. VS's method-name completion, function argument tooltips, etc make it much easier to program in than either emacs or BBEdit. MS also has better debuggers and better compilers than the open-source community.
Although I use Linux all the time (right now, in fact), I must admit that their interface is mediocre compared to OSX (especially for nongeeks; I do wish OSX had virtual desktops and focus-follows-mouse), their IDEs are mediocre compared to Microsoft's, their browsers are mediocre compared to Safari, etc.
If the Open Source community is to spread, we must be better than the competition, or at least comparable. And right now, we are barely comparable, and that expensive proprietary software is worth it in many cases.
While you're rgiht that Linux is not perfectly secure, you must admit that those are rather different classes of vulnerability. The two Windows ones were a remote root exploits in the default configuration, and a root exploit that could be easily used by a webpage/email. The Linux ones were mostly DOS vulnerabilities, most of which would require substantial access (ie, a shell account or more) to exploit, and many of which are not present in a standard configuration (eg, you have to have NFS turned on).
I would say that considering the kinds of vulnerabilities we're talking about, Linux's track record is at least as good as Windows' in this department.
Slashdot Mirror
Adding to the litany of bad "people kill people" jokes:
Hacking tools don't kill servers, malformed packets kill servers.
Guns don't kill people, bullets kill people.
Guns don't kill people, I kill people >:-)
No, it's an RPG wurm. That +3 Fireproof armor will come in handy.
No. XOR gates cannot be used to build anything but linear transforms (or, since you can use the power rails, affine transforms) on your data, represented as a vector of Z_2^n.
In other words, each output of any network of XOR gates is the XOR of some set of the inputs, or its complement.
NAND or NOR, as well as some messier gates like (A AND NOT B), can build any 2-value boolean logic.
Just to educate the masses further, a quantum CNOT(controlled NOT)/XOR gate is not equivalent to a classical XOR gate. A CNOT gate includes a control bit, and retain the information of this bit at the end of the gate.
Of course, in a classical system, you're generally feeding more than one particle into each gate lead, which means you can return that "control bit" simply by tapping the appropriate gate input.
What is fun in Quantum Computing is that you do not need a lot of basic gates(AND, OR, XOR, NOT, etc.), you only need a small number of basic gates to make up the Universal gate.
In classical computing, you only need the NAND gate to make all the other gates. However, you usually make most real circuits out of NAND, NOR, NOT, AOI/OAI (AND-OR-INVERT) and Transmit gates (building blocks for MUXes).
Furthermore, ALL the elementary gates in QC are reversible!! Unlike classical gates, like XOR, the quantum CNOT, for example, is fully reversible.
Um... correct me if I'm wrong... but this would imply that it is impossible to construct an AND gate, right? Because it is impossible to extend a conventional AND gate (or OR, or NOR, or NAND) so that it is reversible. That is, you can't make a reversible gate that takes n inputs to n outputs, such that one of the outputs is the AND of all the inputs.
In fact, any reversible network of gates would have to effect an invertible affine transform on Z_2^n. That sort of limits what you can do with the data, doesn't it (also implies that CNOT is the only basic gate)? Or is there some special quantumy-thing that I'm missing here?
No, it's not. Err... it's not NOT... errr... The gate was a "controlled NOT gate", also known as XOR.
No, that's a maximum rating. It doesn't draw nearly 400W if you aren't going to use all that, otherwise it would have to dissipate 200W on its own, which it certainly cannot.
If he's truly the BOFH, he'll find a way to pin the bandwidth spike on his PHB's pr0n downloading.
Yeah, that would explain the UPSTREAM spike...
If you approach me you had better be able to prove that you have not altered your grades.
Um... depending on the exploit, proving that he did not alter his grades wouldn't be trivial. If the security hole leaves no traces in the logs, it's pretty much impossible to prove that you didn't change something. If you want to prove that you *did* change it, you can get a trusted source to make and sign a "before and after" shot of it (tougher on something you do not have legitimate access to), but if you want to prove that you *didn't* change it, you basically must diff with the backup tapes.
I have dealt with students bringing up exploits to me that they have found work in our system. First I have to verify their claim, second I have to consider the damage they may have done (purposefully or not). If this means a call to security then I am obligated to do that. After that I have to consider fixing my system and damage control.
Oh, come on man, you can do better than that! You're the sysadmin. You have the logs. If anyone can check whether a student damaged something, it's you. And given that the university may take disciplinary action if they find out the student's identity, you shouldn't disclose that identity. They're doing you a service, aren't they?
Including a PGP key is sort of overkill. Just include the hash of some random number, concatenated with your name. Your knowledge of that value proves your hand in the exploit. A key has basically no advantages over a hash in this case, as either could be changed by some party wishing to deny your involvement.
If you really wanted to make sure you could prove your involvement (IMHO there is little point in this), you could mail it through a timestamping service (eg stamper@itconsult.co.uk); they will publish (and mail to you, if you specify an account; maybe Hotmail?) a signature that they remailed it on that date.
At worst, you can boot a terminal in single-user or with Knoppix and
They're 100% machine language. As in, they're patterns of charged and discharged capacitors in RAM, as well as patterns of current flow in the CPU.
So don't quibble.
4. GNC also sells soy-protein. On the protein utilization scale, soy has the lowest value. ie. just 30-40% of soy can be utilized by body, the rest is excreted. Besides, soy protein intake leads to man-boobs.
Uh-oh... I can see a new advertising campaign coming...
ENLARGE YOUR BREASTS! INCREASE YOUR CUP BY 1-3 SIZES WITH OUR NEW HERBAL SOY BREAST ENLARGEMENT PILLS!!!
no text
OpenBSD is used for firewalls pretty often in the academic world, although I couldn't speak for the commercial one...
This is more true than you might guess. Hollerith, the founder of IBM, is famous for building an electronic census-analyzing machine for the goverment.
Well, having a spambox is ok, but then you still have to check through 100 spam messages a day... remember these people's email addresses are posted on their website...
A combination of white lists/black lists, and Baysian filtering stops so close to 100% of spam that it's really silly for anyone to be bitching about spam these days. I don't GET any spam anymore - 0. Not 0.001%, 0 - the integer 0, as in none.
Have you done the power-curve analysis on that? My mother works at a law firm, and they once tried to install a spam filter. It was state-of-the-art, with Bayesian filtering, and white/black lists, and additional whitefilters on top. It blocked most (not all) of their spam. But it also blocked some tiny fraction of legitimate messages.
Even if you have the (extremely impressive) power curves of Paul Graham's Plan for Spam -- and that was on a very well-trained Bayesian filter written by a coding genious -- it is Not Good Enough when missing a legit email could get you sued for millions. Either you risk blocking legit email, or else you have to wade through a pile of spam bigger than that legit email... either way, another protocol would be nice.
I think that you are forgetting the buffer cache in your assessment of the efficiency of swapping. As well as keeping programs in memory, you also want to keep commonly-accessed disk blocks in memory for as long as possible. With a large hard drive, this can take up quite a lot of memory. Swapping application memory that hasn't been used in a long time for cache space that will be reused soon causes a substantial increase in speed.
I run Linux at home without a swapfile (I dual boot and the Linux partition is 1GB), and I've found that when the computer has been running for a while without too many applications, many commands can complete almost instantly and without accessing the disk.
But when several large apps are sleeping (say I have 2 X sessions open, one with my sister's apps while she's afk), commands take much longer to complete. This doesn't happen on other computers, with a swapfile, because the cache stays up-to-date.
It is worth noting that some OS researchers these days are regarding memory as a unified object, consisting of disks, RAM and CPU cache, and attempting to maximize the performance of this single object under standard workloads. This view is nonsensical without swapping capabilities.
... or you're willing to swap drives mid-photo-spree. And this shouldn't be too much of a problem; who takes 150 RAW 6MPixel images in a row?
I haven't used Eclipse or netbeans. I'll have to check them out, especially if they are as good as Visual Studio.
And no, Safari is not Konqueror ported to OSX. Safari's Webcore is KHTML ported to OSX, and *very much improved.* I use Konqueror at home (no OSX there, only Linux), and Safari renders many, many more pages correctly. Hopefully the next version of Konq will use Apple's changes to KHTML.
And that's just the core. Safari's interface is much superior to Konqueror's, at least for web pages (Konq handles other things besides the web; it's also KDE's default local file browser). Safari's googlebar, tabbed browsing, status bar, and bookmarks are all just great. I take issue with the brushed metal, but hey, there are hacks around that.
I've used Konqueror, Mozilla, Chimera/Camino, MSIE (Mac and Windows), Phoenix/Firebird, and Safari extensively, and occasionally Opera and Galeon, and Safari is by far the best from what I've seen. YMMV.
Disclaimer: I use BBEdit and emacs (learned it before I knew about vi, so I have no idea which one's better) to program at school, as I have a Mac.
I would say that what the OSS programming community needs is a better IDE. (This is distinct from what the OSS community as a whole needs, which I would say is a better desktop environment, to attract computer newbs.) While emacs / vim are pretty good programs, they are nothing to Visual Studio. VS's method-name completion, function argument tooltips, etc make it much easier to program in than either emacs or BBEdit. MS also has better debuggers and better compilers than the open-source community.
Although I use Linux all the time (right now, in fact), I must admit that their interface is mediocre compared to OSX (especially for nongeeks; I do wish OSX had virtual desktops and focus-follows-mouse), their IDEs are mediocre compared to Microsoft's, their browsers are mediocre compared to Safari, etc.
If the Open Source community is to spread, we must be better than the competition, or at least comparable. And right now, we are barely comparable, and that expensive proprietary software is worth it in many cases.
Right. Technically. But I would count a vulnerability in something common to most distributions (say, sudo), as a "Linux Vulnerability".
While you're rgiht that Linux is not perfectly secure, you must admit that those are rather different classes of vulnerability. The two Windows ones were a remote root exploits in the default configuration, and a root exploit that could be easily used by a webpage/email. The Linux ones were mostly DOS vulnerabilities, most of which would require substantial access (ie, a shell account or more) to exploit, and many of which are not present in a standard configuration (eg, you have to have NFS turned on).
I would say that considering the kinds of vulnerabilities we're talking about, Linux's track record is at least as good as Windows' in this department.