SDComment *parent = [[SDSlashdotCommentManager defaultManager] getCommentForUid:@"17541932"]; if (parent) { SDComment *this = [parent createReplyWithTitle:@"MOD PARENT UP"]; [parent moderateUp]; if (this) { [this autorelease]; } } /* Not sure if I'm making fun or not:D */
Why not just make more original episodes of Nova, dangit!
OTOH, there's tons of coverage of natural science, physics, applied science and anthropology, but in general not many shows on IT; this is remarkable considering how unlikely it is for the average person to see the Tevatron, but that person probably deals with computer viruses, data encryption and slow internets on a daily basis. There should be a Secret Life of Machines for computers. You could have:
The Secret Life of the Router
The Secret Life of C
The Secret Life of the Hard Drive
And so on. I don't think Hunkin would do it, so we'd have to find someone suitably cheeky but computer-friendly and having an artistic streak. Maybe Woz.
With pulsar GPS you're using changes in the arrival times of the pulses to determine if you're getting closer to the pulsar or getting farther away
Exactly. Pulsars are like VORs but without the magnetic-north synchronization signal. You can figure your velocity to or from by measuring the doppler, but you can't figure your bearing, and therefore, cannot figure your position.
By comparing the difference between your observation of a pulsars frequency and its frequency from a known reference, you can figure out your velocity to or from, but not your position. So first you need a very good reference datum, like from Earth, but even more, you need a very accurate clock on your ship so you can make good measurements of the pulsar's doppler. If you measure a few pulsars at once, you can build a resultant V from a collection of pulsars and work a solution for your V relative to your reference datum, like figuring your groundspeed in 3 dimensions, with wind in 4 dimensions. Fun fun fun.
This is good for speed, and figuring out your position this way is as easy as doing an integral of your V, but it's open-loop and little errors in your speed measurements will give you bad position fix, eventually. I was just thinking about this again today, and if you watched two pulsars at once, and measured the difference in phase between the two, given their frequency and position is known, you could plot your position along a hyperbola with the two pulsars as foci, just like LORAN. Getting a third reading would then give you a fix.
Pulsars are really a boon to celestial navigators, no doubt.
If you go 100km relative to the pulsar you might get mixed up as to which pulse to lock onto and think you hadn't moved.
You wouldn't use the phase of the pulsar signal to get a distance reading, but you could use it's frequency as a way of identifying it uniquely, and then you would triangulate your position by getting a bearing on three or four pulsars; this is essentially how the starburst pattern on the Mariner Plaque is supposed to work for aliens trying to find Earth. It gives the frequency of pulsars near the earth and their bearing to Sol relative to the galactic center. The aliens would identify the pulsars by their frequency, and then use their knowledge of the pulsars' absolute locations to work backward and triangulate the position of Sol (assuming these aliens, a million years from now, are able to backtrack the locations of all the pulsars, as they will all have moved relative to Sol and will have slowed down in their spinning, causing an error in our reported frequencies).
After having written some tools on OS X that do stuff with RTF:
RTF is well documented and you can make an RTF document on all manner of platforms (I've done it in Ruby and Cocoa), but many platforms have extended RTF in their own way in order to support special features. OS X has added a few special methods to RTF files to support Mac OS X typography, and I've noticed that different versions of Word handle document attributes (like headers and page numbers) in different ways.
RTF is great if you want to make up something quick that is ONLY formatted text, but readers have all manner of different ways of interpreting the exact appearance of tables, page layouts and margins, and there doesn't seem to be any manageable common mechanism for including images or other documents, something Word and OO.org excel(pun) at. Even HTML seems to be better at this.
I use RTF output in a few little in-house tools I have, so people can get the text+attributes they create and open them in a text editor of their choice for touching-up and delivery. When my tools have to create something that is supposed to be finished, they make PDFs.
RTF is great for interoperability, but I never expect an RTF file to contain a "finished product," unless the recipient expects quality on par with a Selectric. It is merely a relatively-open serialization format for strings with attributes.
It's not calling curl or the shell from memory, it appears (from the description) to be a return-to-libc-attack. I am not an expert on this particular thing, but a return-to-libc attack is where you use a buffer overflow to overwrite the return address of the stack frame. Under normal circumstances, the rtsp URL parser would return to his calling function, but if an overflow overwrites the return address, you can basically rewrite the stack's memory of who called the URL parser in the first place. So, instead of returning to where Quicktime called it, your computer can be tricked into returning to a different place in memory, like somewhere in libc. Libc has all kinds of dangerous functions, namely system(3), which accepts a string as an argument (which you have also put on the stack with your buffer overflow) and will run an arbitrary program on your computer (like curl, but bash and perl and ruby can do all kinds of damage).
Of note is the fact that this exploit gets around NX, because your payload need not be executable, it merely is a return address and a string to pass into libc. Also of note is that this exploit does not cause privilege escalation; any processes started by the exploit will run under the privileges of the user who clicks on the file, and you will still get a sudo-dialog if the sploit tries to do things as wheel.
If I am misreading this exploit, please correct me. They say "arbitrary code execution" in the summary.
This way you can think of CmdLin as an object, new is just a method that returns a new instance. Voila, one keyword removed, and much cleaner syntax. This would not even offend the Java people because they can have the same syntax with a static method.
D doesn't appear to have meta-classes, so classes are not objects, so an operator may be a better choice for instantiating than a static method; D also appeas not to have dynamic class loading, either. If you're all about the Class.new or [Class new] way of life, you may have to stick with Ruby or Objective-C or Smalltalk.
There is no such thing as an interpreted language. There are languages, and there are interpreters and compilers.
An interesting point, theoretically, but if you write a compiler for a language that has an eval() function (basically any dynamic interpreted language) you would have to include a complete compiler for the language in the language's runtime environment. This also applies to any language that does call- or pass-by-name (since this is just an application of eval() ).
You could compile such programs, but the runtime infrastructure (on any Intel/PowerPC type machine) would be huge, with dispatch tables serialized into the binary to allow relinking names back to memory locations, and a JIT compiler to allow eval to work. The call stack would probably have to be abstracted, and the thing just starts looking like a platform-optimized interpreter, with the added requirement that you have to compile your scripts before running them. (It actually starts looking like Objective-C, JUST KIDDING!:P )
This applies to current hardware arch, but if you built a CPU that, say, ran Java natively, or at least handled strings of arbitrary length as quickly as an x86 handles a UInt32, then an compiler of a dynamic language begins to make more sense.
I think this site is just plain cruel to take the piss out of people who have had severe accidents with fatal results
Just remember, for every moron that tries to light a campfire in an ordinance dump, China kills a democrat, a stray not-so-smart-bomb kills an Iraqi math whiz, and a chemical plant releases phosgene into a river valley.
Stupid people kill themselves a hundreth as often as they kill others, either by accident or malice. And in the second case, it's always the smart people they kill first.
It's an interesting suggestion, but you have to download an entire HTML-XHTML page before the entities can be interpreted, so even if your browser won't display nsfw content, you will still have transferred it and any software tracking your downloads will catch you looking at stuff, even if you didn't see it; the data would definitely find its way onto the proxy server, unless the proxy server honored the nsfw tag and didn't cache these. An "nsfw" attribute on an [a href="" rel="nsfw"] would be the only way to screen your content before you pull it - there could be a firefox plugin that turns these links red, or puts a stop sign next to them, er something.
It's one thing for your boss to see you with a dirty web page, another for them to have a record of you going to it. This solution does nothing for the latter.
Apple already did number 1, it was called "iTV" at WWDC. (You said "announce plans" and that's exactly what they did.)
How about this one: In the wake of an accounting scandal, Apple is found guilty corporately of fraud and is broken-up into an Computer Systems company and a media delivery company. It'd be ironic that after all these years Apple got broken for shady business practices before you-know-who.
Actually high wages for parliamentarians are meant to make them less inclined to take bribe money
That is a good theory, but in the US every 2 or 6 years our parliamentarians are hit with a multi-million dollar liability, known as a "Campaign." The legislators are up to themselves to raise the money for these, and there are no caps on how much they or their opponent can spend.
If you want to make legislators bribe-proof, you have to make it so that they need for no money in the course of their work, which means paying them well, enough to maintain a domicile in the capital, and strictly capping campaign spending (capping fundraising, and all the exceptions and codicils on that, and the attendant free-expression issues, gets more and more unworkable all the time).
I would say all campaigns should be publicly funded, private donations forbidden, and equal money to the top 3 primary victors, but most Americans consider a campaign donation a form a free speech, and thus beyond legitimate restraint. (I think this is bullshit, but there we are.)
I forget the date, it might have been in '05 even, but when they held the first Senate hearings after the Terrorist Surveillance Program (also known as the NSA Warrantless Wiretap program), Charles Shumer asked Alberto Gonzalez, point blank:
Can you tell me right now that you are not opening US First Class mail without a warrant?
Or words to that effect. Alberto's response was pointed without meaning to be:
Well... I won't address all of the outrageous theories that are floating around in the media right now.
If anyone knows where one can find the exchange I'm sure we'd all appreciate it.
Inside your function "TopLevel" block, you still have to qualify your variable assignment of "count" as being a member of "self." In a language with a proper namespace, this qualification is automatic if you use a variable name that is not declared in any parent namespace.
Also, languages with namespaces almost always have a way of "importing" a namespace permanently into the top scope, so you can say:
using namespace TopLevel; count = 1;
Thus, TopLevel::count is assigned, and if you wanted to keep a separate one in the top namespace, you do main::count (er words to that effect, I am not a C++ coder). If you wanted to declare a buncha functions in a scope, using your technique, you'd have to state them in long form every time you wanted to use them, which is equivalent to just giving them a long prefix-name.
You're absolutely right, it's syntactic sugar, and namespaces and objects and execution contexts are similar things, but programmers use them in different ways, and syntax should fit use hand-in-glove.
I've always wondered if you raised a kid the right way if he would be able to have a quantum intuition.
"Paging Mr. Dick, Paging Mr. Phillip K. Dick, you have a visitor at the front desk."
An individual with a quantum-intuitive understanding of the world might be very difficult for the rest of us to recognize. Such a person would have a lot of trouble perceiving cause and effect in the way we do, and would probably have no concept of determinism or even certainty. They would be able to see more dimensions than us (if such theories are physical), and would be unable to correlate these dimensional relations to objects within our understanding: if you are a sphere, you can describe yourself to a plane by saying "I'm a bunch of circles," but this really is incomplete and the plane really would be hopeless to have a complete understanding of you. Such a person may appear at times clairvoyant or at least extremely intelligent, but much of the time incoherent and simply apart from the human race.
In short, such a person would either be autistic or the Mua'Dhib. Read PKD's "Martian Time-Slip" or "Dune" for examples of people with quantum knoweldge or understanding, and how is basically makes them appear mad much of the time.
Thanks for the links. I know this problem isn't proven on OS X, but based on the executive summary I'd suppose it could be an issue, so to Mac OS X people, textutil(1) can read doc and convert to txt, html, rtf, or even webarchive, so you get all the images.
Textutil is in/usr/bin on an install of OS X, and just acts as a wrapper for the OS X text word processing subsystem.
#!/usr/bin/ruby def recurse(x) x += 1 recurse x end recurse 1
If you run this, ruby will not segfault, but the interpreter will raise an Exception, so, you could do this:
begin recurse 1 rescue SystemStackError $stderr.print "recurse went to deep, continuing!...\n" end # and anything after this point will run as if there was no problem.
Just for the record, recursion is a basic tool of programming, as basic as a reference type, and certain languages, like Haskell or Scheme or Lisp, use it implicitly and derive their power from it.
IMDB to the rescue. It screens 1.85 in the US and 1.66 in Europe (that's standard for Europe). Though a non-anamorphic negative exposure is 4:3, the action is always framed to work for the wider ratios (the top and bottom gets cut off).
Amazon only lists a 1.33 version. It's possible they haven't gotten around to re-mastering it.
Modern movies tend to be either 1.85:1 ("flat") or 2.35:1 ("Scope"). Flat films are usually presented on 16:9 TVs (which is 1.76:1) with a pinch of left and right frame clipped (it's really miniscule). Scope films do have to be letterboxed on a 16:9 TV.
DVDs support a variety of resolutions which are letterboxed for the appropriate output. All of the NTSC formats worth mentioning are 480-line.
Fun fact: Almost all mass-entertainment films before the 1950s were 4:3.
Slashdot Mirror
SDComment *parent = [[SDSlashdotCommentManager defaultManager] getCommentForUid:@"17541932"]; :D */
if (parent) {
SDComment *this = [parent createReplyWithTitle:@"MOD PARENT UP"];
[parent moderateUp];
if (this) {
[this autorelease];
}
}
/* Not sure if I'm making fun or not
Also known as the world of the Spiderman movies. You know, the world where MJ is the one thrown from the bridge BUT survives, and they don't get to Gwen until the third movie.
Once you've made one, you can use it to make others. ?. Profit!
Why not just make more original episodes of Nova, dangit!
OTOH, there's tons of coverage of natural science, physics, applied science and anthropology, but in general not many shows on IT; this is remarkable considering how unlikely it is for the average person to see the Tevatron, but that person probably deals with computer viruses, data encryption and slow internets on a daily basis. There should be a Secret Life of Machines for computers. You could have:
And so on. I don't think Hunkin would do it, so we'd have to find someone suitably cheeky but computer-friendly and having an artistic streak. Maybe Woz.
Exactly. Pulsars are like VORs but without the magnetic-north synchronization signal. You can figure your velocity to or from by measuring the doppler, but you can't figure your bearing, and therefore, cannot figure your position.
By comparing the difference between your observation of a pulsars frequency and its frequency from a known reference, you can figure out your velocity to or from, but not your position. So first you need a very good reference datum, like from Earth, but even more, you need a very accurate clock on your ship so you can make good measurements of the pulsar's doppler. If you measure a few pulsars at once, you can build a resultant V from a collection of pulsars and work a solution for your V relative to your reference datum, like figuring your groundspeed in 3 dimensions, with wind in 4 dimensions. Fun fun fun.
This is good for speed, and figuring out your position this way is as easy as doing an integral of your V, but it's open-loop and little errors in your speed measurements will give you bad position fix, eventually. I was just thinking about this again today, and if you watched two pulsars at once, and measured the difference in phase between the two, given their frequency and position is known, you could plot your position along a hyperbola with the two pulsars as foci, just like LORAN. Getting a third reading would then give you a fix.
Pulsars are really a boon to celestial navigators, no doubt.
There are no OEMs of Apple-platform computers, but Apple authorized resellers regularly sell Intel Macs with Windows XP pre-installed.
Textmate?
I think you mean 100,000 km.
You wouldn't use the phase of the pulsar signal to get a distance reading, but you could use it's frequency as a way of identifying it uniquely, and then you would triangulate your position by getting a bearing on three or four pulsars; this is essentially how the starburst pattern on the Mariner Plaque is supposed to work for aliens trying to find Earth. It gives the frequency of pulsars near the earth and their bearing to Sol relative to the galactic center. The aliens would identify the pulsars by their frequency, and then use their knowledge of the pulsars' absolute locations to work backward and triangulate the position of Sol (assuming these aliens, a million years from now, are able to backtrack the locations of all the pulsars, as they will all have moved relative to Sol and will have slowed down in their spinning, causing an error in our reported frequencies).
After having written some tools on OS X that do stuff with RTF:
RTF is well documented and you can make an RTF document on all manner of platforms (I've done it in Ruby and Cocoa), but many platforms have extended RTF in their own way in order to support special features. OS X has added a few special methods to RTF files to support Mac OS X typography, and I've noticed that different versions of Word handle document attributes (like headers and page numbers) in different ways.
RTF is great if you want to make up something quick that is ONLY formatted text, but readers have all manner of different ways of interpreting the exact appearance of tables, page layouts and margins, and there doesn't seem to be any manageable common mechanism for including images or other documents, something Word and OO.org excel(pun) at. Even HTML seems to be better at this.
I use RTF output in a few little in-house tools I have, so people can get the text+attributes they create and open them in a text editor of their choice for touching-up and delivery. When my tools have to create something that is supposed to be finished, they make PDFs.
RTF is great for interoperability, but I never expect an RTF file to contain a "finished product," unless the recipient expects quality on par with a Selectric. It is merely a relatively-open serialization format for strings with attributes.
It's not calling curl or the shell from memory, it appears (from the description) to be a return-to-libc-attack. I am not an expert on this particular thing, but a return-to-libc attack is where you use a buffer overflow to overwrite the return address of the stack frame. Under normal circumstances, the rtsp URL parser would return to his calling function, but if an overflow overwrites the return address, you can basically rewrite the stack's memory of who called the URL parser in the first place. So, instead of returning to where Quicktime called it, your computer can be tricked into returning to a different place in memory, like somewhere in libc. Libc has all kinds of dangerous functions, namely system(3), which accepts a string as an argument (which you have also put on the stack with your buffer overflow) and will run an arbitrary program on your computer (like curl, but bash and perl and ruby can do all kinds of damage).
Of note is the fact that this exploit gets around NX, because your payload need not be executable, it merely is a return address and a string to pass into libc. Also of note is that this exploit does not cause privilege escalation; any processes started by the exploit will run under the privileges of the user who clicks on the file, and you will still get a sudo-dialog if the sploit tries to do things as wheel.
If I am misreading this exploit, please correct me. They say "arbitrary code execution" in the summary.
D doesn't appear to have meta-classes, so classes are not objects, so an operator may be a better choice for instantiating than a static method; D also appeas not to have dynamic class loading, either. If you're all about the Class.new or [Class new] way of life, you may have to stick with Ruby or Objective-C or Smalltalk.
Thanks, looks like good stuff. Now I just have to wait a year for the Cocoa bindings! :D
An interesting point, theoretically, but if you write a compiler for a language that has an eval() function (basically any dynamic interpreted language) you would have to include a complete compiler for the language in the language's runtime environment. This also applies to any language that does call- or pass-by-name (since this is just an application of eval() ).
You could compile such programs, but the runtime infrastructure (on any Intel/PowerPC type machine) would be huge, with dispatch tables serialized into the binary to allow relinking names back to memory locations, and a JIT compiler to allow eval to work. The call stack would probably have to be abstracted, and the thing just starts looking like a platform-optimized interpreter, with the added requirement that you have to compile your scripts before running them. (It actually starts looking like Objective-C, JUST KIDDING! :P )
This applies to current hardware arch, but if you built a CPU that, say, ran Java natively, or at least handled strings of arbitrary length as quickly as an x86 handles a UInt32, then an compiler of a dynamic language begins to make more sense.
Intersting. How are public interfaces to private implementations declared, and how are mixins written?
On dsource.org I can't find any examples of public headers into libraries.
Just remember, for every moron that tries to light a campfire in an ordinance dump, China kills a democrat, a stray not-so-smart-bomb kills an Iraqi math whiz, and a chemical plant releases phosgene into a river valley.
Stupid people kill themselves a hundreth as often as they kill others, either by accident or malice. And in the second case, it's always the smart people they kill first.
It's an interesting suggestion, but you have to download an entire HTML-XHTML page before the entities can be interpreted, so even if your browser won't display nsfw content, you will still have transferred it and any software tracking your downloads will catch you looking at stuff, even if you didn't see it; the data would definitely find its way onto the proxy server, unless the proxy server honored the nsfw tag and didn't cache these. An "nsfw" attribute on an [a href="" rel="nsfw"] would be the only way to screen your content before you pull it - there could be a firefox plugin that turns these links red, or puts a stop sign next to them, er something.
It's one thing for your boss to see you with a dirty web page, another for them to have a record of you going to it. This solution does nothing for the latter.
Apple already did number 1, it was called "iTV" at WWDC. (You said "announce plans" and that's exactly what they did.)
How about this one: In the wake of an accounting scandal, Apple is found guilty corporately of fraud and is broken-up into an Computer Systems company and a media delivery company. It'd be ironic that after all these years Apple got broken for shady business practices before you-know-who.
That is a good theory, but in the US every 2 or 6 years our parliamentarians are hit with a multi-million dollar liability, known as a "Campaign." The legislators are up to themselves to raise the money for these, and there are no caps on how much they or their opponent can spend.
If you want to make legislators bribe-proof, you have to make it so that they need for no money in the course of their work, which means paying them well, enough to maintain a domicile in the capital, and strictly capping campaign spending (capping fundraising, and all the exceptions and codicils on that, and the attendant free-expression issues, gets more and more unworkable all the time).
I would say all campaigns should be publicly funded, private donations forbidden, and equal money to the top 3 primary victors, but most Americans consider a campaign donation a form a free speech, and thus beyond legitimate restraint. (I think this is bullshit, but there we are.)
I forget the date, it might have been in '05 even, but when they held the first Senate hearings after the Terrorist Surveillance Program (also known as the NSA Warrantless Wiretap program), Charles Shumer asked Alberto Gonzalez, point blank:
Or words to that effect. Alberto's response was pointed without meaning to be:
If anyone knows where one can find the exchange I'm sure we'd all appreciate it.
Well.... just an observation:
Inside your function "TopLevel" block, you still have to qualify your variable assignment of "count" as being a member of "self." In a language with a proper namespace, this qualification is automatic if you use a variable name that is not declared in any parent namespace.
Also, languages with namespaces almost always have a way of "importing" a namespace permanently into the top scope, so you can say:
Thus, TopLevel::count is assigned, and if you wanted to keep a separate one in the top namespace, you do main::count (er words to that effect, I am not a C++ coder). If you wanted to declare a buncha functions in a scope, using your technique, you'd have to state them in long form every time you wanted to use them, which is equivalent to just giving them a long prefix-name.
You're absolutely right, it's syntactic sugar, and namespaces and objects and execution contexts are similar things, but programmers use them in different ways, and syntax should fit use hand-in-glove.
"Paging Mr. Dick, Paging Mr. Phillip K. Dick, you have a visitor at the front desk."
An individual with a quantum-intuitive understanding of the world might be very difficult for the rest of us to recognize. Such a person would have a lot of trouble perceiving cause and effect in the way we do, and would probably have no concept of determinism or even certainty. They would be able to see more dimensions than us (if such theories are physical), and would be unable to correlate these dimensional relations to objects within our understanding: if you are a sphere, you can describe yourself to a plane by saying "I'm a bunch of circles," but this really is incomplete and the plane really would be hopeless to have a complete understanding of you. Such a person may appear at times clairvoyant or at least extremely intelligent, but much of the time incoherent and simply apart from the human race.
In short, such a person would either be autistic or the Mua'Dhib. Read PKD's "Martian Time-Slip" or "Dune" for examples of people with quantum knoweldge or understanding, and how is basically makes them appear mad much of the time.
Thanks for the links. I know this problem isn't proven on OS X, but based on the executive summary I'd suppose it could be an issue, so to Mac OS X people, textutil(1) can read doc and convert to txt, html, rtf, or even webarchive, so you get all the images.
Textutil is in /usr/bin on an install of OS X, and just acts as a wrapper for the OS X text word processing subsystem.
Just to offer the alternate case:
If you run this, ruby will not segfault, but the interpreter will raise an Exception, so, you could do this:
Just for the record, recursion is a basic tool of programming, as basic as a reference type, and certain languages, like Haskell or Scheme or Lisp, use it implicitly and derive their power from it.
IMDB to the rescue. It screens 1.85 in the US and 1.66 in Europe (that's standard for Europe). Though a non-anamorphic negative exposure is 4:3, the action is always framed to work for the wider ratios (the top and bottom gets cut off).
Amazon only lists a 1.33 version. It's possible they haven't gotten around to re-mastering it.
Modern movies tend to be either 1.85:1 ("flat") or 2.35:1 ("Scope"). Flat films are usually presented on 16:9 TVs (which is 1.76:1) with a pinch of left and right frame clipped (it's really miniscule). Scope films do have to be letterboxed on a 16:9 TV.
DVDs support a variety of resolutions which are letterboxed for the appropriate output. All of the NTSC formats worth mentioning are 480-line.
Fun fact: Almost all mass-entertainment films before the 1950s were 4:3.